University of Southern California Dissertations and Theses

Impact of language and culture on a neuropsychological screening battery for Hispanics

(USC Thesis Other)

Impact of language and culture on a neuropsychological screening battery for Hispanics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI fiinrss
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction..
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9" black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IMPACT OF LANGUAGE AND CULTURE
ON A NEUROPSYCHOLOGICAL SCREENING BATTERY
FOR HISPANICS
by
Marta Elena Corona-LoMonaco
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement of the Degree
DOCTOR OF PHILOSOPHY
(EDUCATION - COUNSELING PSYCHOLOGY)
May 2000
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UM I Number: 3017996
Copyright 2001 by
Corona-LoMonaco, Marta Elena
All rights reserved.
___ ®
UMI
UMI Microform 3017996
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY O F SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
( V W t t l Itt-u -L Cvrovm . - L o ( V l o n A L C '
under the direction of h&C. Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Dean of Graduate Studies
D ate..........................................
DISSERTATION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
ii
A special thanks is extended to all the people who have contributed to my
personal and professional growth.
To my family: To my grandfather, Agripino Diaz, who was the first
migrant in our family to this country. I will always admire your courageous
decision to commit to the U.S. Bracero Program. It is through your efforts that my
dreams became possible here in the United States.
To all the wonderful parents and families who have inspired me in unique
ways: To my parents and grandparents (Lupe, Rigoberto, Balbina, Agripino), the
Corona-Diaz family, my “Tia Maria”: Thank you for the experience of Mexico, its
values, traditions, culture, language, for religion, and most importantly, family. To
the Wynns: Thank you for your parenting, for the experience of family, for
sharing your Jewish traditions, for teaching me English, for giving me dreams and
the encouragement to pursue them, for the value of an education. To the Bishops:
Thank you for parenting me through the challenges of young adulthood. For your
love, support, encouragement, and constant presence. For all the wonderful
memories. To all my parents: Thank you for the bicultural experience which
represents who I am (personally and professionally). This diverse background
inspired my curiosity and interest to pursue this dissertation, and more importantly,
my career.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
To the women in my life: Thank you Christina, Vicki, Andrea, Cynthia,
Rebekah, Deborah, and Marilyn, for making each stage of my personal and
academic growth so meaningful. A special thanks to Christina, Marilyn, and
Rebekah for your extra-long-distance support.
To my husband: Thank you for believing in me and supporting me (in so
many ways). For facing the challenges of a doctorate with me (all of them).
For your view of life, your love, and your friendship. I dedicate this dissertation to
you.
To the Harvard community: The Latino Team, my supervisors, mentors,
and colleagues, thank you for your constant support and encouragement.
To my dissertation committee: To my dissertation committee: Dr.
Newcomb, for your long-term support and for your constructive criticism. Dr.
Abreau, for your insightful suggestions. Dr. Jellison, for so many years of
mentorship, for sharing your knowledge, for all of your encouragement and
support, for your sincere interest in me (my past and my future), and for being a
part of the committee. Dr. Felix-Ortiz, for your knowledge of culture and for
sharing it with me, thank you. Dr. Thompson, thank you for believing in me and
for furthering my interest in cross-cultural literature and research. Dr. Rhodes,
thank you for introducing me to psychometrics (neuropsychology) and for giving
me a strong clinical foundation.
A special thank you to the NeSBHis Research Team, especially Dr. Marcel
Ponton and Dr. Lorenzo Herrera. Thank you for all the professional and cultural
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
growth you have inspired in me, for supporting this project, and for
dedication to the Latino community.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ..........................................................................................ii
LIST OF TABLES..................................................................................................... ix
LIST OF FIGURES......................................................................................................xvi
ABSTRACT................................................................................................................. xvii
Chapter
1. INTRODUCTION AND LITERATURE REVIEW.......................................1
Importance of Study............................................................................................ 1
Historical Overview............................................................................... 2
L anguage................................................................................................ 5
Moderator Variables .............................................................................7
Acculturation ......................................................................................... 9
Cultural Identity................................................................................... 12
Acculturative S tress............................................................................ 13
Neuropsychological A ssessm ent....................................................... 13
Summary and C ritique..................................................................................... 17
Purpose of Study...............................................................................................20
Research Questions and Hypotheses...............................................................22
Research Question 1 - Demographic Differences on
NeSBHis-R Domain Performance: To What Extent
Will Demographic Differences Among the Current
Sample Influence their Performance on NeSBHis-R
Subtests?.................................................................................................22
Research Question 2 - Influence of Objective
Language Categories on NeSBHis-R Domain
Performance: To What Extent Will Distinct
Objective Language Skills Between Spanish-English
Bilinguals and Spanish Monolinguals Explain
NeSBHis-R Subtest Scores?...............................................................22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Research Question 3 - Influence o f Objective
Language Categories on NeSBHis-R Domain
Performance While Controlling for Gender, Age,
and Total Years of Education: To What Extent Does
the Objective Language Category Variable Still
Explain a Significant Amount of the Variance on
NeSBHis-R Subtest Scores Following the Control of
Gender, Age, and Total Years of Education?.................................. 24
Research Question 4 - Influence of Culture Variables
on NeSBHis-R Domain Performance: To What
Extent Does Culture (e.g., Language Use/Exposure,
Behavior, Values) Explain a Significant Amount of
the Variance on NeSBHis-R Subtest Scores?................................. 24
Research Question 5 - Influence o f Culture Variables
on NeSBHis-R Domain Performance While
Controlling for Gender, Age, and Total Years of
Education: To What Extent do Culture Variables
Still Explain a Significant Amount o f the Variance
on NeSBHis-R Subtest Scores Following the Control
of Gender, Age, and Total Years of Education? ............................25
2. METHODOLOGY.......................................................................................... 27
Participants and Procedures........................................................................... 27
Instruments...................................................................................................... 33
Demographics and Background Information.................................. 33
Woodcock-Johnson Language Proficiency Battery-Revised 34
Cultural Identity Scales..................................................................... 35
Short Acculturation Scales for Hispanics........................................36
Hispanic Stress Inventory .................................................................37
Neuropsychological Screening Battery for Hispanics-Revised ... 38
3. RESULTS.........................................................................................................44
Normative D a ta ...............................................................................................45
Analyses of Hypotheses................................................................................. 46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vu
Research Question 1 - Demographic Differences on
NeSBHis-R Domain Performance: To What Extent
Will Demographic Differences Among the Current
Sample Influence their Performance on NeSBHis-R
Subtests?...............................................................................................46
Hypothesis.............................................................................. 46
Research Question 2 - Influence of Objective Language
Categories on NeSBHis-R Domain Performance: To
What Extent Will Distinct Objective Language Skills
Between-Spanish-English Bilinguals and Spanish
Monolinguals Explain NeSBHis-R Subtest Scores?..................... 54
Hypothesis A ..........................................................................54
Hypothesis B ...........................................................................59
Hypothesis C...........................................................................61
Hypothesis D ..........................................................................61
Hypothesis E ...........................................................................62
Research Question 3 - Influence of Objective Language
Categories on NeSBHis-R Domain Performance While
Controlling for Gender, Age, and Total Years of
Education: To What Extent Does the Objective
Language Category Variable Still Explain a Significant
Amount of the Variance on NeSBHis-R Subtest Scores
Following the Control of Gender, Age, and Total Years
of Education?....................................................................................... 73
Hypothesis ............................................................................. 73
Research Question 4 - Influence of Culture Variables on
NeSBHis-R Domain Performance: To What Extent
Does Culture (e.g., Language Use/Exposure, Behavior,
Values) Explain a Significant Amount of the Variance
on NeSBHis-R Subtest Scores?......................................................... 79
Hypothesis A ..........................................................................81
Hypothesis B ........................................................................... 84
Hypothesis C ........................................................................... 84
Hypothesis D ..........................................................................84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
viii
Research Question 5 - Influence o f Culture Variables on
NeSBHis-R Domain Performance While Controlling for
Gender, Age, and Total Years of Education: To what
extent do Culture Variables Still Explain a Significant
Amount of the Variance on NeSBHis-R Subtest Scores
Following the Control of Gender, Age, and Total Years
o f Education?......................................................................................115
Hypothesis ............................................................................ 115
4. DISCUSSION.................................................................................................. 142
Introduction.....................................................................................................143
Interpretation of R esults................................................................................ 146
NesBHis-R as Compared to Three Demographic
Variables (Gender, Age, and Education) (Research
Question 1)..........................................................................................146
NeSBHis-R as Compared to Objective Language
Categories(Research Questions 2 and 3 ) ........................................ 151
NeSBHis-R as Compared to Culture (Research
Questions 4 and 5 ) ............................................................................. 175
Theoretical Implications.................................................................................185
Clinical Implications.......................................................................................187
Limitations of Study........................................................................................192
Future Directions............................................................................................. 194
REFERENCES............................................................................................................. 199
APPENDICES
A. Sample Flier......................................................................................................208
B. Language Proficiency-Examiner’s Form ..................................................... 209
C. Examiner’s Scoring S heet..............................................................................210
D. Demographics Questionnaire, English Version...........................................213
E. Reasons for Migration to the United States Category: “Other” .............. 221
F. NeSBHis-R Means and Standard Deviations for Women by Age
and years of Education...................................................................................222
G. NeSBHis-R Means and Standard Deviations for Men by Age
and years of Education...................................................................................224
H. NeSBHis-R Domains......................................................................................226
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ix
LIST OF TABLES
Table Page
1. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Gender, Age, and Total Years of Education.................................................47
2. Comparison o f NeSBHis-R Language Domain Subtests by
Gender, Age, and Total Years of Education................................................. 48
3. Comparison o f NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Gender, Age, and Total
Years of Education...........................................................................................50
4. Comparison o f NeSBHis-R Memory Domain Subtests by
Gender, Age, and Total Years of Education................................................. 51
5. Comparison o f NeSBHis-R Visual-Spatial Subtests by Gender,
Age, and Total Years of Education................................................................52
6. Comparison o f NeSBHis-R Motor Subtests by Gender, Age,
and Total Years of Education.........................................................................53
7. Analysis of Variance: Comparison of Objective Language
Categories and Subjective Language Ratings...............................................60
8. Analysis of Variance: Mean Comparison of NeSBHis-R
Reasoning Domain Subtest by Five Objective Language
Categories..........................................................................................................63
9. Analysis of Variance: Mean Comparison of NeSBHis-R
Language Domain Subtests by Five Objective Language
Categories..........................................................................................................65
10. Analysis of Variance: Mean Comparison of NeSBHis-R
Executive Function/Attention-Concentration Domain Subtests
by Five Objective Language Categories........................................................66
11. Analysis of Variance: Mean Comparison of NeSBHis-R
Memory Domain Subtests by Five Objective Language
Categories..........................................................................................................68
12. Analysis of Variance: Mean Comparison of NeSBHis-R
Visual-Spatial Domain Subtests by Five Objective Language
Categories..........................................................................................................70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
13. Analysis of Variance: Mean Comparison of NeSBHis-R Motor
Domain Subtests by Five Objective Language Categories........................ 71
14. Analysis of Covariance: Comparison of NeSBHis-R Reasoning
Domain by Objective Language Categories, Controlling for
Gender, Age, and Total Years o f Education.................................................74
15. Analysis of Covariance: Comparison of NeSBHis-R Language
Domain Subtests by Objective Language Categories,
Controlling for Gender, Age, and Total Years of Education...................... 75
16. Analysis of Covariance: Comparison of NeSBHis-R Executive
Function/Attention-Concentration Domain Subtests by
Objective Language Categories, Controlling for Gender, Age,
and Total Years of Education.........................................................................76
17. Analysis of Covariance: Comparison o f NeSBHis-R Memory
Domain Subtests by Objective Language Categories,
Controlling for Gender, Age, and Total Years of Education...................... 77
18. Analysis of Covariance: Comparison of NeSBHis-R Visual-
Spatial Domain Subtests by Objective Language Categories,
Controlling for Gender, Age, and Total Years of Education...................... 77
19. Analysis of Covariance: Comparison of NeSBHis-R Motor
Domain Subtests by Objective Language Categories,
Controlling for Gender, Age, and Total Years of Education...................... 78
20. Reliability Analysis: Cultural Identity Scales, Language
Domain............................................................................................................. 81
21. Reliability Analysis: Cultural Identity Scales,
Behavior/Familiarity Domain.........................................................................82
22. Reliability Analysis: Cultural Identity Scales, Value/Attitude
Dom ain............................................................................................................. 83
23. Reliability Analysis: Marin Short Acculturation Scale for
Hispanics..........................................................................................................83
24. Comparison of NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Language Domain.................................................85
25. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Behavior/Familiarity Domain..............................86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X I
26. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Value/Attitude Dom ain.........................................86
27. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Marin Variables..................................................................................................87
28. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales and Marin Clusters..................................................87
29. Comparison o f NeSBHis-R Language Domain Subtests by
Cultural Identity Scales Language Domain................................................... 89
30. Comparison o f NeSBHis-R Language Domain Subtests by
Cultural Identity Scales, Behavior/Familiarity Domain...............................90
31. Comparison o f NeSBHis-R Language Domain Subtests by
Cultural Identity Scales, Value/Attitude Dom ain.........................................90
32. Comparison o f NeSBHis-R Language Domain Subtests by
Marin Variables..................................................................................................91
33. Comparison o f NeSBHis-R Language Domain Subtests by
Cultural Identity Scales and Marin Clusters..................................................92
34. Comparison o f NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales,
Language Domain............................................................................................. 95
35. Comparison o f NeSBHis-RExecutive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales,
Behavior/Familiarity Domain.......................................................................... 96
36. Comparison o f NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Culture Identity Scales
Value/Attitude Dom ain.................................................................................... 97
37. Comparison o f NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Marin Variables.................................. 98
38. Comparison o f NeSBHis-RExecutive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales and
Marin Domains................................................................................................100
39. Comparison o f NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales, Language Domain................................................ 101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xii
40. Comparison o f NeSBHis-R Memory Domain Subtests by
Culture Identity Scales Behavior/Familiarity Domain Variables.............102
41. Comparison o f NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales, Value/Attitude Dom ain....................................... 102
42. Comparison o f NeSBHis-R Memory Domain Subtests by Marin
Variables........................................................................................................... 103
43. Comparison o f NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales and Marin Clusters................................................ 103
44. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Cultural Identity Scales, Language Domain................................................ 105
45. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Cultural Identity Scales, Behavior/Familiarity Dom ain.............................106
46. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Cultural Identity Scales, Value/Attitude Dom ain....................................... 107
47. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Marin Variables................................................................................................107
48. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Cultural Identity Scales and Marin Clusters................................................ 108
49. Comparison o f NeSBHis-R Motor Domain Subtests by Cultural
Identity Scales, Language Domain................................................................ 109
50. Comparison o f NeSBHis-R Motor Domain Subtests by Cultural
Identity Scales, Behavior/Familiarity Dom ain............................................110
51. Comparison o f NeSBHis-R Motor Domain Subtests by Cultural
Identity Scales, Value/Attitude Domain....................................................... 111
52. Comparison o f NeSBHis-R Motor Domain Subtests by Marin
Variables........................................................................................................... 112
53. Comparison o f NeSBHis-R Motor Domain Subtests by Cultural
Identity Scales and Marin Clusters..............................................................112
54. Comparison o f NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Language Domain, Controlling for
Gender, Age, and Total Years of Education................................................ 116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X lll
55. Comparison of NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Behavior/Familiarity Domain,
Controlling for Gender, Age, and Total Years o f Education.....................117
56. Comparison of NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales, Value/Attitude Domain, Controlling
for Gender, Age, and Total Years o f Education........................................ 118
57. Comparison of NeSBHis-R Reasoning Domain Subtest by
Marin Variables, Controlling for Gender, Age, and Total Years
of Education....................................................................................................118
58. Comparison of NeSBHis-R Reasoning Domain Subtest by
Cultural Identity Scales and Marin Clusters, Controlling for
Gender, Age, and Total Years of Education...............................................119
59. Comparison of NeSBHis-R Language Domain Subtests by
Cultural Identity Scales, Language Domain, Controlling for
Gender, Age, and Total Years of Education...............................................120
60. Comparison of NeSBHis-R Language Domain Subtests by
Cultural Identity Scales, Behavior/Familiarity Domain,
Controlling for Gender, Age, and Total Years of Education.....................121
61. Comparison of NeSBHis-R Language Domain Subtests by
Cultural Identity Scales, Value/Attitude Domain, Controlling
for Gender, Age, and Total Years o f Education.........................................122
62. Comparison of NeSBHis-R Language Domain Subtests by
Marin Variables, Controlling for Gender, Age, and Total Years
of Education.................................................................................................... 123
63. Comparison of NeSBHis-R Language Domain Subtests by
Cultural Identity Scales and Marin Clusters, Controlling for
Gender, Age, and Total Years o f Education................................................ 124
64. Comparison of NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales
Language Domain, Controlling for Gender, Age, and Total
Years of Education......................................................................................... 126
65. Comparison of NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales,
Behavior/Familiarity Domain, Controlling for Gender, Age,
and Total Years o f Education....................................................................... 127
with permission of the copyright owner. Further reproduction prohibited without permission.
X IV
66. Comparison of NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales,
Value/Attitude Domain, Controlling for Gender, Age, and
Total Years ofEducation............................................................................... 128
67. Comparison of NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Marin Variables,
Controlling for Gender, Age, and Total Years ofEducation.....................129
68. Comparison of NeSBHis-R Executive Function/Attention-
Concentration Domain Subtests by Cultural Identity Scales and
Marin Clusters, Controlling for Gender, Age, and Total Years
ofEducation.....................................................................................................130
69. Comparison of NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales, Language Domain, Controlling for
Gender, Age, and Total Years ofEducation................................................132
70. Comparison of NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales, Behavior/Familiarity Domain,
Controlling for Gender, Age, and Total Years ofEducation.....................132
71. Comparison o f NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales Value/Attitude Domain, Controlling for
Gender, Age, and Total Years ofEducation................................................133
72. Comparison of NeSBHis-R Memory Domain Subtests by Marin
Variables, Controlling for Gender, Age, and Total Years of
Education......................................................................................................... 133
73. Comparison of NeSBHis-R Memory Domain Subtests by
Cultural Identity Scales and Marin Clusters, Controlling for
Gender, Age, and Total Years ofEducation................................................134
74. Comparison of NeSBHis-R Visual-Spatial Domain Subtests by
Culture Identity Scales Language Domain, Controlling for
Gender, Age, and Total Years ofEducation................................................135
75. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Culture Identity Scales Behavior/Familiarity Domain,
Controlling for Gender, Age, and Total Years ofEducation..................... 135
76. Comparison of NeSBHis-R Visual-Spatial Domain Subtests by
Culture Identity Scales Value/Attitude Domain, Controlling for
Gender, Age, and Total Years ofEducation................................................136
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
XV
77. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Marin Variables, Controlling for Gender, Age, and Total Years
ofEducation..................................................................................................... 136
78. Comparison o f NeSBHis-R Visual-Spatial Domain Subtests by
Cultural Identity Scales and Marin Clusters, Controlling for
Gender, Age, and Total Years ofEducation................................................ 136
79. Comparison o f NeSBHis-R Motor Domain Subtest by Cultural
Identity Scales Language Domain, Controlling for Gender, Age,
and Total Years ofE ducation........................................................................138
80. Comparison o f NeSBHis-R Motor Domain Subtest by Cultural
Identity Scales, Behavior/Familiarity Domain, Controlling for
Gender, Age, and Total Years ofEducation................................................ 138
81. Comparison o f NeSBHis-R Motor Domain Subtest by Cultural
Identity Scales Value/Attitude Domain, Controlling for Gender,
Age, and Total Years ofEducation...............................................................139
82. Comparison o f NeSBHis-R Motor Domain Subtest by Marin
Variables, Controlling for Gender, Age, and Total Years of
Education..........................................................................................................139
83. Comparison o f NeSBHis-R Motor Domain Subtest by Cultural
Identity Scales and Marin Clusters, Controlling for Gender,
Age, and Total Years ofEducation...............................................................140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
XVI
LIST OF FIGURES
Figure Page
1. Short Acculturation Scale for Hispanics................................................................11
2. Cultural Identity Scales............................................................................................13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xvii
ABSTRACT
Purpose
Due to the lack of appropriate tests and norms for use with the Hispanic
population in the United States, misdiagnosis o f cognitive functioning is very
common in research and clinical settings. The current study presents both
normative data and an assessment model to be used with Hispanic populations.
The current results provide a unique contribution to the assessment of Hispanic
patients by providing preliminary normative data which controls for acculturative
stress while clarifying the influence that language proficiency and culture have on
neuropsychological test performance. The categorization of language dominance
groups is operationalized and two models of culture measurement are tested.
Method
Seventy-eight Hispanic subjects were tested with the NeSBHis-R and the
WJLPB-R. Subjects completed a demographics questionnaire and three
standardized measures of culture: acculturation (Marin), cultural identity (CIS),
and acculturative stress (HIS). Subjects were recruited with fliers from churches
and community centers in Los Angeles. All subjects were volunteers and did not
receive compensation for participation. The final sample included 43 men and 35
women, and the largest ethnic representation was Mexican (n=35). Ages ranged
ffoml6 to 65, most were first generation (n=66), right-handed individuals (=76),
and total years of education ranged from 0 - 20.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xvm
Results
The current study confirmed five objective language categories. When
these groups were compared to performance on the NeSBHis-R, significant
ANOVA and ANCOVA results were noted on all six Domains (Nonverbal
Reasoning, Language, Memory, EF-AC, Visual-Spatial, and Motor), and
following the control of Gender, Age, and Education, the Nonverbal Reasoning,
Language, Memory, and EF-AC Domains remained significant. The identified
Culture measures (CIS and Marin) proved to be reliable instruments, and when
compared to NeSBHis-R Domains, Pearson r correlations showed that culture
contributed to all six Domains, and following the control for Gender, Age, and
Education, the Nonverbal Reasoning, Language, and EF-AC Domains remained
significant. Also, partial support was established for the Multidimensional model
of culture measurement. The need to include moderator variables as part of
neuropsychological test batteries when testing Latino/a clients is supported and
highly recommended.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CH APTER 1
INTRODUCTION AND LITERATURE REVIEW
1
Importance o f Study
According to the 1990 Census Bureau, Hispanics constitute approximately
23 million people in the United States (Prewitt, 1999). The Director of the Bureau
of the Census further reports that the 1999 estimate o f the Hispanic population is
31 million, suggesting that “The Hispanic population increased by more than 9.1
million people over this 9-year period—more than any other group” (Prewitt,
1999, p. 381). Yet, adequate psychological assessment instruments for the study
of Hispanics are limited and usually rely on small sample sizes which do not allow
for stratification by age and education. For instance, appropriate measures in the
Spanish language, adequate Hispanic norming samples, as well as research with
Spanish-speaking populations in the United States, are scarce and underutilized
(Ponton, Satz, Herrera, Ortiz, Urrutia, Young, D'Elia, Furst, & Namerow, 1996).
As a result, when working with Spanish-speaking populations, examiners have
relied on direct instrument translations and the use o f third-party interpreters
during the psychological assessment. However, due to varying levels of
bilingualism, cultural differences, and acculturation issues, these methods have
been identified by several clinicians and researchers as a potential for misdiagnosis
(e.g., false positives, false negatives) (Cervantes & Acosta, 1992; Geisinger, 1992;
LaCalle, 1987; Mahurin, Espino, & Holifield, 1992; Ponton et al., 1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
Other issues which complicate the Spanish language and, therefore, the
translation process include not only Hispanics1 differences in national, ethnic, and
racial lines, but also differences in regional subcultures, traditions, values, and
dialects (Mahurin, Espino, & Holifield, 1992). Beyond the uniqueness of the
Hispanic experience, there is also limited multicultural assessment training of
mental health professionals (Dana, 1993) and even fewer Hispanic mental health
professionals trained in multiculturalism (1.5%) (Cervantes & Acosta, 1992), yet
many Hispanics continue to be assessed (Geisinger, 1992).
Historical Overview
Historically, psychological assessment can be traced back to the late 1700s
when interests in individual differences and mental abilities were first being
recognized. It was during the 1800s, that individuals such as Cattel, Wundt, and
Galton proposed theories regarding the objective measurement o f mental abilities.
By the 20th century (1905), the Binet-Simon Intelligence Test was developed and
Turman coined the concept of intelligence quotient (1916) (Reynolds &
Kamphaus, 1990). Other forms of assessment followed shortly after, such as the
clinical interview and personality assessment. More recently, psychological
assessment in the 21st century has expanded and now includes neuropsychological
assessment, behavioral assessment, stress and coping, and adaptation to new
cultures, to name a few (Groth-Mamat, 1990). The assessment o f language
proficiency has also been expanded within the field and includes both linguistic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
and psychological theories. For instance, some clinicians and researchers
conceptualize “language proficiency” as four components of communicative
competence (e.g., grammatical competence, discourse competence, sociolinguistic
competence, and strategic competence) (Sandoval & Duran, 1998).
Hence, the birth and growth of psychological assessment over the past
three centuries has resulted in both a thriving industry and a source of new insight
regarding human behavior and mental abilities. However, this large contribution
has also generated general concerns and problems in both academic and clinical
settings. In the 1970s and early 1980s, class-action suits were filed against the
validity of intelligence tests when a disproportionate amount of Black students
were assigned to educatively mentally retarded classrooms, and a number of
Mexican-American children were misdiagnosed as mentally retarded (Reynolds &
Kamphaus, 1990). Similar concerns were also noted with adult minority
populations when confounds were identified between cultural identity and
psychopathology on the M M PII & II across non-white samples (Dana, 1995).
Also, while the Wechsler Adult Intelligence Scale - Revised (WAIS-R) led to
underestimations of cognitive intelligence levels among Spanish-speaking adults,
the Escala de Inteligencia Wechsler para Adultos (EIWA) (Spanish form of the
WAIS-R) led to overestimations of intelligence (Lopez & Taussig, 1991; Lopez &
Romero, 1988). Academically, standardized aptitude and achievement tests have
also been criticized as biased against minority students (Geisinger, 1992), and the
accuracy o f predicting academic achievement in higher education with aptitude
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
tests has shown to be confounded by English and Spanish language proficiency,
family social economic status, and education background (Geisinger, 1992).
The misdiagnosis of minorities and the identification o f instrument
confounds has, however, spurred great interest in the area o f test bias and how it
relates to psychological assessment instrumentation (e.g., test construction, norms,
and item content), process (e.g., administration and interpretation), and cross-
cultural influences (e.g., language and cultural identity). Test bias is present
"when a test score has different meanings or implications for a relevant, definable
subgroup o f test takers that are different from the meanings or implications of the
remainder o f the test takers" (Geisinger, 1992, p. 25). These issues are
particularly relevant with the Hispanic population which is not only culturally but
linguistically different as well.
In response to the plethora of psychometric concerns, ethical codes
incorporated general guidelines which were sensitive to the areas of culture,
language, and test instrumentation (Standards, 1985; APA, 1992; Prediger, 1994).
However, the AP A code has also been criticized for not providing examples of
culturally sensitive clinical practice and for not addressing the current controversy
and development of multicultural assessment interventions (Dana, 1994). Similar
limitations apply for the American Counseling Association standards (Prediger,
1994) and the Standards for Educational and Psychological Testing (1985). Given
the field’s consistent concern about issues of language, culture, and test
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
instrumentation, these topics will now be discussed in greater detail. Theories and
research are reviewed and emphasis is placed on the Hispanic population.
Language
It is important to consider how, for Hispanics, language could involve any
of the following categories: Spanish monolingual, Spanish-dominant bilingual,
English-dominant bilingual, equally balanced bilingual, and English monolingual
(Rodriguez, 1992). Hence, the complexity of language among Hispanics raises
considerable challenges in the process of psychometric test selection,
administration, and interpretation. Providing that the clinician is bilingual, the first
challenge involves deciding how much o f the test battery will consist of English
and/or Spanish language tests. Next, is the challenge o f test interpretation which
inevitably involves an assessment o f language proficiency, its influence on
language specific subtests, and its relationship to other neurocognitive domains
(e.g., nonverbal reasoning, attention, learning and memory). For example, one
must consider if performance on other neurocognitive domains (e.g., nonverbal
reasoning, attention) is directly associated with the identified construct, the level of
language proficiency, or both (Sandoval & Duran, 1998).
Specifically, a review o f the literature shows that attempts to understand
the linguistic diversity of Hispanics and other racial groups has revealed
discrepancies between monolingual and bilingual test results (Bialystok, 1986;
Cervantes & Acosta, 1992; Garcia, 1986). One of the theories that has been
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
empirically supported includes the First Language Interference Theory (Pialorsi,
1981). However, one must also consider the Threshold Theory which states that
"there may be levels of linguistic proficiency that bilingual children must attain to
avoid cognitive deficits and to allow the cognitive benefits" (Ricciardelli, 1992, p.
301). For example, it has been empirically demonstrated that overall bilingual
superiority is true only for children who have attained a high degree of
bilingualism (Ricciardelli, 1992). The Threshold Theory has also been empirically
supported in a longitudinal study with an adult Hispanic population (Bahrick, Hall,
Goggin, Bahrick, & Berger, 1994) and appears to be a widely accepted theory
among many scholars who argue that “balanced, proficient bilingualism is an
intellectual asset” (Sandoval & Duran, 1998, p. 192).
However, the study o f “bilingualism” has typically focused on children,
and particular emphasis has been placed on the assessment of metalinguistic
ability: the analysis of linguistic knowledge (crystallized ability) and the control
of linguistic processing (fluid ability) (Bialystok, 1986 & 1988). Cross-cultural
findings have shown that bilingualism, as compared to English monolingual
speakers, increased a child’s ability to solve problems involving high levels of
control (fluid ability) (Bialystok, 1986 & 1988). Differing levels of bilingualism
(high versus low bilingual groups) have also been studied and suggest differences
between groups (Bialystok, 1988; Diaz, 1985). Studies involving
neuropsychology and bilingualism are less available, but an analysis o f cognitive
complexity and attentional control of cognitive development in a child sample
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
suggests that bilingual children exhibit stronger executive functioning skills over
monolingual children (Bialystok, 1999).
Overall, developmental research has offered a great contribution to the
understanding o f multilingual ability and cognition, yet similar models have
received less attention with adult populations. For example, attempts to define
balanced, proficient bilingualism (Spanish-English) relative to neurocognitive
performance of adult Hispanics and how it differs from non-balanced bilinguals
and even monolinguals remains to be tested empirically. However, the process o f
determining what balanced, proficient bilingualism looks like relative to
neurocognitive performance and how it differs from non-balanced bilinguals and
even monolinguals remains to be tested empirically.
Moderator Variables
In addition to language proficiency, culture has also been identified as an
important variable in the consideration of learning styles, behavior, and attitudes.
Studies that have looked at both academic and professional settings suggest that
bicultural students have shown greater difficulty adapting to the school
environment because of the challenge o f cultural differences (Matute, 1986) and
that Chicano professionals experience a greater struggle between issues of family
support and educational success. On the other hand, when the cultural implication
of language was considered, bilingualism among a sample of Mexican-Americans
(N = 1,034) was associated with higher self-confidence than subjects who were not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
(this study controlled for social economic status and school experience) (Huang,
1995). Thus, it is believed that different cultural environments likely explain, in
part, the development of cognitive abilities (Ardila, 1990).
Consequently, the cross-cultural concerns that are highlighted in both
research and ethics, have encouraged professionals to discuss the importance of
including moderator variables in all assessment batteries (Dana, 1993 & 1994;
Betancourt & Lopez, 1993). Moderator variables are defined as a correction for
cultural differences that are needed to determine how much variance is accounted
for by culture in test results (Dana, 1994 & 1993). Moreover, it is believed that the
purpose of any moderator variable is to prevent genuine differences from being
ignored, disregarded, or minimized; however, methods of selecting moderator
variable instruments, and knowledge of how they interact with psychological
assessment domains (e.g., neurocognition and personality) has not been explored
empirically.
In fact, a review of the literature demonstrated that the process of selecting
moderator variables is complicated by the fact that there is not a clear
differentiation between measures o f culture. A variety o f constructs which are
associated with the measurement o f culture were identified in the literature, yet
efforts to identify similarities and differences between the constructs has not been
attempted. Some o f these constructs include: Acculturation (e.g., Cuellar, Arnold,
& Maldonado, 1995; Marin, Sabogal, Marin, Otero-Sabogal, & Perez-Stable,
1987; Montgomery & Orozco, 1984), Cultural Awareness (Arbona, Flores, &
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Novy, 1995; Keefe, 1992), Cultural Loyalty (Miranda & White, 1993), Cultural
Identity (Felix-Ortiz, Newcomb, & Myers, 1994), Ethnic Identity (Arbona et al.,
1995; Keefe, 1992; Phinney, 1992), Ethnic Loyalty (Arbona et al., 1995; Keefe,
1992; Salgado De Snyder, 1987), Acculturative Stress, and Discrimination Stress
(Cervantes, Padilla, & Salgado De Snyder, 1991). Given the multitude of culture
constructs and measures, appropriate selection of moderator variables is an unclear
procedure. Therefore, efforts to control for cultural confounds becomes
compromised. In an attempt to clarify which culture variables could satisfy the
function o f a moderator variable, all the identified constructs were reviewed, and
three constructs were selected based on: (a) Clear distinctions in the construct
being measured, and (b) the benefit such constructs could provide to
neuropsychological assessment. Focus is now placed on the identified culture
constructs: Acculturation, cultural identity, and acculturative stress.
Acculturation
Acculturation is the process by which the immigrant's (minority group's)
attitudes and behaviors change toward (adopt) those of the dominant cultural group
(majority group) as a result of exposure to the new cultural group. Thus,
acculturation (culture change) represents adaptation to another culture, which is
distinct from assimilation. Assimilation represents the absorption of another
culture with “resemblance” as its goal. Acculturation and assimilation, therefore,
are two different constructs. The acculturation construct appears to be one of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
most widely used cultural measures in research (Cuellar et al., 1995) because it has
been empirically associated with mental health status (Cuellar et al., 1995; Marin
et al., 1987) and neuropsychological test performance among Latinos (Ponton,
Satz, Herrera, Ortiz, Urrutia, Young, D’Elia, Furst, & Namerow, 1996).
However, despite the importance of these findings, there is a need to
further investigate the acculturation construct because it has been criticized for
relying heavily on language status and generation status (Negy, 1993). The
argument against the single domain of language as a measure of culture is that
learning the national language does not necessarily represent the degree to which
the individual has adopted the core values of the host culture (Negy, 1993). In
addition, language has different salience with various groups and virtually none for
some (Phinney, 1992). Hence, an assessment of the domain of language alone
only informs us about cultural awareness, and is less informative on the issue of
ethnic or cultural identity as a general acculturative experience.
It is similarly inappropriate to assume that increased generation level alone
also increases acculturation. For instance, it is important to consider the process of
enculturation which speaks to the process of later generations returning to their
roots in an attempt to learn about their culture of origin (Matsumoto, 1996).
Moreover, some individuals reside in communities where Spanish is the dominant
language and in which religion, “familism,” and the traditional Mexican cultural
values are practiced (regional differences) (Negy, 1993). In these cases, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
individual is likely to present a traditional cultural and linguistic identity regardless
o f a high generation level.
The acculturation construct has also been criticized for implying a
unidimensional process (Felix-Ortiz et al., 1994). Acculturation theory has
traditionally taken a "unidimensional/ unidirectional" perspective, which measures
acculturation on one dimension; for example, the single domain o f "language"
being measured on a single dimension from "Hispanic" to "Anglo" (Figure 1).
However, this "unidimensional/unidirectional" model
Language Use
Hispanic -------------------------------------------------------------------►Anglo
Ethnic Loyalty
Media
Hispanic ^ p. Anglo
Ethnic Social
Hispanic ^ -------------------------------------------------------------------► Anglo
Relations
Figure 1. Short Acculturation Scale for Hispanics (Marin et al., 1987)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
leads the individual on a journey toward "assimilation" and does not fully specify
alternate paths such as "biculturalism" (e.g., dual cultural experience) and
"marginalism" (e.g., feeling caught between two cultures). Assimilation operates
on the assumption that the process of becoming similar to the majority culture
requires loss o f the non-majority orientation. Assimilation does not validate
situational and biological realities such as needing both cultures to negotiate
different situations and how people of color are always distinguishable regardless
of their level o f assimilation. Assimilation also limits the possibility of an
individual negotiating more than two ethnic identities as part of their cultural
identity (e.g., Native American, Latino, and American) (Oetting & Beauvais,
1991). Therefore, current research has moved away from unidimensional models
and more toward "multidimensional/multi-domain" models in which multiple
domains (e.g., language and values) are each measured (e.g., "Low" to "High") on
multiple dimensions of ethnicity (e.g., Mexican-American and Euro-American).
The individual's experience o f their own culture is therefore measured separate
from their experience of the majority culture/host United States culture.
Cultural Identity
An example of a multidimensional/multi-domain model is the Cultural
Identity Scales (CIS) (Felix-Ortiz et al., 1994). "Cultural Identity" is a "descriptor
that allows for a process that occurs as part of personality formation versus a
process that occurs between groups" (Felix-Ortiz et al., 1994, p. 29). The CIS is a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
multidimensional/multiple domain measure because it assesses the individual's
sense of familiarity with American and Latino/a cultures separately on Language,
Behavior, and Values/Attitudes (Figure 2). In other words, the CIS provides
distinct scores for each Latino/a domain (Spanish Language, Latino/a Behavior,
Latino/a Values and Attitudes) and for each American domain (English Language,
American Behavior, American Values and Attitudes). Thus, the CIS not only adds
the multidimensional aspect to the cultural moderator, but it also provides multiple
domains. As opposed to relying on language as the main determinant of the
acculturation process, the CIS also assesses the influence of behavior and values,
which, in turn, provides a broader measure of culture.
Language
Low .4 -------------------------------------------- —--------------------High
Low -4--------------------------------------------------- ► High
Behavior
<-----------
English
-------------►
f e -
^ -------------
Spanish
^ w
American
^ -------------
Latino/a
f e -
^ -------------
<-----------
Feminism
w
-------------►
High
Low ^ ^ High
Values
Low ^ ------------------------------------------------- — ---------► High
Feminism
Low ^ _______________________________________— ► High
Respeto
Figure 2. Cultural Identity Scales (Felix-Ortiz et al., 1994)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
Acculturative Stress
While the literature shows that multidimensional models significantly
improve the original unidimensional model, the acculturative stress moderator
continues to be seen as a distinct measure which stands separate from the
“acculturation” and “cultural identity” measures. Acculturative stress is defined as
the stress associated with changing one's cultural orientation (Cervantes, Padilla, &
Salgado de Snyder, 1991). Hence, acculturative stress captures the experience of
the individual rather than the group process association with acculturation
(Salgado De Snyder, 1987). In general, the literature suggests that acculturative
stress should be measured in addition to, and separate from, other cultural
variables, such as cultural identity and acculturation (Betancourt & Lopez, 1993),
and that cultural identity should also be measured separately from acculturation
(Felix-Ortiz et al., 1994). In other words, the current literature review suggests
that each of these variables (acculturation, acculturative stress, cultural identity)
measure distinct aspects of culture (Saldana, 1995).
Neuropsychological Assessment
Neuropsychology seeks to understand how an individuals’ behavior is
influenced by different areas of the brain. Behavior encompasses three main
functions: (a) cognitive (e.g., attention, language, visual-spatial skills, memory,
planning, and organization); (b) mood (e.g., how one feels); and (c) personality
style (e.g., how one interacts with people, and disposition). A neuropsychological
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
assessment includes an extensive interview, review o f available records, the
administration of neuropsychological tests, and the interpretation/feedback/
recommendations in oral and written format. These neuropsychological
techniques are frequently applied in both research and clinical settings in order to
clarify how brain-behavior relationships manifest secondary to disease and illness.
While neuropsychology has become a strong tradition in Latin America
and Spain ( Ardila, 1990), in the United States, neuropsychology is still severely
deficient in the areas of "cross-cultural issues, bilingualism, and the applicability
of testing procedures to Hispanics or to the monolingual Spanish-speaking
population" (Cervantes & Acosta, 1992, p. 212; Franzen & Lovell, 1987;
Wedding, Horton, & Webster, 1986). Yet the influence that culturally learned
abilities have on neuropsychological test performance is a concern among
professionals in the field (e.g., Ardila, 1995). It is believed that performance on
neuropsychological tests is influenced by a variety o f moderating variables such as
culture, ecological demands, primary language, and educational level. “Test
scores are associated therefore, not only with the subject's learning opportunities,
but also with those variables which a culture dictates are worthy o f cognitive
application. Different cultural environmental contexts will result in the
development of different patterns o f abilities" (Ardila, 1995, p. 145).
Ardila (1995) addresses key focal points of neglected research in cross-
cultural neuropsychology: (a) normalization o f current basic neuropsychological
instruments, in different cultural contexts; (b) the development o f new
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
neuropsychological instruments, appropriate for different cultural contexts; (c) the
analysis of educational factors and subcultural variations in relation to test
performance; (d) the analysis o f cognitive disturbances in cases of brain pathology
in different cultural and educational contexts; (e) the search for commonality in
neuropsychological performance among existing human groups; and (f) the
analysis of the origins of cognitive activity.
The need for new models o f psychological assessment is evident, yet only
two studies were noted in the current literature review which make an effort to
respond to the numerous professional concerns. Taussig and colleagues present a
study (Taussig, Henderson, & Mack, 1992) in which 18 neuropsychological tests
were translated and administered to a Spanish-speaking sample. In this study,
significant differences were noted between the identified Alzheimer patients and
the normal comparison group. Moreover, significant differences were noted
between Spanish- and English-speaking patients on two of the subtests (Fuld
Object Memory Test and Trails A & B); the latter group showed stronger
performance.
Ponton and his colleagues (1996) also provide an important contribution to
cross-cultural neuropsychology. With the development and standardization o f the
Neuropsychological Screening Battery for Hispanics (NeSBHis), Ponton et al.
offer "the first report available in the literature that provides normative data
stratified by age, gender, and education" for Hispanics using a moderately large
sample (N = 300) of Spanish monolingual and English-Spanish bilingual subjects
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
(Ponton, et al., 1996, p. 101). Moreover, they included the Short Acculturation
Scale for Hispanics (Marin) (Marin et al., 1987) in their battery. In fact, the initial
NeSBHis findings suggest that acculturation is a significant moderator variable.
Partial Correlational analyses which controlled for age and education,
demonstrated that the Marin total scores significantly impact on the following
neuropsychological subtests: Digit Span Forward, Digit Span Backward, Digit
Span Total, Digit Symbol, Color Trails 1, Color Trails 2, and Ravens Total (Poster
Presentation, 1997).
Summary and Critique
Neuropsychological assessment has only recently experienced significant
growth as a profession, and although the most notable advancements have
occurred within the last century, little has been explored relative to ethnic and
cultural differences in test construction and interpretation. The results from the
initial studies, however, speak to the importance of addressing ethnic and cultural
differences in addition to effects of education, age, and gender. Hence, there is
consensus among research, ethical, and clinical perspectives about the important
role that culture plays in the assessment process and the need to clarify the
intervention and interpretation phases relative to cultural confounds on test
performance.
Upon review of the literature associated with the measurement of culture, it
was evident that this is an area that has received a great deal of attention.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
However, it was also noted that less focus has been devoted to understanding of
how culture variables differ (e.g., cultural identity, acculturation, and acculturative
stress) and most importantly, how to apply culture variables as moderators in
clinical and research settings. Therefore, three levels of culture measurement need
to be considered in future research.
First, as professionals, we need to develop a better understanding of the
constructs we use when referring to cultural experiences. Although numerous
constructs were identified in the literature, emphasis was placed on three:
acculturation, cultural identity, and acculturative stress. The current effort to
clarify culture definitions and provide distinct descriptions o f each construct
suggests that each would inform clinical data in a unique way. Second, culture
measurement also requires an assessment o f dimensions. Third, multiple domains
should be considered when measuring culture. For example, the need to look
beyond the domain of language is a notable contribution. Other domains, such as
behavior and values, are now given equal importance in the cultural identity
model.
While the current culture data shows important distinctions in measurement
and definition, more research is needed which will advance our understanding of
these measures as moderator variables. Specifically, it will be important to clarify
how culture measures compare to one another in different clinical populations.
For example, there is a need to investigate how measures o f acculturative stress,
cultural identity, and acculturation differ or resemble one another when compared
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
to neuropsychological test performance. In general, the selection of moderator
variables as part o f the neuropsychological assessment process still needs to be
clarified.
Hence, the current literature review showed that there is considerable
concern about potential cross-cultural confounds associated with psychological
assessment. One o f these confounds includes language differences; however,
language is a very diverse construct which needs to be assessed and understood at
several levels (e.g., cultural “subjective” and linguistic “objective” levels). The
cultural (subjective) assessment of language is typically addressed in culture
measures, such as those which were reviewed for this study (e.g., Short
Acculturation Scale for Hispanics and Cultural Identity Scales). The linguistic
(objective) assessment of language is emphasized in bilingual research and theory
(e.g., Metalinguistic Abilities and Threshold Theory).
Although language differences are mentioned in the ethics codes and despite
significant empirical and theoretical attempts to understand the cultural and
linguistic aspects of bilingual and monolingual functioning, the issue of language
proficiency (subjective and objective) has not been directly addressed in recent
efforts to provide normative data for the neuropsychological assessment of
Hispanics (e.g., Taussig et al., 1992; Ponton et al., 1996 studies). Discrepancies
have been identified in the research relative to bilingual and monolingual theories,
and culture measurement places considerable emphasis on language use and
exposure. Therefore, it is important to consider different levels of language
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
proficiency (linguistic and cultural) when assessing social, cognitive, and
neuropsychological functioning. Specifically, neuropsychological research is still
deficient in its understanding of how the diverse language background of Hispanic
individuals influences neurocognitive functioning (e.g., nonverbal reasoning,
attention, executive functioning, perceptual and organizational skills).
In general, the issue of test administration and language, across
psychometric specialties, has focused on the caution of testing non-English
proficient clients (e.g., Spanish monolingual). It is evident that clinicians should
avoid testing such clients with translators, informal translation of instruments, or
non-Hispanic normative data. However, there is still a need to further our
understanding o f how language proficiency levels among the Hispanic population
interact with neuropsychological functioning. Most importantly, we need to
consider both culture and language in any assessment of an Hispanic client/patient.
Purpose of the Study
The purpose of the current study was to present an alternative assessment
approach which had not yet been empirically evaluated. With the advancement of
the NeSBHis to the current NeSBHis-Revision (NeSBHis-R) standardization,
additional culture and language issues are addressed. Three culture instruments
were added to the NeSBHis-R in an attempt to further inform the impact of
acculturation on neuropsychological test performance. This additional information
was used to examine the potential confounds and distinctions between the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
constructs of acculturation and cultural identity while controlling for acculturative
stress. In addition, the current study sought to clarify the process of selecting
moderator variables and the impact that they have on neuropsychological test
performance. The current study also served as a validity study for both the CIS
and Marin measures, and further informed the unidimensional and
multidimensional distinction between these culture measures. Finally, a language
proficiency variable was added to the NeSBHis-R in an attempt to operationalize
language categories (e.g., bilingual and monolingual) and to address the need to
compare the performance of monolingual and bilingual subjects on this
neuropsychological battery.
In sum, the purpose of the present study was to examine how culture
(Acculturation, Acculturative Stress, Cultural Identity) and language (unbalanced
and balanced bilingual versus Spanish monolingual) contribute to the cognitive
functioning of Hispanic individuals. The relationship between two cultural
variables (acculturation, cultural identity) was analyzed separately from levels o f
proficiency among distinct language categories (Spanish monolingual, balanced
and unbalanced Spanish-English bilinguals). Finally, culture and language results
were compared to specific domains of neuropsychological functioning (language,
learning and memory, attentional speed o f processing, visuospatial functioning,
and motor functioning); demographic controls were considered in these analyses
(gender, age, and education).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
Research Questions and Hypotheses
Research Question 1 - Demographic Differences on
NeSBHis-R Domain Performance: To What Extent Will
Demographic Differences Among the Current Sample
Influence their Performance on NeSBHis-R Subtests?
Hypothesis: Age, gender, and total years of education will all show
significant correlations with NeSBHis-R subtest scores. Specifically, there will be
a significant difference between men and women on NeSBHis-R subtest scores,
age will be negatively correlated with NeSBHis-R subtest performance (as age
increases, NeSBHis-R test performance will decrease), and total years o f education
will be positively correlated with NeSBHis-R subtest scores (as total years of
education increases so will NeSBHis-R test scores).
Research Question 2 - Influence of Objective Language
Categories on NeSBHis-R Domain Performance: To What
Extent Will Distinct Objective Language Skills Between
Spanish-English Bilinguals and Spanish Monolinguals
Explain NeSBHis-R Subtest Scores?
Hypothesis A: The current sample’s bilingual and monolingual objective
language proficiency, as assessed by the Woodcock-Johnson Language
Proficiency Battery and as compared to six subjective language proficiency
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
ratings, can be operationalized to fit five distinct language categories: (a)
Balanced Bilingual “Strong,” (b) English Dominant Bilingual “Strong,” (c)
Spanish Dominant Bilingual “Strong,” (d) Spanish Monolingual “Strong,” and (e)
Spanish Monolingual “Weak.”
Hypothesis B: Objective language categories will explain a significant
amount o f variance on NeSBHis-R domain/subtest performance.
Hypothesis C: “Strong” Balanced Bilinguals will perform significantly
better on NeSBHis-R subtests than “Strong” English Dominant Bilinguals and
“Strong” Spanish Dominant Bilinguals.
Hypothesis D: Bilingual categories (“Strong” Balanced Bilinguals,
“Strong” English Dominant Bilinguals, and “Strong” Spanish Dominant
Bilinguals) will perform significantly better on NeSBHis-R subtests than “Strong”
Spanish Monolinguals.
Hypothesis E: All “Strong” objective language categories (Bilingual and
Monolingual) will perform significantly better on NeSBHis-R subtests than
“Weak” Spanish Monolinguals.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
Research Question 3 - Influence of Objective Language
Categories on NeSBHis-R Domain Performance While
Controlling for Gender. Age, and Total Years o f Education:
To What Extent Does the Objective Language Category
Variable Still Explain a Significant Amount of the Variance
on NeSBHis-R Subtest Scores Following the Control of
Gender. Age, and Total Years of Education?
Hypothesis: Objective Language Categories will still explain a significant
amount of the variance associated with NeSBHis-R domain/subtest scores after
controlling for gender, age, and total years of education.
Research Question 4 - Influence of Culture Variables on
NeSBHis-R Domain Performance: To What Extent Does
Culture fe.g.. Language Use/Exposure. Behavior. Values’ )
Explain a Significant Amount of the Variance on NeSBHis-
R Subtest Scores?
Hypothesis A: The culture measures identified for the current study, the
Cultural Identity Scales and the Marin, will each yield the same constructs with the
current sample.
Hypothesis B. The CIS and Marin measures will show similar results on
comparative culture variables. For example, it is predicted that both the CIS and
Marin will evidence similar results on language specific variables: CIS Language
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
domain; Marin Language Use/Ethnic Loyalty variable; Marin Media variable).
Similarly, it is predicted that both the CIS and the Marin will yield the same results
on behavior specific variables: CIS Behavior domain and the Marin Ethnic Social
Relations variable.
Hypothesis C: NeSBHis-R subtests will yield the strongest correlations
with language-based culture variables (CIS Language Domain, Marin Language
Use/Ethnic Loyalty, and Marin Media), followed by behavior variables (CIS
behavior domain and Marin Ethnic Social Relations), and values variables (CIS
Values Domain). Values variables, as compared to language and behavior
variables, will show the weakest correlations with NeSBHis-R subtests.
Hypothesis D. The CIS multidimensional model (e.g., separate scores for
English/American variables and Spanish/Latino variables) will be a more sensitive
model in that it will show bicultural influences on NeSBHis-R that the Marin
unidimensional model will not.
Research Question 5: Influence of Culture Variables on
NeSBHis-R Domain Performance While Controlling for
Gender. Age, and Total Years ofEducation: To What
Extent do Culture Variables Still Explain a Significant
Amount of the Variance on NeSBHis-R Subtest Scores
Following the Control o f Gender. Age, and Total Years of
Education?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
Hypothesis: Culture (CIS and Marin) will still be significantly correlated
with NeSBHis-R domain performance after controlling for gender, age, and total
years o f education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 7
CHAPTER 2
METHOD
Participants and Procedures
The subject sample consisted of 83 Hispanic who were recruited through
fliers and oral advertisements in different community centers and churches from
the greater Los Angeles area over a one year period (summer 1997 to summer
1998). A sample o f the flier, which was also used as the verbal script when
delivering oral advertisements, can be found in Appendix A.
Three qualification criteria were identified for inclusion in the study. First,
unremarkable medical and psychiatric histories were required of the subjects.
Specifically, qualified subjects were not under a current medication regimen and
their history was not significant for head trauma, major medical or psychiatric
conditions, or substance abuse. Two subjects were eliminated from the study due
to significant histories of mild head injury. Second, it was required that subjects
could be classified as “Latino/a” or “Hispanic” ethnicity as defined by their
personal ethnic identification, and “family’s place o f origin”—Latin America. It
was also required that subjects were either Spanish monolingual or Spanish-
English bilingual speakers. Third, subjects could not have had previous exposure
to psychometric testing, personally, clinically, or professionally. One subject was
eliminated from the study due to professional training in psychometric assessment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Data was collected by the NeSBHis Research Team which consists of a
group of psychologists, neuropsychologists, and psychology graduate students; all
members identified as bicultural (Latino-American) and bilingual (English-
Spanish). Data collectors were trained for NeSBHis-R administration and scoring
by a licensed neuropsychologist on the team. The training consisted o f three
phases: (a) The identified data collector was required to observe the
administration o f the NeSBHis-R by a licensed neuropsychologist, (b) she/he then
was asked to administer the NeSBHis-R to the same licensed neuropsychologist,
and (c) she/he was observed in administration. The training did not involve
subjects who were part of the current study because there was concern that the
observation component would influence the test performance.
The first phase of the study involved a brief interview with the subject in
which she/he was reminded of the purpose of the study: “to administer
neuropsychological tests to individuals who do not present with major medical,
psychiatric, or substance abuse histories in an effort to further develop
neuropsychological interventions with Latino/Hispanic populations.” Subjects
were also told that their data would remain confidential and that examiners could
not provide raw data nor interpretation of the results to the subject. Referrals were
readily available for subjects who were interested in obtaining professional
consultation and the NeSBHis Research Team was prepared to discuss any cases
for whom a referral would be clinically indicated. No referrals were clinically
required nor requested. The brief interview also consisted o f the examiner
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
determining the birth place of the subject and their choice of English or Spanish
language questionnaires. This data was used to determine if the subject would
receive all the questionnaires in English or Spanish and if the subject would
receive either the “immigrant” or “non-immigrant” version on one o f the four
questionnaires. In addition, the brief interview also required the examiner’s oral
administration of the first page of the demographics questionnaire. This process
was identified as a screening because it addressed questions specific to the
subjects’ medical, psychiatric, and substance use history. If the subject qualified
for the study they were asked to complete the remainder of the questionnaires on
their own (e.g., demographics, culture, and instruments). If the subject was
illiterate or if the subject had questions regarding the completion o f the
questionnaires, the examiner was available for assistance. In some cases, all the
questionnaires were administered orally (e.g., if the subject was illiterate or had
very limited schooling). All questionnaires were reviewed by the examiners to
ensure that all items were completed.
The second phase of the study placed emphasis on the assessment of
language skills. Since the subject had already provided self-assessment of their
subjective English and Spanish language skills when completing the demographic
questionnaire, the examiner was asked to provide a similar subjective rating of the
subject’s English and Spanish language skills. Subjective language ratings, by
both the subject and the examiner, were completed prior to the administration of
objective tests. To ensure the examiner’s subjective assessment of both languages,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
the examiners were required to verbally administer the Woodcock-Johnson
Language Use Survey (WJLUS) to the subject in the opposite language that was
used for the brief interview (WJLUS English Form and WJLUS Spanish Form)
(Woodcock, 1991). For example, if the brief interview was administered in
Spanish, then the WJLUS was administered in English. Thereafter, the examiner
was required to complete a subjective assessment o f the subject’s English and
Spanish language skills, entitled “Language Proficiency—Examiner’s Form”
(Appendix B). Next, selected subtests from the Woodcock-Johnson Language
Proficiency Battery (WJLPB) (Woodcock, 1991) were administered in English
first and then in Spanish. The third phase o f the study required that the examiner
administer the NeSBHis-R to the subject. The order o f questionnaire and test
administration was always the same and was ensured with the completion of a
scoring sheet by the examiner (Appendix C). Following the administration of the
battery, the examiner scored all of the data, which was also back scored by another
NeSBHis Research Team member.
Two additional criteria for inclusion in the current study were identified
following a review of the sample demographics. Since the current study was
identified as a normative data collection, as expected, only one subject endorsed a
clinically significant level of stress on the Hispanic Stress Inventory (HSI). This
subject was eliminated from the study and the remaining sample was defined as a
control for the five levels of stress that are assessed by the HSI: (a)
Occupational/Economic Stress, (b) Parental Stress, (c) Family/Cultural Conflict,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
(d) Marital Stress, and (e) Immigration Stress. Finally, upon review o f the
objective language proficiency data (WJLPB Standard Scores), only five
categories were identified: (1) “Strong” Balanced Bilingual, (2) “Strong” English
Dominant Bilingual, (3) “Strong” Spanish Dominant Bilingual, (4) “Strong”
Spanish Monolingual, and (5) “Weak” Spanish Monolingual. Therefore, one
subject was eliminated due to a classification of English monolingual. The final
sample consisted of 78 subjects.
A description of the final sample (N=78) is based on a review o f the items
endorsed on the demographics questionnaire (Appendix D). When presented with
a choice of language to complete the questionnaires, 44 subjects requested
questionnaires in the Spanish language, 23 requested questionnaires in the English
language, 4 were indifferent but selected Spanish, and 7 were indifferent but
selected English. The demographic data indicates that the sample consisted of 43
men and 35 women, the majority were right handed (n=76), and the age range was
16 to 65 (16-25 = 20 subjects; 26-39 = 33 subjects; 40-65 = 25 subjects). The
level o f education ranged from no formal education to 20 years o f education (0-11
years o f education = 24 subjects; 12-14 years of education = 23 subjects; 15-20
years o f education = 31 subjects). Moreover, 14 subjects received their education
only in the United States, 34 in Latin America, and 30 in both the United States
and Latin America. It was also noted that 25 subjects never earned a degree, 24
subjects earned a High School diploma or G.E.D., 4 subjects earned an Associate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
o f Arts degree, 12 subjects earned a Bachelor o f Arts degree, 11 subjects earned a
Masters degree, and 3 subjects earned a Ph.D., J.D., or M.D.
A consideration o f the subject’s birthplace revealed that 13 subjects were
bom in the United States, and 65 were bom in Latin America. Also, 47 were bom
in the city and 31 were bom in a rural area. O f those subjects who were bom in
Latin America, 35 were from Mexico, 6 from Columbia, 5 from El Salvador, 4
from Honduras, 3 from Cuba, 3 from Venezuela, and less than 3 represented each
o f the following: Puerto Rico, Argentina, Nicaragua, Uruguay, and Bolivia. The
majority o f the subjects were first generation migrants (n=66) and their reason for
migration to the United States included economic (n=33), political (n=6), social
(n=13), and other reasons (n=30, see Appendix E). The remaining sample
consisted of second (n=7), third (n=6), and fourth (n=l) generation subjects.
Regardless o f birthplace, most o f the sample spoke Spanish as their first language
(n=76). Information regarding social economic status (gross yearly income) was
also collected and showed that 33 subjects earned less than $30,000, 22 subjects
earned between $30,000 and 59,999, 8 subjects earned between $60,000 and
$99,999, and one subject earned more than $100,000. The sample, therefore,
largely represents the middle to middle-lower social economic group. Moreover,
17 subjects were full-time students and reported financial aid and scholarship
support as their only source of income. It was also noted that 29 subjects were
identified as married, 5 as separated, 9 as divorced, and 35 as never married.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
Information regarding subjects’ parents was also collected. The majority
of parents were bom in Mexico (38 mothers and 38 fathers), less than 5 parents
were bom in the United States, and the remaining parents were bom in the same
Latin American countries that were identified by the subjects. The level of
education of parents showed that the majority had less than 12 years of education
(fathers: 52; mothers: 54), 12 to 14 had a high school level o f education (fathers:
14; mothers: 12), and a similar number had a college or professional level of
education (fathers: 14; mothers: 14).
Instruments
Six instruments were utilized in the current study: (a) Demographics and
Background Information, (b) Cultural Identity Scales (CIS), (c) Short
Acculturation Scale for Hispanics (Marin), (d) Hispanic Stress Inventory (HSI), (e)
Woodcock-Johnson Language Proficiency Battery - Revised, English and Spanish
Editions (WJLPB-R), and (f) Neuropsychological Screening Battery for
Hispanics - Revised (NeSBHis-R).
Demographics and Background Information
The following demographic information was obtained by the self-report
demographic questionnaire, English and Spanish forms: age, gender, handedness,
reason for migration (personal and/or family), education (total years, degree,
location), social economic status (gross yearly income), country o f origin (rural
town versus city), first language spoken, marital status, parent data (education,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
birthplace), and subject’s generation level. Generation level was operationalized
into four categories: First generation Hispanics (respondent, and all parents and
grandparents bom outside U.S.), second generation Hispanics (respondent bom in
U.S. and all parents and grandparents bom outside U.S.), third generation
Hispanics (respondent and at least one parent bom in the U.S.), and fourth
generation Hispanics (at least one parent and one grandparent bom in the U.S.).
The demographic questionnaire also included a review o f medical history which
assessed for alcoholism, medications, vision and hearing deficits, epilepsy,
tuberculosis, meningitis, encephalitis, diabetes, arthritis, brain tumor, cancer,
embolism, convulsions, glaucoma, vascular problems, automobile accidents, loss
of consciousness, and psychological/psychiatric treatment or hospitalizations.
Woodcock-Johnson Language Proficiency
Battery - Revised
English and Spanish language proficiency was measured with the
Woodcock-Johnson Language Proficiency Battery - Revised (WJLPB-R) and the
Woodcock Language Proficiency Battery - Revised, Spanish version (WJLPB-R,
Spanish). Both the WJLPB-R and the WJLPB-R, Spanish versions consist of three
domains: oral language (lenguaje oral), reading (lectura), and written language
(lenguaje escrito) and are reliable and valid measures. Specifically, for the
WLPB-R, internal consistency reliability coefficients are generally in the high .80s
and low .90s for the subtests and in the mid .90s for the clusters. Moreover, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
WJLPB-R has relatively high criterion validity correlations between the Broad
English Ability cluster and all criterion measures (Stanford-Binet IV, Boehm Basic
Concepts, Bracken Basic Concepts, K-ABC, McCarthy, and PPVT-R).
The Spanish form of the Woodcock Language Proficiency Battery -
Revised (WLPB-R, Spanish) measures the same abilities as the English form (oral,
reading, and written) and contains parallel tests and clusters (Woodcock & Munoz-
Sandoval, 1995a & 1995 b). The translation process involved approximately 30
examiners in five Spanish-speaking countries (Costa Rica, Mexico, Peru, Puerto
Rico, and Spain) as well as a board of consulting editors. Extensive re-norming,
additions, and revisions were added to the Spanish WJLPB-R from the original
Spanish WJLPB. The WJLPB-R, Spanish cluster reliability coefficients are
generally in the .90s and since the WJLPB-R English and Spanish batteries are
parallel in content and structure, the WJLPB-R validity results were generalized to
the Spanish WJLPB-R. Moreover, a validity study that is unique to the Spanish
WJLPB-R found a correlation between the WLPB-R Oral Language (Lenguaje
oral) cluster with the Language Rating Scale (PreLAS), and the Woodcock-Munoz
Language Survey English and Spanish Forms (W-M Survey) (Woodcock &
Munoz-Sandoval, 1995a & 1995b).
Cultural Identity Scales
Cultural identity was measured with the Cultural Identity Scales (CIS).
The CIS assesses the individual's sense of familiarity with both American and
permission of the copyright owner. Further reproduction prohibited without permission.
36
Latino(a) cultures and was made available in both English and Spanish formats. It
compares four cultural identity groups (Highly Bicultural, Latino/a, American, and
Low-Level Bicultural) on language, behavior, and attitude/value scales. The CIS
is a valid and reliable instrument. When correlated with the Marin (Marin et al.,
1987) Short Acculturation Scale, generation status, and length o f time in the
United States, criterion validity was significant for all the CIS language scales (p
< .001), three of the four behavior/familiarity scales, and two o f the three
attitude/value scales. Evidence o f convergent and divergent validity was also
found. Familiarity With American Culture was positively correlated with
California residency (r = .21) and with English proficiency (r = .31), and Latino(a)
Activism was negatively correlated with California residency (r= -.15).
Moreover, Perceived Discrimination was negatively correlated with California
residency (r= -.15) and Respeto (respect) was negatively correlated with English
proficiency (r= -.17). Cronbach alpha reliability coefficients were obtained for
each of the ten scales and ranged from .69 to .91. The largest coefficients were
found for the language scales (.87 - .91), followed by comparable values/attitudes
scale coefficients (.72 - .81), and behavior/familiarity coefficients (.69 - .89).
Short Acculturation Scale for Hispanics
Acculturation was measured with the Short Acculturation Scale for
Hispanics (Marin) (Marin et al., 1987) and was made available in both English and
Spanish formats. The Marin is a reliable and valid instrument. With respect to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
reliability, the alpha coefficient for the 12 common items was .92. Among these
12 Marin items, 3 factors were identified: Language (.90), Media (.86) and Ethnic
Social Relations (.78). Validity coefficients were also significant. The
correlations between the subject's total Marin score and their generation level,
length o f residence, and own Likert scale evaluation (e.g., very Latino(a) to very
American) were .65, .70, and .76 respectively (p< .001). Moreover, the correlation
between the Marin Acculturation Index (generation plus length of residence in the
U.S. plus subject's own evaluation), and the 12 scale items was .83 (p< .001).
Finally, a T-Test showed that the non-Hispanic scores (M=4.63) differed
significantly from the Hispanic scores (M=2.72).
Hispanic Stress Inventory
Acculturative stress was measured with the Hispanic Stress Inventory
(HSI) (Cervantes, Padilla, & Salgado de Snyder, 1991) and was made available in
both English and Spanish formats. The HSI is a measure of psychosocial stress for
Hispanics which taps a range of psychosocial stressor events for both recent
Hispanic immigrants and U.S. bom Hispanics. The HSI is a reliable and valid
instrument that measures "constructs related more to psychosocial stress that is
culturally relevant than to symptoms of anxiety, depression, somatization, or
generalized distress" (Cervantes, Padilla, & Salgado de Snyder, 1991, p.447).
Internal consistency reliability coefficients were in the moderate to high range (.77
to .91) and Test-Retest Reliability on the Immigrant version ranged from .61 to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
.86. Content validity was based on the ratings of five judges. There was complete
agreement (five out o f five judges) on the assignment of a total o f 79 items (45%
o f the items). It is important to note that
Only those items that were judged to be conceptually clean and relevant to
the population were retained. If an item was not categorized equivalently
by at least four of the five judges, it was discarded unless the item was seen
as very clinically important despite the conceptual overlap. Two additional
items were retained for this reason" ( Cervantes, Padilla, & Salgado de
Snyder, 1991, p. 447).
Construct validity (eigen values o f greater than 1.00) demonstrated five factors for
the immigrant subsample and four factors for the U.S.-bom subsample. The
average factor loading for the immigrant sample was .550 and for the U.S.-born
sample it was .555.
Neuropsychological Screening Battery
for Hispanics - Revised
Neuropsychological functioning was measured with the
Neuropsychological Screening Battery for Hispanics, Revised (NeSBHis-R),
which is based on the original NeSBHis (Ponton et al., 1996). The NeSBHis
battery includes 10 instruments which are administered in Spanish and is designed
to assess the domains o f language, memory, mental control (attention/
concentration), psychomotor speed, visuo-spatial functioning, and nonverbal
reasoning. Normative data of the NeSBHis is based on a sample of 300 Spanish
monolingual and Spanish-English bilingual subjects. Factor analytic methodology
was used to determine the construct validity of the NeSBHis, and the results
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
demonstrated a stable factor structure. The following five factors yielded
significant eigen values (> 1.00): language or verbal factor, verbal
learning/memory factor, attentional/speed of processing factor, a visuo-spatial
factor; and an independent or pure motor factor. The ten subtests which are
associated with each of the identified neuropsychological factors/domains and
which make up the NeSBHis battery are presented. This review is taken from
Ponton et al., (1996).
The Language Domain is assessed with two subtests: The Ponton-Satz
Boston Naming Test (PS-BNT) and a Controlled Oral W ord Association test FAS
(Verbal Fluency T est). The PS-BNT measures confrontational naming in which
the subject is required to verbally identify a series of 30 pictures. The FAS is a
measure of verbal fluency which requires the subject to generate as many words as
possible following a letter prompt (e.g., F - A - S); the subject is given a 60 second
time limit per letter.
The Memory Domain assesses for both verbal and nonverbal memory. The
former was assessed with the WHO-UCLA Auditory Verbal Learning Test
(WHO-UCLA AVLT). This test consists o f 15 words which are read to the
subject five times. The subject is asked to recall as many words as possible
following each verbal presentation o f the list by the examiner. Then an
interference trial is presented (e.g., new list of words) which the subject is also
asked to recall. Immediately after the subject’s recall o f the interference list,
she/he is asked to recall the original list of words (e.g., the list repeated five times);
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
this is the “short-term recall.” Twenty minutes later, the subject is asked to recall
the same list o f words (e.g., the list repeated five times); this is the “long-term
recall.” Finally, the subject is read a list of 30 words in which she/he is required to
identify the words that were part o f the original list of 15 words; this is the
“recognition trial.” Nonverbal memory is assessed with the Rey-Osterreith
Complex Figure test (Rey-O Delay). The subject is asked to draw (e.g., with blank
paper and pencil), from memory, a complex figure that was presented 10 minutes
earlier.
The Visual-Spatial Domain is also assessed with two tests. The first is the
Block Design subtest from the Escala de Inteligencia Wechsler para Adultos
(EIWA - Spanish form o f the Wechsler Adult Intelligence Scale - Revised
“WAIS-R”). This test measures visual-spatial (perceptual organizational) and
constructional skills. The EIWA Block Design differs from the WAIS-R Block
Design in that there is a different scoring system: (a) there are fewer bonus points
for faster performance, and (b) 3 o f the 10 items are different from the WAIS-R.
The second Visual-Spatial Domain subtest is the copying o f the Rey-Osterreith
Complex Figure (Rey-O Copy) onto blank paper (e.g., paper-pencil task). The
Rey-O Copy and assesses visual-spatial, planning and organization, and fine motor
skills. The Taylor 36 point scoring system was used for the Rey-O (Copy).
The Mental Control Domain includes three tests. The first two are the
EIWA Digit Span (Forward, Backward, Total) and Digit Symbol subtests. The
administration o f the Digit Span differs from the EIWA administration in that both
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
sets of digits are administered per trial. The Digit Span test measures attention,
working memory, and cognitive manipulation skills. The Digit Symbol subtest
measures sustained attention, speed of information processing, and short-term
memory. The third Mental Control Domain subtest is the Color Trails (1 and 2)
which measures sustained attention, sequencing ability, speed of processing, and
divided attention. The Color Trails test was developed with the intention of
minimizing cultural bias by not including letters o f the alphabet (e.g., minimize
penalty for limited education). The Color Trails 1 subtest requires the subject to
connect circles by number order (e.g., 1 to 25). The Color Trails 2 subtest also
asks the subject to connect circles in number order; however, an alternating color
sequence is added to the task.
The Motor (psychomotor) Domain includes the Pin Test which is a
measure of fine motor dexterity. It requires the subject to punch as many holes as
possible using a pin and metal plate. The pin is inserted through small holes in a
metal plate and the number of hole punches (paper between metal plates) is
compared to a 30 second time limit. Both dominant hand and non-dominant hand
functioning is assessed . The final domain, Reasoning, is comprised of the
Raven’s Standard Progressive Matrices (Ravens). The Ravens is considered to be
a test o f nonverbal reasoning and consists of a booklet o f 60 nonverbal
diagrammatic puzzles. Each page of the booklet contains a puzzle in which there
is a part missing, and the subject is required to select one of six items which best
completes the puzzle.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
In the current study, neuropsychological assessment was measured with a
revised version of the NeSBHis—the NeSBHis-R. The NeSBHis-R differs from
the original NeSBHis battery in that it adds three neuropsychological tests:
Stroop; Grooved Pegboard; Verbal Fluency, Animals and Fruits. The Stroop
measures both reading and naming fluency as well as an individual’s ability to
shift hishe/her “perceptual set to conform to changing demands” (Lezak, 1983, p.
523). There are three parts to the Stroop, and all scores are based on completion
time. First, the subject is asked to name colors that present on a visual stimuli, row
by row (red, green, blue). Then the subject is asked to name colors, in Spanish,
that present on a visual stimuli. Then the subject reads words from a visual
stimuli, row by row (red, green, blue). Finally, the subject is asked to name the
color of the ink and ignore the word, row by row (e.g., words printed in colors that
do not correspond to the word: red written in green ink). While the first two parts
of the Stroop assess primarily for attention, concentration, and speed of processing
(e.g., rapid naming), the third part also assesses for cognitive flexibility or
inhibitory abilities. Hence the Spanish form of the Stroop was added to the Mental
Control NeSBHis-R Domain, which is now referenced as the Executive
Function/Attention-Concentration Domain (EF-AC).
On the Motor Domain, the Pin Test was replaced by the Grooved Pegboard
test. The Grooved Pegboard test “consists of a small board containing a 5 x 5 set
of slotted holes angled in different directions. Each peg has a ridge along one side
requiring it to be rotated into position for correct insertion” (Lezak, 1983, p. 532).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Time completion, in seconds, is scored for both the dominant and non-dominant
hand. Hence, the Grooved Pegboard serves as a measure of psychomotor speed
and precision so it was added to the NeSBHis-R Motor Domain. Finally, on the
Language Domain, the semantic categories of Animals and Fruits were added to
the Verbal Fluency Test. As with the FAS version, the subject is asked to name as
many animals and fruits as possible and a one minute time limit is allowed for
each category.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 4
CHAPTER 3
RESULTS
In this chapter, the hypotheses are restated and results which tested
relationships between NeSBHis-R subtests, the Objective Language Category
variable, and culture variables are presented. The first analyses compared each
NeSBHis-R domain by three demographic variables: age, gender, and total years
o f education. Then the Objective Language Category variable was introduced and
separate analyses were run in an attempt to verify the operational definitions o f
each language category. This was done by comparing six subjective language
variables to the Objective Language Category variable. Then, the Objective
Language Category variable was compared to each NeSBHis-R domain/subtest.
Thereafter, additional comparisons were made between NeSBHis-R subtest scores
and the Objective Language Category variable, this time controlling for age,
gender, and total years of education. A similar pattern was followed in comparing
NeSBHis-R domains/subtests to the culture variables. First, reliability analyses
were run for each culture measure. Then each culture variable and cluster was
compared to performance on NeSBHis-R subtests. These results were followed by
a series of analyses which compared NeSBHis-R performance and culture
variables while controlling for age, gender, and total years o f education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
Ail results are presented by the following order of NeSBHis-R Domains:
(a) Reasoning, (b) Language, (c) Executive Function/Attention-Concentration (EF-
AC), (d) Memory, (e) Visual-Spatial, and (f) Motor. It is also important to note
that specific EF-AC (Stroop and Color Trails 1 and 2) and Motor (Grooved
Pegboard, Dominant and Non-Dominant) subtest scores represent the total time
required to complete the task so lower test scores on these subtests are associated
with better performance.
Normative Data
Normative data is frequently compiled to assist with the interpretation of
neuropsychological test performance. A normative sample (e.g., subjects deny
major medical, psychiatric, and head trauma histories), completes one or more
neuropsychological tests and then the results are stratified by age. Other
stratification variables also frequently include gender and level of education.
Stratification variables are important to include because they often explain a
significant amount of the variance on neuropsychological test performance. The
current study compiled two normative data charts: one for women (Appendix F)
and one for men (Appendix G). Each chart is stratified by age (16-25; 26-39; 40-
65) and years of education (0-11; 12-14; 15-20). While it was important to
distinguish between the identified age and education categories, the current sample
yielded a small number of subjects per cell; therefore, it was decided that age and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
total years o f education would be treated as continuous variables for the purpose of
answering all proposed Research Questions and hypotheses.
Analyses o f Hypotheses
Research Question 1 - Demographic Differences on
NeSBHis-R Domain Performance: To What Extent Will
Demographic Differences Among the Current Sample
Influence their Performance on NeSBHis-R Subtests?
Given that normative data is most frequently stratified by age, gender, and
total years o f education, and since these three variables were identified in the
original NeSBHis study, they were a primary focus o f the current study.
Hypothesis. The first hypothesis stated that age, gender, and total years of
education would all be correlated with neuropsychological subtest scores.
Specifically, it was predicted that there would be a significant difference between
men and women on neuropsychological test performance, that age would be
negatively correlated with neuropsychological test performance (as age increases,
neuropsychological test performance will decrease), and that education would be
positively correlated with neuropsychological test performance (as total years of
education increases so will neuropsychological test performance).
On the NeSBHis-R Reasoning Domain (Table 1), the Ravens subtest total
score was significantly correlated with age, gender, and total years o f education
and confirms the stated hypothesis. The Ravens subtest has been associated with
with permission of the copyright owner. Further reproduction prohibited without permission.
47
theories of intelligence (e.g., Spearman’s g) (Raven, Raven, & Court, 1993) and is
identified in the original NeSBHis study as a measure o f nonverbal intelligence
(Ponton et al., 1996); however, it is important to note that a strong “education
effect” was identified with the current sample which showed that as the total years
of education increased, so did Ravens total scores. Also, as expected, it was noted
that as age increased, Ravens total scores decreased. Gender differences were also
identified and favored men on the Ravens total scores.
Table 1
Years of Education
Total Years of
Gender Age Education
Male(N=43) Female(N=35)
M SD M SD
Ravens
(Total) 44.65* 10.65 39.40* 12.26 -.27 * .64 **
t (76) = 2.02, *p < .05; **p<.01.
r (78), *p < .05; ♦*£<.01.
On the NeSBHis-R Language Domain, most subtests were significantly
correlated with total years of education, but not with age or gender (Table 2). The
hypothesis was confirmed in that education was significantly correlated with three
of the four Language Domain subtests. The “Animals” category from the Verbal
Fluency subtest was not significantly correlated with education and, therefore,
suggests that verbal access to semantic categories is less associated with academic
exposure. The stated hypothesis was not confirmed with respect to age and gender
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
on Language Domain subtests. Therefore, the current results suggest that
education, although a weak correlation, demonstrates a significantly stronger
correlation with NeSBHis-R Language Domain performance than age and gender.
Table 2
Years of Education
Gender Age
Total Years of
Education
Male(N=43)
M SD
Female(N=35)
M SD
Ponton-Satz Boston
Naming Test
(Total) 23.37 5.11 22.31 4.07 .22 .26 *
Verbal Fluency
(FAS Total) 35.00 9.79 34.23 10.04 .13 .27 *
Verbal Fluency
(Animals) 18.49 4.38 17.23 4.86 .07 .16
Verbal Fluency
(Fruits) 13.51 3.73 13.17 3.05 .17 .28 *
t (76), *g < .05; **p<.01.
r (78), *p< .05; **p< .01.
The results of the NeSBHis-R Executive Function/Attention-Concentration
(EF-AC) Domain subtests varied when compared to education, age, and gender
variables (Table 3). The education variable was consistent with the proposed
hypothesis as it was positively correlated with all EF-AC subtests. The strongest
education and EF-AC correlations were on the Stroop Word and Digit Symbol
subtests. The former subtest requires rapid reading skills. The latter subtest
requires similar academic skills because the subject is asked to learn and copy
symbols (e.g., this paper-pencil task is similar to the concept of writing). Age was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 9
significantly correlated with EF-AC subtests which emphasized visual-
motor/visual-perceptual and speed of processing skills. For example, weak to
moderate correlations were noted on the Digit Symbol, Color Trails 1, and Color
Trails 2 subtests. The strongest age correlation was identified on the Color Trails
2 subtest which requires higher order visual-spatial, fine motor, and mental control
skills. Moreover, age correlations were not significant on EF-AC subtests which
require verbal output. Therefore, the proposed hypothesis, relative to age effects,
was confirmed in that psychomotor speed and higher order cognitive flexibility
was negatively correlated with performance on EF-AC subtests. However, the
proposed hypothesis relative to age effects was also disconfirmed because verbal
performance on EF-AC subtests was not affected with the current sample. Gender
results did not support the proposed hypothesis because findings were not
significantly correlated with any of the EF-AC subtests, with the exception of the
Digits Forward subtest. On this test, men recalled one additional digit forward.
The results o f the NeSBHis-R Memory Domain demonstrated
discrepancies between oral and fine-motor skills when compared to education, age,
and gender variables (Table 4). On the education variable, the proposed
hypothesis was disconfirmed when compared to Memory Domain subtests that
required oral output as education was not significantly correlated on any of the
AVLT tasks. However, the proposed hypothesis was confirmed on the memory
subtest which required fine-motor output. For instance, on the Rey-Osterreith
(Delay), a weak correlation was noted when compared to total years of education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
A similar oral versus fine motor discrepancy was noted on the age variable. The
gender variable was not consistent with the proposed hypothesis in that none o f the
Memory Domain subtests showed significant differences between men and
women.
Table 3
Comparison o f NeSBHis-R Executive Function./Attention-Concentration Domain
Subtests bv Gender. Age, and Total Years of Education
Gender Age
Total Years
of Education
MaIe(N=43)
M SD
FemaIe(N=35)
M SD
Stroop Color 67.65 13.98 70.40 10.26 .10 -.27 *
Stroop Word 44.93 8.76 44.11 6.84 -.18 -.40 **
Stroop
Interference 116.19 26.25 114.49 24.34 .02 -.34 **
Digit Span
Total 10.00 2.25 9.51 1.22 -.19 .32 **
Digit Span
Forward 5.56* 1.14 5.09* .74 -.10 .28 **
Digit Span
Backward 4.47 1.33 4.46 .78 -.21 .25 *
Digit Symbol 58.84 13.92 59.86 14.78 -.34 ** .61 **
Color Trails 1 39.42 14.21 41.54 15.67 .40 ** -.35 **
Color Trails 2 86.51 28.47 91.77 31.81 .51 ** -.51 **
t (76), *£ < .05; **p<.01.
r (78), *p<.05; **£<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Table 4
Years o f Education
Gender Age
Total Years
of Education
MaIe(N=43)
M SD
Female(N=35)
M SD
AVLT
(Total: Lists 1-5) 54.21 7.18 52.80 5.68 -.01 .16
AVLT
(Short-Term
Recall) 12.16 2.21 11.97 1.85 -.00 .03
AVLT
(Long-Term Recall) 12.16 2.63 12.26 1.67 .07 .06
AVLT
(Recognition) 14.49 .96 14.34 1.00 -.21 .08
Rey-Osterreith
(Delay) 21.53 6.49 19.30 6.00 -.41 ** .28 *
t (76), *£<.05; **£<.01.
r (78), *£< .05; **£<.01.
Discrepancies were noted on the NeSBHis-R Visual-Spatial Domain when
compared to education, age, and gender variables (Table 5). The proposed
hypothesis was confirmed with reference to the education variable as both the
Block Design and Rey-Osterreith (Copy) subtests demonstrated moderate
correlations. This statistical result suggests that with the current sample, visual-
spatial skills improved as total years of education increased. On the age variable,
discrepancies were noted between visual-spatial subtests but there was still support
the proposed hypothesis. For example, the correlation between age and
performance on the Block Design subtest was moderate, and when the Rey-
Osterreith (Copy) scores were compared to the age variable, the correlation
approached significance, r (78) = .21, £ = .08. Gender results were not consistent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
with the proposed hypothesis when compared to performance on the Block Design
subtest, and although the hypothesis was confirmed relative to Rey-Osterreith
(Copy) scores, it is important to note that the significance was weak because men
favored women by only one point on this subtest.
Table 5
Years of Education
Gender Age
Total Years
of Education
Male(N=43) Female(N=35)
M SD M SD
Block Design
(Total) 36.93 6.78 34.31 6.37 -.44 ** .45 **
Rey-Osterreith
(Copy) 34.22* 1.99 33.13* 2.24 -.21 .51 **
J(76), *£ < .05; **£<.01.
I (78), *p<.05; **£<.01.
The NeSBHis-R Motor Domain reflected discrepancies between gender,
age, and education variables (Table 6). Results associated with the education
variable were consistent with the proposed hypothesis as moderate correlations
were reflected on both dominant and non-dominant hand performance on the
Grooved Pegboard subtest. This finding suggests that as total years of education
increases, so does speed of fine-motor skills. The proposed hypothesis was further
supported by the comparison of age to Grooved Pegboard scores as non-dominant-
hand performance was statistically significant and dominant-hand performance
approached significance, r (78) = . 19, £ = .08. Although the age correlations were
weak as compared to Grooved Pegboard performance, it was notable that age
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
explained greater variance than gender differences on the Grooved Pegboard
subtest. Gender did not confirm the proposed hypothesis as Grooved Pegboard
performance differences between men and women were not statistically
significant.
Table 6
Education
Gender Age
Total Years
of Education
Male(N=43) Female(N=35)
M SD M SD
Grooved Pegboard
(Dominant Hand) 69.53 13.09 71.60 14.46 .19 -.46 **
Grooved Pegboard
(Non-Dominant
Hand) 76.47 16.25 80.20 19.06 .32 ** -.37 **
t (76), *g < .05; **£< .01.
1(78), *£<.05; **£< .01.
A summary o f the findings associated with research question one, shows
that total years o f education was positively correlated with all NeSBHis-R
Domains, but showed discrepancies on the Memory Domain. For instance, the
AVLT subtests were not significantly correlated with education but the Rey-
Osterreith (Delay) subtest was. Correlations between NeSBHis-R Domains and
the age variable showed similar findings. All Domain comparisons were
significant except the AVLT subtests. Also, an age finding which differed from
the education findings was that, on the EF-AC Domain, nonverbal output subtests
were significant while verbal subtests were not. The gender variable was not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
significantly correlated with NeSBHis-R Domains. Overall, as compared to the
current sample’s performance on NeSBHis-R Domains, education demonstrated
the strongest correlation with higher NeSBHis-R subtest scores, less significance
was noted with the age variable, and the gender variable was insignificant. The
gender variable did not show significance as compared to NeSBHis-R Domain
performance.
Research Question 2 - Influence o f Objective Language
Categories on NeSBHis-R Domain Performance: To What
Extent Will Distinct Objective Language Skills Between
Spanish-English Bilinguals and Spanish Monolinguals
Explain NeSBHis-R Subtest Scores?
This question concerns the extent to which objective language skills
(Spanish-English bilingual and Spanish monolingual) explain performance on
neuropsychological subtests. The original NeSBHis study did not address
bilingual versus monolingual language differences; therefore, the following results
sought to operationalize bilingual and monolingual skills and then compared the
Threshold Theory of language to neuropsychological test performance.
Hypothesis A. It was hypothesized that bilingual and monolingual
language proficiency could be operationalized to fit the five Objective Language
Categories that were identified upon review of the current sample’s performance
on the Woodcock-Johnson Language Proficiency Battery-Revised (e.g., Standard
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
scores): (a) Balanced Bilingual “Strong,” (b) English Dominant Bilingual
“Strong,” (c) Spanish Dominant Bilingual “Strong,” (d) Spanish Monolingual
“Strong,” and (e) Spanish Monolingual “Weak.”
The Balanced Bilingual category required that the subject’s standard scores
fall within the same standard deviation on both Spanish and English WJLPB-R
subtests (e.g., both average; both above average). The unbalanced bilingual
categories (e.g., English Dominant Bilingual and Spanish Dominant Bilingual)
required that the subject’s Spanish and English WJLPB-R standard scores reflect a
difference of at least one or more standard deviation(s) difference between
languages (e.g., English = above average and Spanish = average). The Spanish
Monolingual category required that the subject obtain a standard score less than 79
on the English WJLPB-R subtests. Distinctions between the level of proficiency
for each category were also based on the WJLPB-R subtest standard scores. The
“Strong” level reflected standard scores above 90 and the “Weak” level reflected
standard scores below 79. For example, the “Balanced Bilingual Strong” category
required WJLPB-R standard scores above 90 on both Spanish and English
subtests. The “English Dominant Bilingual” category required standard scores
above 90 on English WJLPB-R subtests and standard scores between 80 and 89 on
Spanish subtests. The “Spanish Dominant Bilingual” category required standard
scores above 90 on Spanish WJLPB-R subtests and standard scores between 80
and 89 on English subtests. The “Spanish Monolingual Strong” category required
standard scores above 90 on Spanish WJLPB-R subtests and standard scores below
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
79 on English subtests. The “Spanish Monolingual Weak” category required
standard scores below 89 on Spanish WJLPB-R subtests and standard scores
below 79 on English subtests. Similar below average to impaired performance on
WLPB-R subtests was considered to identify the “Weak” counterparts of the
Balanced Bilingual Strong, English Dominant Bilingual Strong and Spanish
Dominant Bilingual Strong groups; however, the current sample did not meet this
criteria. Therefore, only the five Objective Language categories that represent the
current sample were statistically analyzed.
Six subjective language ratings were identified to verify the five Objective
Language Categories: (a) Subject’s self-report o f Spanish language proficiency as
reflected by a 5-point Likert scale; (b) Subject’s self-report of English language
proficiency as reflected by a 5-point Likert scale; (c) Examiner’s rating of the
subject’s Spanish language proficiency as reflected by a 5-point Likert scale; (d)
Examiner’s rating o f the subject’s English language proficiency as reflected by a
5-point Likert scale; (e) Subjects’ self-report of Spanish language proficiency as
reflected by items endorsed on the CIS Spanish Proficiency variable (Scale ranged
from 3 “low” to 12 “high”); and (f) Subject’s rating o f English language
proficiency as reflected by items endorsed on the CIS English Proficiency variable
(Scale ranged from 3 “low” to 12 “high”). Also note that for each subject, all
subjective language measures were completed prior to the administration of the
objective tests (e.g., WJLPB-R and NeSBHis-R).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
As demonstrated by Table 7, an Analysis of Variance (ANOVA) confirmed
five objective language categories (e.g., Balanced and Unbalanced Bilingual,
Monolingual, and Strong and Weak) when compared to subjective language
ratings, and post hoc comparisons noted significant differences between the five
categories. Three general findings are highlighted following a review o f the post
hoc results: (a) Differences between the five categories (b) Differences between
the subject and examiner ratings, (c) Differences between the 5-point Likert scale
on the demographics questionnaire and the CIS Language Proficiency variables.
First, the post hoc results further confirmed the five objective language
categories. Support for each language category is presented below. On the English
Dominant Bilingual (Strong) Objective Language Category, English language
subjective ratings were significantly higher than English language subjective
ratings o f Spanish Dominant Bilinguals (Strong) and Spanish Monolinguals
(Strong and Weak). Also, Spanish language ratings for English Dominant
Bilinguals (Strong) were not significantly stronger than any other Objective
Language Category. These findings support the Objective Language Category of
English Dominant Bilinguals (Strong).
Upon review of the Spanish Dominant Bilingual Strong Objective
Language Category, it was expected that the subjective Spanish language ratings
would be stronger than the English Dominant Bilinguals (Strong) and Spanish
Monolinguals (Weak); these expectations were confirmed by the post hoc test. It
was also expected that subjective English language ratings by Spanish Dominant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Bilinguals (Strong) would be stronger than Spanish Monolinguals (Strong and
Weak); this expectation was also confirmed by the post hoc tests.
Upon review o f the Spanish Monolingual (Strong and Weak) ratings, post
hoc tests also favored these categories. Indeed subjective ratings showed that
Spanish Monolinguals (Weak) did not rate themselves as stronger than any other
Objective Language Category, and the examiners demonstrated a similar rating
with the additional note o f Spanish proficiency o f Spanish Monolinguals (Weak)
as stronger than that of the English Dominant Bilinguals (Strong).
Moreover, the Spanish Monolingual Strong category was supported by the
higher subjective Spanish language proficiency ratings endorsed by these subjects
as compared to the subjective Spanish language proficiency ratings o f Spanish
Monolinguals whose level of proficiency was “Weak” and English Dominant
Bilinguals (Strong). Also, Spanish Monolingual (Strong) ratings o f English
language proficiency were only stronger than that o f Spanish Monolinguals
(Weak).
The Balanced Bilingual Strong category was also confirmed. For instance,
English language ratings of Balanced Bilinguals (Strong) were stronger than all
other categories except the English Dominant Bilinguals (Strong), and Spanish
language ratings of Balanced Bilinguals (Strong) were only stronger than English
Dominant Bilinguals (Strong) and Spanish Monolinguals (Weak).
In addition to reviewing the Objective Language Categories, ratings by
examiners were also compared to ratings by subjects. It was noted that English
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
language ratings were consistent between subjects and examiners. However, upon
review o f Spanish language ratings between subjects and examiners, similarities
and differences were noted. For example, while both subjects and examiners rated
Spanish proficiency differences between the four “Strong” categories, only
subjects rated differences between the “Strong” and “Weak” categories. In
general, subjects were more sensitive in their ability to distinguish levels of
Spanish language proficiency.
A final note is made with reference to Table 7. It was also apparent that
the 5-point Likert scale on the demographics questionnaire was less sensitive to
language proficiency differences than the CIS language proficiency variables. For
example, while both measures showed significant differences between English
proficiency ratings, the 5-point Likert scale only identified one difference between
Objective Language Categories. The CIS Spanish Proficiency variable identified
all salient differences that were required to confirm the Objective Language
Categories. Hence, the last two issues, differences between subject and examiner
ratings, and differences between measures, suggest that one needs to be very
sensitive to the issue of subjective language assessment.
Hypothesis B. It was hypothesized that the Objective Language Category
variable would explain a significant amount of variance on neuropsychological
domain/subtest performance.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Table 7
Analysis o f Variance: Comparison of Objective Language Categories and
Subjective Language Ratings
Balanced
Bilingual
Strong
English
Dominant
Bilingual
Strong
Spanish
Dominant
Bilingual
Strong
Spanish
Mono
lingual
Strong
Spanish
Mono
lingual
Weak
(a) (b) (c) (d) (e)
M
SD
M
SD
M
SD
M
SD
M
SD F
Spanish
Language
Rating by
Subject
4.67
.59
4.00
.82
4.89
.47
(c> b)
4.72
.57
4.27
.88
3.74**
English
Language
Rating by
Subject
4.56
.62
(a > c,d,e)
4.90
.32
(b > c,d,e)
3.61
.61
(c > d,e)
2.94
.87
(d > e)
2.00
.76
40.15**
Spanish
Language
Rating by
Examiner
4.44
.92
(a > b)
3.50
.85
4.72
.67
( O b )
5.00
.00
(d > b)
4.87
.35
(O b )
10.01**
English
Language
Rating by
Examiner
4.72
.57
(a > c,d,e)
5.00
.00
(b > c,d,e)
3.67
.77
( O e )
3.00
.84
(d > e)
2.13
.99
34.41**
CIS
Spanish
Proficiency
11.00
1.28
(a > b.e)
8.60
1.78
11.39
1.09
(c > b.e)
10.83
1.50
(d > b.e)
8.27
1.03
16.96**
CIS
English
Proficiency
10.89
1.37
(a > c.d,e)
11.50
1.08
(b > c.d,e)
9.11
1.02
(c > d,e)
7.72
1.81
(d > e)
4.93
1.71
44.10**
F (73), * £ < .05; ** £ < .0 1 .
Note: ( > ) indicates significant differences between objective language categories
(Tukey * £ < .05)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
Hypothesis C. Because the current sample did not represent both “Strong”
and “Weak” bilinguals, it was not possible to directly test the Threshold Theory.
The theory states that levels of linguistic proficiency among bilingual children are
likely correlated with cognitive benefits and deficits (Ricciardelli, 1992). The
intention o f the current study was to test the threshold theory with an adult sample
in which “levels” o f linguistic proficiency were defined as “Strong” and “Weak.”
The goal was to test strong and weak levels across all five Objective Language
Categories. However, since all bilinguals in this sample were identified as
“Strong” bilinguals, Hypothesis C focused on testing the following category
differences: “Strong Balanced Bilinguals,” “Strong English Dominant
Bilinguals,” “Strong Spanish Dominant Bilinguals,” “Strong Spanish
Monolinguals,” and “Weak Spanish Monolinguals.” The current sample did not
allow for the comparison o f “Strong” and “Weak” across all Objective Language
Categories. However, the general concept o f the theory, bilingual proficiency is
negatively and positively correlated with cognitive performance, is applied to the
current adult sample. Thus, “Strong” levels o f bilingualism are the focus of
Hypothesis C. It was hypothesized that “Strong” Balanced Bilinguals would
perform significantly better on neuropsychological subtests than “Strong” English
Dominant Bilinguals and “Strong” Spanish Dominant Bilinguals.
Hypothesis D. Differences between bilingual (Strong) and monolingual
(Strong) categories were considered. It was hypothesized that all bilingual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
categories (Strong Balanced and Strong Unbalanced) would perform significantly
better on neuropsychological subtests than Spanish Monolinguals (Strong).
Hypothesis E. Differences between “Strong” bilingual and monolingual
categories and the “Weak” monolingual category were considered. It was
hypothesized that all “Strong” categories (bilingual and monolingual) would
perform significantly better on neuropsychological subtests than “Weak” Spanish
Monolinguals.
Results associated with Hypotheses B through E are now discussed by each
NeSBHis-R domain. Following an Analysis of Variance between the Objective
Language Category variable and the NeSBHis-R Reasoning Domain (Table 8),
Hypothesis B was confirmed. In other words, objective language categories
significantly explained some of the variance on Ravens test performance. Post hoc
tests further confirmed Hypothesis E, as both bilingual and monolingual “Strong”
categories, performed significantly better on the Ravens subtest than Spanish
Monolinguals for whom the level of proficiency was “Weak.” Post hoc tests of
Hypotheses C and D, however, were not confirmed. Significant differences were
not noted between “Strong” balanced and unbalanced bilinguals nor between
bilingual and monolingual “Strong” categories when compared to performance on
the Ravens subtest. These findings suggest that while the Ravens subtest is a
nonverbal test, strong language skills will likely benefit an individual’s nonverbal
reasoning skills.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
Table 8
A n a ly sis o f V a ria n c e : M ean C o m p a riso n o f N e S B H is-R R e a so n in g D o m a in
S u b te st b v F iv e O b je c tiv e L a n g u a g e C a te g o rie s
English Spanish Spanish Spanish
Balanced Dominant Dominant Mono Mono
Bilingual Bilingual Bilingual lingual lingual
“Strong” “Strong” “Strong” “Strong” “Weak”
(a) (b) (c) (d) (e)
M M M M M
SD SD SD SD SD F
Ravens 47.11 46.80 47.61 41.06 27.64 12.17 **
Total 8.04 7.94 6.55 11.92 10.49
(a>e) (b>e) (Oe) (d>e)
F (4, 73) * £ < .05; * * £ < .0 1 .
Note: (>) indicates significant differences between Objective Language
Categories
(Tukey: £ < .05)
Following an Analysis o f Variance between the Objective Language
Category variable and the NeSBHis-R Language Domain (Table 9), Hypothesis
B was confirmed on three of four subtests. A significant F was noted on the
Ponton-Satz Boston Naming Test (PS-BNT) and two of the Verbal Fluency Test
categories (FAS Total and Animals), and approached significance on the third
Verbal Fluency Category (Fruits), F (4,73) = 2.41, £ = .06.
Post hoc comparisons of Hypothesis C showed that Strong Balanced
Bilinguals did not perform better than the Strong Unbalanced Bilinguals. Rather,
Strong Balanced Bilinguals and Strong Spanish Dominant Bilinguals performed
better than Strong English Dominant Bilinguals on the PS-BNT and two o f the
Verbal Fluency Categories (FAS Total and Animals). This finding suggests that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
stronger or equal Spanish language skills, as compared to English language skills,
are most associated with better performance on the NeSBHis-R Language Domain.
Post hoc comparisons of Hypothesis D demonstrated that Strong Spanish
Monolinguals performed better on the NeSBHis-R Language Domain than Strong
English Dominant Bilinguals; however, significant differences were not noted
between Strong Spanish Bilinguals (balanced and unbalanced) and the Strong
Spanish Monolinguals. Hypothesis E was confirmed in that Strong Spanish
Dominant Bilingual subjects performed significantly stronger than Weak Spanish
Monolingual subjects on the NeSBHis-R Language Domain with the exception of
the Verbal Fluency “Fruits” subtest. Strong Balanced Bilinguals and Strong
Spanish Monolinguals only performed better than Weak Spanish Monolingual
subjects on the Verbal Fluency subtest (FAS Total).
Following an Analysis o f Variance between the Objective Language
Category variable and the NeSBHis-R EF-AC Domain (Table 10), significant F’s
were noted on all but one of the neuropsychological subtests: Digit Span Forward.
Post hoc comparisons were associated with Hypothesis C. Among bilingual
subjects, no differences were noted between Strong Spanish Dominant Bilinguals
and Strong Balanced Bilinguals; however, both o f these groups performed better
than Strong English Dominant Bilinguals on the Stroop subtest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Table 9
A n a ly sis o f V a ria n c e : M e a n C o m p a ris o n o f N e S B H is-R L a n g u a g e D o m a in
S u b te sts b v F iv e O b je c tiv e L a n g u a g e C ateg o ries
Balanced
Bilingual
“Strong”
English
Dominant
Bilingual
“Strong”
Spanish
Dominant
Bilingual
“Strong”
Spanish
Mono
lingual
“Strong”
Spanish
Mono
lingual
“Weak”
(a) (b) (c) (d) (e)
M
SD
M
SD
M
SD
M
SD
M
SD F
Ponton-
Satz
Boston
Naming
Test (Total)
23.61
3.27
(a>b)
18.20
6.78
25.61
3.33
(Ob,e)
23.78
4.78
(d>b)
20.71
2.27
6.52 **
Verbal
Fluency
(FAS
Total)
35.17
8.42
(a>e)
28.60
10.54
39.00
8.81
(Ob,e)
39.83
7.61
(d>b,e)
26.07
7.50
7.75**
Verbal
Fluency
(Animals)
17.78
4.10
(a>b)
12.70
3.95
20.67
3.85
(Ob,e)
19.78
4.32
(d>b)
15.93
3.00
8.64**
Verbal
Fluency
(Fruits)
13.33
2 . 6 8
1 1 . 2 0
2.90
14.39
3.74
14.44
4.12
1 2 .2 1
2.39
2.41
F (4, 73), * £ < .0 5 ; * * £ < .0 1 .
Note: (>) indicates significant differences between objective language categories
(Tukey’s £ < .05)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
T able 10
A nalysis o f V arian ce: M e a n C o m p a riso n o f N e S B H is-R E x ecu tiv e
F u n c tio n /A tte n tio n -C o n c e n tra tio n D o m a in S u b te sts b v F iv e O b jectiv e L a n g u a g e
Categories
Balanced
Bilingual
“Strong”
English
Dominant
Bilingual
“Strong”
Spanish
Dominant
Bilingual
“Strong”
Spanish
Monolingual
“Strong”
Spanish
Monolingual
“Weak”
(a) (b) (c) C d) C e)
M
SD
M
SD
M
SD
M
SD
M
SD F
Stroop Color 68.17
8.08
(a>b)
81.70
9.58
65.06
10.83
COb)
63.56
12.73
Cd>b)
72.43
14.27 5.09**
Stroop Word 41.28
5.03
(a>b.e)
52.00
6.90
42.06
6.64
Cob.e)
42.28
6.42
Cd>b,e)
49.64
9.63 6.84**
Stroop Inter
ference
103.72
22.14
(a>b)
135.20
19.14
109.78
14.93
115.94
29.34
122.93
29.55 3.41**
Digit Span
Total
10.22
1.06
9.20
1.99
11.00
2.11
(O d,e)
9.28
1.64
8.71
1.68 4.63**
Digit Span
Forward
5.50
.71
4.90
.99
5.78
1.35
5.22
.89
5.07
.83 1.84
Digit Span
Backward
4.72
.75
(a>e)
4.30
1.25
5.22
1.06
COd.e)
4.11
.96
3.71
1.07 5.45**
Digit
Symbol
67.56
9.92
(a>tLe)
67.20
9.91
(b>e)
62.67
10.93
(O e)
54.28
15.09
45.14
11.82 9.44**
Color Trails
1
34.17
12.16
(a>d.e)
32.20
8.57
(b>d.e)
36.28
8.69
(O d)
49.22
18.80
48.07
13.98 5.48**
Color Trails
2
80.61
29.24
(a>e)
70.40
14.63
(b>d,e)
78.83
17.62
COe)
101.22
33.58
109.71
31.34 5.20**
F(4,73), * e < .0 5 ; * * g < .0 1 .
Note: (>) indicates significant differences between objective language categories
(Tukey’s jj < .05)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 7
Hypothesis D was confirmed on the Color Trails 1 subtest as all three
bilingual categories performed significantly better than the Strong Spanish
Monolingual subjects. Also, Strong Spanish Dominant Bilinguals performed
better than Strong Spanish Monolinguals on the Digit Span Total and the Digit
Span Backward. Hypothesis E showed variable post hoc results. For instance,
Strong English Dominant Bilinguals performed better than Weak Spanish
Monolinguals only on the Digit Symbol and Color Trails (1 and 2) subtests.
Strong Balanced Bilinguals performed better than Weak Spanish Monolinguals on
the Stroop (Word), Digit Span (Backward), Digit Symbol, and Color Trails (1 and
2) subtests. Strong Spanish Dominant Bilinguals performed better than Weak
Spanish Monolinguals on Stroop (Word), Digit Span (Total and Backward), Digit
Symbol, and Color Trails 2. Finally, Strong Spanish Monolinguals performed
better than Weak Spanish Monolinguals only on the Stroop Word subtest. Overall,
on the EF-AC Domain, it was noted that Strong Bilinguals perform better than
Strong and Weak Monolinguals on EF-AC tasks which include a fine motor,
visual-spatial component. Also, rapid naming o f colors and words proved to be
more difficult for Strong English Dominant Bilinguals than for Strong Balanced
Bilinguals and Strong Spanish Dominant Bilinguals.
An Analysis o f Variance test was run to identify relationships between the
Objective Language Category variable and the NeSBHis-R Memory Domain
(Table 11). Hypothesis A was only confirmed on the NeSBHis-R AVLT (Long-
Term Recall) subtest. All other memory domain subtests were not significant;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Table 11
Analysis o f Variance: Mean Comparison o f NeSBHis-R Memory Domain
Subtests bv Five Objective Language Categories
Balanced
Bilingual
“Strong”
English
Dominant
Bilingual
“Strong”
Spanish
Dominant
Bilingual
“Strong”
Spanish
Mono-lingual
“Strong”
Spanish
Mono
lingual
“Weak”
(a) (b) (c) (d) (e)
M
SD
M
SD
M
SD
M
SD
M
SD F
AVLT 53.67 50.40 56.83 53.28 51.93 2.04
(Total: 7.56 6.90 5.72 5.71 5.94
Lists 1-5)
AVLT 11.83 11.10 13.06 12.22 11.64 1.93
(Short-
Term
Recall)
1.98 1.91 1.43 2.07 2.56
AVLT 12.06 10.50 13.17 12.56 11.93 2.69**
(Long-
Term
Recall)
2.07 3.37 1.54
(O b)
1.76 2.27
AVLT 14.11 14.70 14.72 14.28 14.43 1.20
(Recog
nition)
1.41 .48 .75 .96 .76
Rey-
Osterreith
(Delay)
19.89
6.34
20.05
5.27
23.75
6.07
19.61
7.90
18.75
4.08
1.65
F (4,73), * £ < .05; ** 2 < -01.
Note: (>) indicates significant differences between objective language categories
(Tukey’s £ < .05)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
however, the NeSBHis-R AVLT (Total: Lists 1-5 and Short-Term Recall)
approached clinical significance (p = .09 and p = .12, respectively). Post hoc tests
provided some support for Hypothesis C in that Strong Spanish Dominant
Bilinguals obtained better scores on the AVLT (Long-Term-Recall) than Strong
English Dominant Bilinguals. Hypotheses D and E were not confirmed. Hence,
with the current sample, verbal output subtests (AVLT) on the NeSBHis-R
Memory Domain were slightly more sensitive to the Objective Language
Categories than their nonverbal counterpart (e.g., Rey-Osterreith Delay). Overall,
the Memory Domain demonstrated minimal influence by the Objective Language
Categories, which speaks to the strength of these subtests as measures of learning
and memory.
The NeSBHis-R Visual-Spatial Domain was compared to the Objective
Language Category variable using an Analysis of Variance test (Table 12).
Hypothesis B was confirmed on the Visual-Spatial Domain as demonstrated by
significant F’s on both the Block Design and Rey-Osterreith (Copy) subtests.
Hypothesis C was not confirmed; there were no differences between bilingual
categories. Relative to Hypothesis D, there was one difference between Strong
Bilinguals and Strong Spanish Monolinguals: Strong Spanish Dominant
Bilinguals performed better on the Block Design subtest than Strong Spanish
Monolinguals. Hypothesis 5 also favored Strong Spanish Dominant bilinguals in
that they performed better than Weak Spanish Monolinguals on both the Block
Design and Rey-Osterreith (Copy) subtests. The current results suggest that strong
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
bilingual skills may favor strong monolingual skills relative to visual-spatial
constructional abilities, and that Strong Spanish Dominant Bilinguals do better
than Weak Spanish Monolinguals on visual-spatial tasks. More notably, however,
the current findings show that bilingual and monolingual categories explain a
limited amount of variance on the NeSBHis-R Visual-Spatial Domain.
Table 12
Analysis of Variance: Mean Comparison of NeSBHis-R Visual-Spatial Domain
Subtests by Five Objective Language Categories
Balanced
Bilingual
“Strong”
English
Dominant
Bilingual
“Strong”
Spanish
Dominant
Bilingual
“Strong”
Spanish
Mono
lingual
“Strong”
Spanish
Mono
lingual
“Weak”
(a) (b) (c) (d) (e)
M
SD
M
SD
M
SD
M
SD
M
SD F
Block
Design
(Total)
36.17
4.85
37.50
7.12
39.67
4.35
(c>d.e)
33.72
8.29
31.57
5.97
4.08**
Rey-
Osterreith
(Copy)
34.36
1.68
32.75
3.07
34.78
L.06
(O e)
33.61
2.48
32.43
1.83
3.69**
E (4,73), * £ < .0 5 ; ** £ < • 0 1 .
Note: (>) indicates significant differences between objective language categories
(Tukey’s £ < .05)
The NeSBHis-R Motor Domain was compared to the Objective Language
Category variable using an Analysis o f Variance test (Table 13). While both
Dominant and Non-Dominant Grooved Pegboard F results were significant, post
hoc findings were not. Therefore, while objective language skills explain some of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
the variance relative to fine motor speed and precision, significant bilingual and
monolingual differences among the current sample were not evidenced on this
domain. Similarly, Hypotheses C through E on the Motor Domain were not
confirmed.
Table 13
Analysis of Variance: Mean Comparison of NeSBHis-R Motor Domain Subtests
bv Five Objective Language Categories
Balanced
Bilingual
“Strong”
(a)
English
Dominant
Bilingual
“Strong”
(b)
Spanish
Dominant
Bilingual
“Strong”
(c)
Spanish
Mono-lingual
“Strong”
(d)
Spanish
Mono
lingual
“Weak”
(e)
M M M M M
SD SD SD SD SD F
Grooved 65.06 69.50 66.61 75.11 77.07 2.62*
Pegboard
(Dominant
Hand)
7.46 7.03 10.15 20.20 13.87
Grooved 71.89 74.30 71.44 86.06 87.36 3.68*
Pegboard
(Non-
Dominant
Hand)
9.56 6.41 14.50 24.58 17.44
*
F (4,73), * g < .05; * * p < .01.
Note: (>) indicates significant differences between objective language categories
(Tukey’s p < .05)
In sum, a comparison o f the Objective Language Category variable to
NeSBHis-R Domains showed that the most salient findings were noted on the
NeSBHis-R Reasoning, Language, EF-AC, and Visual-Spatial Domains
(Hypothesis B). Results specific to Hypothesis C were variable. Balanced
Bilinguals (Strong) did not perform significantly better than Spanish Dominant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Bilinguals (Strong) on any NeSBHis-R subtest, but significant differences were
noted on the Language Domain and the EF-AC Domain, which favored Balanced
Bilinguals (Strong) to English Dominant Bilinguals (Strong). Also, Spanish
Dominant Bilinguals (Strong) performed significantly better than English
Dominant Bilinguals (Strong) on the NeSBHis-R Language, EF-AC, and Memory
Domains. Hypothesis D was insignificant relative to the NeSBHis-R Reasoning,
Motor, and Memory Domains. However, on the Language Domain, Strong
Spanish Monolinguals performed significantly better than Strong English
Dominant Bilinguals. Also, Strong Spanish Dominant Bilinguals did better than
Spanish Monolinguals (Strong) on the Visual-Spatial Domain (Block Design
subtest). Most notably, on the EF-AC Domain all three Bilingual categories
performed significantly better than Strong Spanish Monolinguals on the Color
Trails 1, and Spanish Dominant Bilinguals did better than Spanish Monolinguals
(Strong) on the Digit Span Total and Backward subtests. Finally, specific to
Hypothesis E, statistical significance was noted on the NeSBHis-R Reasoning
Domain. All Strong Categories (Bilingual and Monolingual) performed
significantly better than the Spanish Monolingual W eak Category on the Ravens
subtest. Variable significance was noted on the Language, EF-AC, and Visual-
Spatial Domains. The M otor and Memory Domains were not significant.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
Research Question 3 - Influence o f Objective Language
Categories on NeSBHis-R Domain Performance While
Controlling for Gender. Age, and Total Years o f Education:
To What Extent Does the Objective Language Category
Variable Still Explain a Significant Amount of the Variance
on NeSBHis-R Subtest Scores Following the Control of
Gender. Age, and Total Years o f Education?
This question follows the first two Research Questions. The importance of
considering demographic variables in a normative data collection was outlined in
Research Question 1. Research Question 2 focused on the variable o f Objective
Language Categories as an additional demographic variable that should be
considered when evaluating neuropsychological test results. Research Question 3
combines Research Questions 1 and 2 in an effort to explain whether the Objective
Language Category variable is still significant following the control of gender,
age, and total years o f education.
Hypothesis. As a follow-up to Research Question 2, it was hypothesized
that the Objective Language Category variable would still explain a significant
amount of variance associated with NeSBHis-R domain performance after
controlling for gender, age, and total years of education.
Analyses o f Covariance (ANCOVA) were used to address Hypothesis A.
First, the NeSBHis-R Reasoning Domain was compared to the Objective
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
Language Category variable after controlling for gender, age, and total years of
education (Table 14). The ANCOVA results show that objective language
categories still explain a significant amount o f the variance on Ravens subtest
performance.
Table 14
Analysis of Covariance: Comparison of NeSBHis-R Reasoning Domain by
Objective Language Categories. Controlling for Gender. Age, and Total Years of
Education
Objective Language Categories
_______ Ravens Total_________________________ 3.97**________________
F (4, 70), * p< .05; ** p < .0 1 .
A series of ANCOVA tests were run to determine if the NeSBHis-R
Language Domain was still significantly related to the Objective Language
Category variable after controlling for gender, age, and total years of education. In
comparison to the results from Research Question 2, performance on the PS-BNT
and Verbal Fluency (FAS Total and Animals) was still significantly related to the
Objective Language Category variable and performance on the Verbal Fluency
“Fruits” subtest was still not significant (Table 15).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Table 15
Analysis o f Covariance: Comparison of NeSBHis-R Language Domain Subtests
bv Objective Language Categories. Controlling for Gender. Age, and Total Years
ofEducation
Objective Language Categories
Ponton-Satz Boston Naming Test (Total)________________ 4.94**____________
Verbal Fluency (FAS Total)___________________________ 5.32**____________
Verbal Fluency (Animals)_____________________________ 7.10**____________
Verbal Fluency (Fruits)________________________________ 1.82_____________
F (4, 70), * £ < .0 5 ; ** p < .0 l.
The NeSBHis-R Executive Function/Attention-Concentration (EF-AC)
Domain was also still significantly influenced by the Objective Language
Category variable after controlling for gender, age, and total years of education. In
comparison to the results from Research Question 2, performance on the Stroop
(Color, Word, and Interference) and Digit Span Backward remained significant;
however, Digit Span Total, Digit Symbol, and the Color Trails (1 and 2) subtests
were no longer significant (Table 16). These results suggest that the Objective
Language Category variable continues to influence NeSBHis-R EF-AC Domain
subtests which require verbal output. For example, these verbal output subtests
remained significant and the Digit Span Total Score variable approached
significance, F(4, 70) = 2.41, p = .06. On the other hand, NeSBHis-R EF-AC
domain subtests which do not require verbal output and which are more reliant on
fine-motor visual tracking skills, were no longer significant.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
Table 16
Analysis of Covariance: Comparison of NeSBHis-R Executive
Function/Attention-Concentration Domain Subtests by Objective Language
Objective Language Categories
Stroop Color 7.15**
Stroop Word 5.96**
Stroop Interference 3.87**
Digit Span Total 2.41
Digit Span Forward .83
Digit Span Backward 3.72**
Digit Symbol 1.79
Color Trails 1 1.58
Color Trails 2 .39
F (4, 70), * £ < .05; * * p < .0 1 .
ANCOVA tests were also run to determine if the NeSBHis-R Memory
Domain subtests would remain largely unrelated to the Objective Language
Category variable after controlling for gender, age, and total years of education. In
comparison to the results from Research Question 2, performance on the AVLT
Long-Term Recall remained significant but all other NeSBHis-R subtests (AVLT
Total, AVLT Short-Term Recall, AVLT Recognition, and Rey-Osterreith (Delay)
were still not significantly related to the Objective Language Category variable
(Table 17).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
T a b le 17
Objective Laneuaee Cateeories. Controlling for Gender. Aee. and Total Years of
Education
Objective Language Categories
AVLT (Total: Lists 1-5) 1.52
AVLT (Short-Term Recall) 1.87
AVLT (Long-Term Recall) 2 .8 8 *
AVLT (Recognition) 1 .0 1
Rey-Osterreith (Delay) 1.27
F (4, 70), * £ < .05; ** g < .01.
The NeSBHis-R Visual-Spatial Domain, as compared to the Objective
Language Category variable, after controlling for gender, age, and total years of
education, was not significant. In comparison to the results from Research
Question 2, performance on both the Block Design and Rey-Osterreith (Copy)
subtests were no longer significant when compared to the Objective Language
Category variable when the additional control for gender, age, and total years of
education was considered (Table 18). These results suggest that NeSBHis-R
subtests are less reliant on bilingual and monolingual skills than other
demographic variables.
Table 18
Analysis of Covariance: Comparison of NeSBHis-R Visual-Spatial Domain
Subtests bv Objective Language Categories. Controlling for Gender. Age, and
Total Years of Education
Objective Language Categories
Block Design (Total)____________________________. 8 6
Rey-Osterreith (Copy)__________________________ 1.61
F (4, 70), * p < .05; * * g< .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
Finally, the NeSBHis-R Motor Domain remained largely unrelated to the
Objective Language Category variable after controlling for gender, age, and total
years o f education. In comparison to the results from Research Question 2,
performance on the Dominant Grooved Pegboard subtest no longer showed a
significant relationship to the Objective Language Category variable. Also, the
Non-Dominant Grooved Pegboard subtest was still not significantly related to the
Objective Language Category variable following a similar control for demographic
variables. Hence, the NeSBHis-R Motor Domain subtests continue to show
minimal to no influence by bilingual and monolingual language ability (Table 19).
Table 19
Analysis of Covariance: Comparison of NeSBHis-R Motor Domain Subtests by
Objective Language Categories. Controlling for Gender. Age, and Total Years of
Education
Objective Language Categories
Grooved Pegboard (Dominant Hand)_____________________ .46_____________
Grooved Pegboard (Non-Dominant Hand)_________________.77_____________
F (4, 70), * p < .05; ** p < .01.
Overall, Research Question 3 findings showed that with the additional
control for gender, age, and education, the Objective Language Category variable
continued to explain a significant amount of variance on the NeSBHis-R
Reasoning and Language Domains and was still largely unrelated to the Memory
Domain. The most notable changes, with the additional control, were on the
Visual-Spatial and Motor Domains; these NeSBHis-R Domains were no longer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
significant as compared to the Objective Language Category variable. In addition,
the EF-AC NeSBHis-R Domain remained significant as compared to the Objective
Language Category variable on subtests which require verbal output, but was no
longer significant on subtests which include a visual-spatial, fine-motor speed
component.
Research Question 4 - Influence of Culture Variables on
NeSBHis-R Domain Performance: To What Extent Does
Culture (e.g.. Language Use/Exposure. Behavior. Values')
Explain a Significant Amount o f the Variance on NeSBHis-
R Subtest Scores?
This question addresses two issues. First, it is a follow-up to the original
NeSBHis study in that it seeks to further clarify the influence o f culture on
neuropsychological test performance. Second, the following analyses also
attempted to clarify the role o f individual variables versus culture clusters on
NeSBHis-R performance. Culture variables and clusters are operationalized by
two culture measures: Cultural Identity Scales (CIS) and Marin Short
Acculturation Scale for Hispanics (Marin). Many culture measures have been
criticized for placing too much emphasis on language use and proficiency. As a
result, other aspects o f culture such as behavior and values are missed. The
current Research Question attempted to clarify the influence o f language
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
use/proficiency as well as behavior and values relative to neuropsychological test
performance.
The CIS includes 1 0 variables (3 = language; 4 = behavior; 3 = values).
Three CIS clusters were considered: (a) CIS English/American Cluster, (b) CIS
Spanish/Latino Cluster; and (c) CIS Spanish Latino - All Variables - Cluster.
Since the CIS includes four additional Spanish/Latino variables, it is important to
note that the CIS Spanish/Latino Cluster and the CIS English/American Cluster (a
and b above) represent one variable from each CIS Domain (Language, Behavior,
Values) which directly correspond between English/American and Spanish/Latino
questionnaire items. The final cluster, “Spanish/Latino - All Variables - Cluster”
(c above), represents the exact title because all seven Spanish/Latino questionnaire
items were considered. Following is a breakdown of each cluster and their
respective questionnaire items: (a) CIS English/American Cluster: English
Proficiency; Familiarity with American Culture; Feminism (American Value), (b)
CIS Spanish/Latino Cluster: Spanish Proficiency; Familiarity with Latino Culture;
“Respeto” (Latino Value), (c) CIS Spanish/Latino - All Variables - Cluster:
Spanish Proficiency; Spanish Language Preference; Familiarity with Latino
Culture; Latino Activism; Preferred Latino Affiliation; Perceived Discrimination;
and “Respeto” (Latino Value).
The Marin variables include: Language Use and Ethnic Loyalty, Media,
and Ethnic Social Relations. The Marin Cluster (total) combines the three Marin
variables. It is also important to note that while the CIS provides separate scores
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
for English/American and Spanish/Latino variables, the Marin does not.
Therefore, note that negative correlations on the Marin represent increased Spanish
language use and exposure and that positive correlations represent increased
English language use and exposure. All variables within the NeSBHis-R and
culture measures are presented as continuous variables. Also, as with previous
Research Questions, it is important to note that the Stroop, Color Trails, and
Grooved Pegboard subtest scores are based on the time required to finish the test;
therefore, lower scores on these subtests reflect better performance.
Hypothesis A. It was hypothesized that the CIS and the Marin measures
would yield the same constructs with the current sample.
The CIS consists of three domains: (a) Language Domain, (b)
Behavior/Familiarity Domain, and (c) Value/Attitude Domain. The language
domain consists of three variables. A reliability analyses o f each variable and its
respective items demonstrated strong correlations for each variable (Table 20).
Table 20
Reliability Analysis: Cultural Identity Scales. Language Domain
Spanish Spanish Language English
Proficiency Preference Proficiency
Item Numbers: 1,2,3 31, 32, 33, 34, 35 4, 5,6
Reliability
Coefficients: .9358 .8631 .9592
The Behavior/Familiarity domain consists of four variables. A reliability analyses
of each variable and its respective items demonstrated strong correlations for three
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
of the four variables (Table 21). The Latino Activism variable showed a weak
correlation (.57); however, when the items were collapsed the correlations were
significantly stronger. For example, when item number 24 “How often do you
participate in civic or political activities (e.g., demonstrations and protests) for
your Latino Community?” and item number 25 “How often do you participate in
minority organizations or fraternities in your community or school?” were
collapsed, the correlation was .70. AJso, when item 28 “If I could choose a
teacher, I would ... (5-point Likert scale: 1 = accept only a Latino/a; 5 = accept
only a White)” and item 29 “If I could choose my doctor, I would ... (5-point
Likert scale: 1 = accept only a Latino/a; 5 = accept only a White)” were collapsed,
the correlation was .74. With the current sample, these two categories, activism in
your community and choice o f professional representation represented distinct
issues. Nevertheless, the Latino Activism variable was not changed for the current
analyses; therefore, the results for the Latino Activism variable were interpreted
with caution.
Table 21
Reliability Analysis: Cultural Identity Scales. Behavior/Familiaritv Domain
Familiarity with
American
Culture
Familiarity with
Latino Culture Latino Activism
Preferred Latino
Affiliation
Item
Numbers: 8. 10. 12, 14 7. 9, 11, 13 24, 25, 28, 29 26, 27 .3 0
Reliability
Coefficients: .8741 .8664 .5702
(24,25 = .7010)
(28,29 = .7363)
.7930
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
The Value/Attitude domain consisted of three variables. A reliability analyses of
each variable and its respective items demonstrated strong correlations for each
variable (Table 22).
Table 22
Reliability Analysis: Cultural Identity Scales. Value/Attitude Domain
Perceived “Respeto” Feminism
Discrimination (Latino Value) (American Value)
Item Numbers: 21. 22. 23 18, 19 20 15, 16. 17
Reliability Coefficients: .7517 .7801 .8784
Overall, the CIS was found to be a reliable measure when compared to the current
sample and served as a primary measure for the purpose of testing subsequent
hypotheses and answering additional Research Questions.
The Marin consists of three variables: (a) Language Use/Ethnic Loyalty,
(b) Media, and (c) Ethnic Social Relations. A reliability analyses o f each variable
and its respective items demonstrated strong correlations for each variable (Table
23).
Table 23
Reliability Analysis: Marin Short Acculturation Scale for Hispanics
Language Use
Ethnic Loyalty Media
Ethnic Social
Relations
Item
Numbers: 1, 2, 3, 4, 5 6 , 7,8 9, 10, 11, 12
Reliability
Coefficients: .8800 .9054 .8260
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
Thus, the Marin was found to be a reliable measure when compared to the current
sample and therefore also served as a primary measure for the purpose of testing
subsequent hypotheses and answering additional Research Questions.
Hypothesis B. It was hypothesized that the CIS and Marin measures would
show similar results on comparative culture variables. For example, it was
predicted that both the CIS and Marin would evidence similar results on Language
specific variables: CIS Language domain, Marin Language Use/Ethnic Loyalty
variable, and Marin Media variable. Similarly, it was predicted that both the CIS
and the Marin would yield the same results on Behavior specific variables: CIS
Behavior Domain and Marin Ethnic Social Relations variable.
Hypothesis C. NeSBHis-R subtest scores will show the strongest
correlations with Language based culture variables (CIS Language Domain; Marin
Language Use/Ethnic Loyalty; and Marin Media), followed by Behavior variables
(CIS Behavior Domain; Marin Ethnic Social Relations), and Values variables will
show little or no significant correlations (CIS Values Domain).
Hypothesis D. It was hypothesized that the CIS multidimensional model
(e.g., separate scores for English/American variables and Spanish/Latino variables)
would be a more sensitive model in that it would show bicultural influences on
NeSBHis-R that the Marin unidimensional model would not.
First, the NeSBHis-R Reasoning Domain was compared to each CIS and
Marin variable and cluster. The first series o f correlations focused on the CIS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
Language Domain variables. On this CIS domain, significant correlations were
identified on all three variables and the highest correlation was noted on the
“English Proficiency” variable (Table 24).
Table 24
Scales. Laneuaee Domain
Spanish Spanish Language English
Proficiency Preference Proficiency
Ravens Total .31 ** .49 ** .62 **
r (78), *£< .05; **£<.01.
The second series of correlations focused on the CIS Behavior/Familiarity Domain
variables. On this domain, significant and positive correlations were identified on
three o f the four variables (Table 25). The highest correlation was noted on the
“Familiarity with American Culture” variable suggesting a moderate relationship
with performance on the Ravens test. A weak correlation was identified between
the “Familiarity with Latino Culture” variable and the Ravens total score. Thus,
with the current sample, knowledge and access to American culture increased
Ravens total scores at a greater rate than having knowledge of, and access to,
Latino culture. The “Latino Activism” variable was not significantly correlated
with performance on the Ravens subtest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
T a b le 25
Comparison of NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales. Behavior/Familiarity Domain
Familiarity Familiarity Preferred
with American with Latino Latino Latino
Culture Culture ActiviiisL. Affiliation
Ravens Total .45 ** .36** . 0 3 ^ .23 *
I (78), *£ < .05; * * p < .0 1 .
The third series o f correlations focused on the CIS Value/Attitude Domain
variables. On this domain, significant correlations were identified on two of the
three variables (Table 26). The highest correlation was on the “Feminism”
variable, but was closely followed by the “Respeto” variable suggesting that
identification with both American and Latino values is correlated (weak
correlations) with better performance on the Ravens test. The “Perceived
Discrimination” variable was not significantly correlated with performance on the
Ravens test.
Table 26
Comparison of NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales. Value/Attitude Domain
Perceived “Respeto” Feminism
Discrimination (Latino Value) (American Value)
Ravens Total .15 .35 ** .40 **
r (78), *£<.05; **£<.01.
The NeSBHis-R Reasoning Domain was also compared to the Marin
variables. The results demonstrated positive correlations on all three variables as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
compared to performance on the Ravens test (Table 27). The strongest correlation
was on the Media variable.
Table 27
Comparison o f NeSBHis-R Reasoning Domain Subtest bv Marin Variables
Language Use/
Ethnic Loyalty Media Ethnic Social Relations
Ravens Total .35 ** .51 ** .44 **
r(78), *2 < 05; **£<.01.
A comparison of both CIS and Marin clusters was also completed and all
the correlations were in the moderate range (Table 28).
Table 28
Scales and Marin Clusters
Cultural Cultural Identity
Identity Scales Cultural Identity Scales
English/ Scales Spanish/Latino Marin
American Spanish/Latino (All Variables) Cluster
Cluster Cluster Cluster (Total)
Ravens Total .54 ** .43 ** .46 ** .50 **
r (78), *£< .05; **£<.01.
Overall, there was a moderate correlation between culture and the
NeSBHis-R Reasoning domain. The CIS reflected significant correlations on both
English/American and Spanish/Latino variables of each domain (Language,
Behavior, and Values); however, it was also noted that the strongest correlations
were associated with English/American variables and clusters. Both measures
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
suggested that greater familiarity with the American culture and greater English
language use/proficiency improves performance on the Ravens test.
It was noted that Hypothesis B was confirmed because there were not
significant differences noted between the CIS and the Marin results; therefore,
both measures identified positive and significant correlations secondary to
language and behavior assessment as compared to the NeSBHis-R Reasoning
Domain. Hypothesis three was not confirmed because both the CIS and the Marin
showed moderate correlations on both the Language and Behavior variables.
Actually, the CIS Values variable also showed significant correlations in similar
ranges. Hence, the Language variables did not prove to be the strongest
correlation, as compared to Behavior and Values variables. Lastly, it was noted
that Hypothesis D was confirmed. For instance, while both measures suggest that
greater familiarity and exposure with American culture and higher English
proficiency/use improves performance on the Ravens test, the CIS was able to
identify that Spanish proficiency and identification with Latino Behavior and
Values were also significantly correlated with Ravens total scores. This finding
shows support for the multidimensional model.
Next, the NeSBHis-R Language Domain was compared to each CIS and
Marin variable and cluster. The first series of correlations focused on the CIS
Language Domain and significant correlations were only identified on the
“Spanish Proficiency” variable (Table 29). In fact, moderate, positive correlations
between “Spanish Proficiency” and the Ponton-Satz Boston Naming Test and two
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
o f the Verbal Fluency tests were noted. This result indicates that with the current
sample, only “Spanish Proficiency” and not “Spanish Language Preference” nor
“English Proficiency” is significantly related to the NeSBHis-R Language
Domain. As one would expect, only Spanish language proficiency impacts
Spanish language tests.
Table 29
Comparison o f NeSBHis-R Language Domain Subtests bv Cultural Identity Scales
Language Domain
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Ponton-Satz Boston
Naming Test (Total)
.57** -.08 . 0 0
Verbal Fluency
(FAS Total)
4 5 **
. 0 2 . 1 2
Verbal Fluency
(Animals)
.41 ** -.16 -.09
Verbal Fluency
(Fruits)
.36 ** - . 1 0 - . 0 1
f(78), *p< .05; **£<.01.
The second series o f correlations focused on the CIS Behavior/Familiarity
Domain (Table 30). On this CIS domain, significant and positive correlations
were only noted on the “Familiarity with Latino Culture” variable. These corre
lations were in the weak to moderate range. Weak correlations were identified on
two o f the Verbal Fluency tests (FAS Total and Animals). Stronger correlations
were noted on the PS-BNT and the third Verbal Fluency subtest (Fruits). Thus,
with the current sample, access to or familiarity with American culture does not
appear to improve performance on the NeSBHis-R Language Domain.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
T a b le 30
Scales. Behavior/Familiaritv Domain
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Ponton-Satz
Boston Naming
Test (Total)
.08 .42 ** -.04 .05
Verbal Fluency
(FAS Total)
.13 -.08 .2 1
Verbal Fluency
(Animals)
- . 1 2 .25 * -.13 .06
Verbal Fluency
(Fruits)
. 1 0 .43 ** .03 - . 0 1
r (78), *£< .05; **£< .01.
The third series o f correlations focused on the CIS Value/Attitude Domain.
On this CIS domain, none of the variables were significantly correlated with any
o f the NeSBHis-R Language Domain subtests (Table 31). This result implies that
the current sample’s values (American and Latino), as assessed by the CIS, did not
impact performance on any of the NeSBHis-R Language subtests.
Table 31
Scales. Value/Attitude Domain
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Ponton-Satz Boston
Naming Test (Total)
-.06 .00 .03
Verbal Fluencv
(FAS Total)
.11 .14 .14
Verbal Fluency
(Animals)
.00 -.14 .02
Verbal Fluency (Fruits) -.16 -.02 -.07
r (78), *£ < .05; **£ < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
The NeSBHis-R Language Domain was also compared to the Marin
variables and significant correlations were only noted on the “Language
Use/Ethnic Loyalty” variable (Table 32). These correlations were negative and
while they were weak, they suggest that as Spanish language use and proficiency
increases, so do scores on the PS-BNT and the Verbal Fluency (Animals) subtests.
Unlike the CIS, which yielded significant results on both the Spanish Proficiency
and the Behavior/Familiarity domains, only language use/proficiency was
significant on the Marin. Also, the CIS yielded stronger correlations.
Nevertheless, both the Marin and the CIS were consistent in the identification of
higher Spanish Language proficiency/use relative to stronger performance on
NeSBHis-R Language subtests.
Table 32
Comparison of NeSBHis-R Language Domain Subtests bv Marin Variables
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Ponton-Satz Boston
Naming Test (Total)
-.33 ** . 0 2 .06
Verbal Fluency
(FAS Total)
. 0 1 . 1 0 .0 1
Verbal Fluency
(Animals)
-.30 ** -.05 -.07
Verbal Fluency
(Fruits)
- . 2 1 .04 .08
r (78), *g < .05; **£ < .0 1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
A comparison of both CIS and Marin clusters was also completed and the
results noted that only the CIS was significantly correlated with the NeSBHis-R
Language Domain (Table 33). Among the CIS clusters, the “Spanish/Latino”
Cluster was positively correlated with three of the four NeSBHis-R Language
Domain subtests (PS-BNT, Verbal Fluency: FAS Total and Fruits). Moreover,
the Verbal Fluency (FAS Total) was also significantly correlated with the more
comprehensive “Spanish/Latino (All Variables)” cluster.
Table 33
and Marin Clusters
Cultural
Identity Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Ponton-Satz
Boston Naming
Test (Total)
.01 .35 ** .1 0 -.15
Verbal Fluencv
(FAS Total)
.18 ' .41 ** .27* .04
Verbal Fluency
(Animals)
-.07 .19 .03 - .2 0
Verbal Fluency
(Fruits)
- .0 2 .28 * .04 -.07
r (78), *£<.05; **£<.01.
While both the CIS and the Marin measures noted that increased Spanish
language use/proficiency increased performance on most NeSBHis-R Language
subtests, only the CIS noted a positive correlation between behavior (“Familiarity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
with Latino Culture”) and all NeSBHis-R Language subtests. Similarly, the Marin
cluster was not significantly correlated with NeSBHis-R Language subtests, but
the CIS Spanish/Latino Cluster was.
Relative to the identified hypotheses, it was noted that Hypothesis B (the
CIS and Marin measures will show similar results on comparative culture
variables) showed variable results. The Language-based culture variable was most
consistent between the CIS and the Marin, but while the CIS was significantly
correlated with all the NeSBHis-R Language subtests, the Marin was only
significantly correlated with PS-BNT and the Verbal Fluency “Animals” subtests.
The Behavior variables were only significantly correlated with the CIS.
Hypothesis C was confirmed as Language was most significant, Behavior
followed, and Values was not significant. Hypothesis D was not confirmed as
both culture measures were only significantly correlated with Spanish/Latino
culture variables when compared to the NeSBHis-R Language subtests. However,
it is important to note that only the CIS Spanish/Latino cluster was significantly
correlated with the NeSBHis-R Language subtests. It is likely that the Marin’s
single continuum (e.g., Spanish to English) is not as sensitive as the CIS’s separate
continuum’s (e.g., one continuum for Spanish and a separate continuum for
English) with separate high and low scales for each continuum.
Focus is now placed on comparing the NeSBHis-R Executive Function/
Attention-Concentration (EF-AC) Domain to each CIS and Marin domain/variable
and cluster. The first series of correlations focused on the CIS Language Domain
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
variables (Table 34). On this CIS domain, significant correlations were identified
on all three CIS Language variables. On the Stroop (Color, Word, and
Interference), stronger Spanish proficiency showed improved performance on this
test. On the Color Trails subtests (1 and 2), greater Spanish language preference
and stronger English language proficiency was correlated with improved test
performance. Both the Stroop and the Color Trails subtest correlations were in the
moderate range. On the Digit Span subtest (Total and Forward), weak and positive
correlations were identified on the Spanish proficiency variable indicating that as
Spanish proficiency increased so did performance on these subtests. The Digit
Span Backward subtest suggested that both Spanish and English proficiency
contribute to improved recall on this subtest, although these correlations were
weak. The Digit Symbol subtest reflected significant correlations on all three CIS
Language variables; however, the strongest correlation was on the English
Proficiency variable. This last result suggests that while Spanish Proficiency
influences improved Digit Symbol test performance, this correlation was weak.
Higher Spanish Language Preference and English Proficiency showed moderate to
strong correlations with improved Digit Symbol performance showing a slight
preference for higher English Proficiency skills.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
T a b le 3 4
Subtests bv Cultural Identitv Scales. Laneuaee Domain
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Stroop Color
_ 4 4 **
.03 -.05
Stroop Word -.46 ** -.08 -.15
Stroop Interference -.48 ** - . 1 1 - . 2 0
Digit Span
Total
.39 ** .19 .2 1
Digit Span
Forward
.39 ** . 1 0 .07
Digit Span
Backward
.27 * .2 1 .26 *
Digit Symbol .30 ** .57 ** .64 **
Color Trails 1 -.07 -.43 ** -.45 **
Color Trails 2 -.18 -.43 ** -.54 **
r (78), *£<.05; **£<.01.
The second series o f correlations focused on the CIS Behavior/Familiarity
Domain variables (Table 35). On this CIS domain, significant correlations were
noted on each CIS variable. On the Stroop (Color, Word, and Interference) greater
Familiarity with Latino Culture improved test performance, although these
correlations were in the weak to moderate range. On the Digit Span subtests
(Total, Forward, and Backward), higher Latino Activism showed a weak
correlation with improved test performance; however, these correlations should be
interpreted with caution based on the low reliability o f the Activism variable as
compared to the current sample. Greater Preference for Latino Affiliation was also
positively correlated with the Digit Symbol (Total subtest); this was also a weak
with permission of the copyright owner. Further reproduction prohibited without permission.
96
correlation. The Digit Symbol and Color Trails (2) subtests showed that greater
Familiarity with American Culture improved test performance; these correlations
were in the weak to moderate range. The Color Trials (!) subtest scores also
showed significant improvement when Familiarity with Latino Culture increased,
although Familiarity with American Culture still had a stronger influence. On the
Color Trails (1) subtest, however, less Familiarity with American Culture
improved test performance. This last correlation was in the weak range.
Table 35
Subtests bv Cultural Identity Scales. Behavior/Familiaritv Domain
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Stroop Color . 0 0 -.31 ** -.05 .06
Stroop Word - . 2 0 -.40 ** - . 0 2 .1 1
Stroop
Interference
-.13 -.32 ** - . 1 2 .16
Digit Span
Total
.09 . 2 0 .33 ** . 2 2 **
Digit Span
Forward
.0 1 .19 .30 ** .18
Digit Span
Backward
. 1 1 .14 .29 ** .2 1
Digit Symbol
4 9 **
. 2 2 .13 .14
Color Trails 1
3 3 **
- . 1 1 - . 1 0 -.06
Color Trails 2 -.38 ** -.24* - . 1 2 -.14
r (78), *p< .05; **£<.01.
The third series of correlations focused on the CIS Value/Attitude Domain
variables. On this CIS domain, significant correlations were noted in all three CIS
variables (Table 36). On the Stroop subtests, stronger identification with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
Feminism Value improved test performance (Color, Word, and Interference). The
Stroop Word subtest also showed that greater identification with Latino Values
improved test performance. On the Digit Symbol subtest, higher identification
with both American and Latino Values improved test performance. The Color
Trails subtest correlations (1 and 2) showed that greater Perceived Discrimination
and greater identification with American Values (Feminism) improved test
performance. The latter (American Values) showed a stronger correlation with
improved Color Trials performance but all these correlations were in the weak to
moderate range. The Digit Span subtests were not significant as compared to the
CIS Value/Belief Domain.
Table 36
Subtests bv Culture Identity Scales Value/Attitude Domain
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Stroop Color -.04 -.14
. 3 4 **
Stroop Word .06 -.25 * -.25 *
Stroop Interference .04 -.16 -.29 *
Digit Span
Total
. 1 0 .09 .06
Digit Span
Forward
. 1 2 .03 -.03
Digit Span
Backward
.04 .09 . 1 1
Digit Symbol .18 .38 **
4 3 **
Color Trails 1 -.28 * -.16 -.33 **
Color Trails 2 -.23 * -.19
. 4 5 **
r (78), *p< .05; **p<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
The NeSBHis-R Language Domain was also compared to the Marin
variables and the results demonstrated significant correlations on all three Marin
variables (Table 37). All three Marin variables were significantly correlated with
the Digit Symbol and Color Trails (1 and 2) subtests. One additional correlation
was noted on the Stroop (Color) subtest as compared to the “Language Use/Ethnic
Loyalty” variable. The Stroop (Color) subtest showed a weak correlation and
suggests that increased English language use lowers test performance. The Digit
Symbol correlations were in the weak to moderate range and also showed that
increased English language use and exposure improves test performance.
However, the Color Trails subtests (1 and 2) generally revealed moderate
correlations which suggest that increased English Language use and exposure
improves test performance.
Table 37
Subtests bv Marin Variables
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Stroop Color . 2 2 * - . 1 0 .04
Stroop Word .0 1 -.16 -.08
Stroop Interference .04 -.06 -.13
Digit Span Total . 0 1 . 2 0 .19
Digit Span Forward -.07 . 1 2 . 1 2
Digit Span Backward .06 .19 . 2 0
Digit Symbol .43 ** .55 **
3 7 **
Color Trails 1 -.32**
. 3 9 ** - 3 5 **
Color Trails 2 -.25 * -.43 ** - . 2 2 **
r (78), * p < .05; **p<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
Both CIS and Marin clusters were compared and the results noted
significant correlations on all clusters (Table 38). The Stroop (Color, Word, and
Interference) was only significantly correlated with the CIS clusters. The negative
correlation suggests that higher scores on all clusters improve test performance.
So it is likely that greater identification with culture improves Stroop test
performance. The Digit Span subtest was positively correlated with the
Spanish/Latino CIS clusters showing that higher Spanish/Latino identification
improved Digit Span performance and that English/ American identification does
not. Finally, the Color Trails (1 and 2) and Digit Symbol subtests were significant
on both the CIS and Marin clusters, suggesting that greater identification with
culture variables in general, but particularly American Culture, improved test
performance.
The NeSBHis-R EF-AC Domain showed that Digit Span subtests (Total,
Forward, and Backward) were most correlated with greater Spanish language
proficiency/use, which is consistent with the verbal output that is required by these
subtests. The Color Trails subtests showed that greater identification with culture,
especially American Culture, increased test performance. Conversely, greater
identification with Spanish/Latino variables improved Stroop performance. While
both Spanish/Latino and English/American variables (CIS and Marin) increased
Digit Symbol performance, there was preference for the latter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T a b le 38
C o m p a riso n o f N e S B H is-R E x e c u tiv e F u n c tio n /A tte n tio n -C o n c e n tra tio n D o m a in
S u b te sts b v C u ltu ra l Id e n tity S c a le s a n d M a rin D o m a in s
Cultural
Identity Scales
English/
American
Cluster
Cultural
Identity
Scales
S panish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Stroop Color -.08 -.35 ** -.16 .09
Stroop Word -.25 * -.48 ** -.25 * -.08
Stroop Interference
-.25 * -.42 ** -.26 * -.04
Digit Span Total .16 .28 * .35 ** .13
Digit Span Forward
.04 .25 * .29* .05
Digit Span Backward
.2 0 .2 2 * .30 ** .16
Digit Symbol
.60** .39 ** .50 ** .53 **
Color Trails 1 -.45 ** -.19 -.39 ** -.41 **
Color Trails 2
-.53 ** -.28 *
- 42 **
-.35 **
r (78), *£< .05; **£<.01.
Overall, Hypothesis B was supported. The Language variable was
confirmed on both the Marin and the CIS (e.g., Digit Symbol performance
increased as English Proficiency/Use increased), and the Behavior variable was
also consistent (e.g., Digit Symbol and Color Trails 1 and 2 both showed better
scores as identification with American behaviors increased). Hypothesis C was
only partially confirmed because both CIS and Marin Language and Behavior
variables appeared to show similar significance on NeSBHis-R EF-AC subtests.
However, although the Values variables were significantly correlated with EF-AC
subtest scores, it was still less than the approximately equal influence o f both
Language and Behavior variables. Hypothesis D was confirmed as the Marin only
permission of the copyright owner. Further reproduction prohibited without permission.
101
showed English/American significance, but the CIS showed both
English/American and Spanish/Latino significance on EF-AC subtests.
Results which compare the NeSBHis-R Memory Domain, the CIS, and the
Marin are now presented. The first series o f correlations focused on the CIS
Language Domain variables. None o f the NeSBHis-R Memory Domain subtests
were significantly correlated with the three CIS Language variables with the
exception of the AVLT (Long-Term Recall). This subtest was positively
correlated with the “Spanish Proficiency” variable; however, this correlation was
very weak (Table 39). The remaining correlations associated with the CIS, the
Marin, and the NeSBHis-R Memory Domain were not statistically significant
(Tables 40-43).
Table 39
Laneuaee Domain
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
AVLT
(Total: Lists 1-5)
.17 .05 -.04
AVLT
(Short-Term Recall)
.13 . 0 0 -.04
AVLT
(Long-Term Recall)
.23 * -.03 -.06
AVLT
(Recognition)
- . 1 0 - . 0 1 . 0 2
Rey-Osterreith
(Delay)
.08 . 1 2 .19
r (78), *£ < .05; **£ < .0 1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
T ab le 4 0
C o m p a riso n o f N e S B H is-R M e m o ry D o m a in S u b te sts b v C u ltu re Id e n tity S cales
Behavior/Familiarity Domain Variables
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
AVLT (Total:
Lists 1-5)
- . 0 1 .07 . 0 2 -.14
AVLT (Short-
Term Recall)
-.09 .03 .13 -.13
AVLT (Long-
Term Recall)
-.13 .08 .14 -.18
AVLT
(Recognition)
. 0 1 - . 0 1 .04 -.04
Rey-Osterreith
(Delay)
-.03 .08 .05 -.03
r (78), *p< .05; * * £ < .0 1 .
Table 41
Comparison o f NeSBHis-R Memory Domain Subtests bv Cultural Identity Scales.
Value/Attitude Domain
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
AVLT
(Total: Lists 1-5)
-.06 .03 . 0 2
AVLT
(Short-Term Recall)
- . 0 1 -.03 -.06
AVLT
(Long-Term Recall)
- . 1 0 -.06 -.06
AVLT
(Recognition)
.08 .19 .19
Rey-Osterreith
(Delay)
. 2 0 .05 .15
r (78), *£ < .05; **£ < .0 1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T a b le 42
C o m p a riso n o f N e S B H is -R M e m o ry D o m a in S u b te sts bv M a rin V a ria b le s
103
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
AVLT
(Total: Lists 1-5)
-.05 .08 -.08
AVLT
(Short-Term Recall)
-.10 .07 -.11
AVLT
(Long-Term Recall)
-.18 .03 -.17
AVLT
(Recognition)
.01 .07 -.06
Rey-Osterreith
(Delay)
-.01 .22 .14
I (78), *2 <-05; **2 <.01.
Table 43
Comparison o f NeSBHis-R Memory Domain Subtests bv Cultural Identity Scales
and Marin Clusters
Cultural Identity
Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Settles
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
AVLT
(Total: Lists 1-5)
. 0 2 .13 .08 - . 0 2
AVLT
(Short-Term
Recall)
- . 0 2 .09 .09 -.06
AVLT
(Long-Term
Recall)
-.06 .11 .05 -.13
AVLT
(Recognition)
. 1 0 .08 .08 .01
Rey-Osterreith
(Delay)
.09 .07 . 1 2 .1 2
1(78), * 2 < -05; **2< 01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
Culture variables (CIS and Marin) were not significantly correlated with
the NeSBHis-R Memory Domain subtests. Only one correlation was noted and
while it was weak it was on the Spanish Proficiency variable. Since the AVLT
relies on verbal output, this correlation is likely explained by the language
component of both the AVLT test and the CIS Spanish Proficiency variable.
Therefore, the general findings of the NeSBHis-R Memory Domain and culture
variables suggest that the AVLT and the Rey-Osterreith Delay subtests are likely
most associated with learning and memory skills and less associated with
American or Latino cultural factors.
Overall, due to the largely insignificant findings relative to culture and the
NeSBHis-R Memory Domain, Hypotheses B, C, and D were only partially
supported. Hypothesis B was not confirmed because only the CIS showed a
significant correlation. Hypothesis C was supported because only the Language
variable associated with culture was significant, Behavior and Values were not.
Finally, Hypothesis D was confirmed because only the CIS was able to identify a
significant correlation with the Memory Domain. While this finding could be
related to multidimensional model, it is more likely that culture has a minimal
statistical relationship with the NeSBHis-R. Memory Domain.
Next, the NeSBHis-R Visual-Spatial Domain was compared to each CIS
and Marin variables and clusters. The first series o f correlations focused on the
CIS Language Domain variables (Table 44). On this CIS domain, significant
correlations were identified on each CIS variable when compared to the Block
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Design subtest. However, while “Spanish Proficiency” and “Spanish Language
Preference” both revealed weak correlations, the “English Proficiency” correlation
was in the moderate range. In other words, while both Spanish and English
language variables were positively correlated with Block Design subtest scores,
stronger English proficiency showed a significantly higher influence on improved
Block Design performance. On the other hand, significant and positive
correlations relative to the Rey-Osterreith (Copy) subtest were noted on both
Spanish and English proficiency variables, but both correlations were in the weak
range.
Table 44
Scales. Laneuaee Domain
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Block Design
(Total)
.26 * .26 * .44 **
Rey-Osterreith
(Copy)
3 7 **
. 2 0 .35 **
r (78), *£< .05; **£<.01.
The second series of correlations focused on the CIS Behavior/Familiarity
Domain (Table 45) and while all significant correlations were positive, they were
also relatively weak. Both the Block Design and Rey-Osterreith (Copy) were
significantly correlated with increased “Familiarity with American Culture.”
Increased “Familiarity with Latino Culture” was correlated with the Rey-Osterreith
but not with the Block Design subtest and “Preferred Latino Affiliation” was only
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
correlated with the Block Design subtests. “Latino Activism” was not
significantly correlated with the NeSBHis-R Visual-Spatial subtests. Hence, both
visual-spatial subtests showed slight improvement with both American and Latino
behavior domains.
Table 45
Scales. Behavior/Familiaritv Domain
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Block Design .23 ** .14 .06 .28 *
(Total)
Rey-Osterreith .25 * .30 ** -.05 .15
(Copy)
1(78), *p< .05; **£<.01.
On the CIS Value/Attitude Domain (Table 46), the “Feminism” variable was
positively correlated with both the Block Design and Rey-Osterreith subtests;
however, these correlations were weak. A similar weak and positive correlation
was associated with the Block Design subtest and the CIS “Perceived
Discrimination” variable. The Latino Value (Respeto) was not significantly
correlated with NeSBHis-R Visual-Spatial subtests.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
Table 46
Comparison o f NeSBHis-R Visual-Spatial Domain Subtests bv Cultural Identity
Scales. Value/Attitude Domain
“Respeto” Feminism
Perceived (Latino (American
Discrimination Value) Value)
Block Design (Total) .30 ** .18 .29 **
Rey-Osterreith (Copy) .16 .18 .25 *
r (78), *£ < .05; **£<.01.
A comparison of the Marin variables and the NeSBHis-R Visual-Spatial
Domain yielded weak correlations and all but one was significant, the Rey-
Osterreith (Copy) compared to the “Language Use/Ethnic Loyalty” variable (Table
47). Thus, increased English language use and exposure reflected higher visual-
spatial subtest scores.
Table 47
Comparison of NeSBHis-R Visual-Spatial Domain Subtests bv Marin Variables
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Block Design (Total) . 2 2 * .29 * .29 *
Rey-Osterreith (Copy) - . 0 2 .27 * .25 *
I (78), *p<.05; **p<.01.
Similar weak correlations were noted upon comparison o f the NeSBHis-R
Visual-Spatial Domain to the CIS and Marin clusters. All but one comparison was
significant, the Marin Cluster as compared to the Rey-Osterreith (Copy)
(Table 48).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
T a b le 4 8
Scales and Marin Clusters
Cultural
Identity Scales
English/
American
Cluster
Cultural
Identity Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Block Design
(Total)
.36 ** .24* .35 ** .30
**
Rey-Osterreith
(Copy)
.28 * .32 ** .26 * .16
r (78), * 2 < 05; * * 2 < -01.
Weak correlations were noted between the NeSBHis-R Visual-Spatial
Domain and culture variables (CIS and Marin). Nevertheless, significant variables
suggest that the Block Design subtest is associated with better English language
skills, as this was the highest correlation noted on the Visual-Spatial Domain. In
general, greater knowledge o f the American culture, greater identification with
American values, and greater use o f the English language were associated with
both the Block Design and Rey-Osterreith (Copy) subtests.
Overall, the hypotheses associated with Research Question 4 showed
confirmation. Hypothesis B, for example, was supported in that both the CIS and
Marin yielded positive and significant correlations between Language and
Behavior Culture Variables and NeSBHis-R Visual-Spatial subtests. On the other
hand, Hypothesis C showed minimal confirmation because while English language
was still the strongest correlation on Block Design, other Language, Behavior, and
Values correlations with both Block Design and Rey-O were not significantly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
different from one another. Finally, Hypothesis D was confirmed because the CIS
demonstrated significant correlations on both the English/American and
Spanish/Latino domains while the Marin only yielded significance which was
associated with English/American identification.
The final series of analyses which were associated with Research Question
4 include the comparison the NeSBHis-R Motor Domain subtest to the culture
variables (CIS and Marin). First, this NeSBHis-R domain was compared to the
CIS Language Domain variables (Table 49). The results indicate that there is a
significant and moderate relationship between higher “English Proficiency” and
better scores on the Grooved Pegboard subtest (D and ND). Other significant
correlations were noted on the Grooved Pegboard subtest (N and ND) as compared
to Spanish language CIS variables; these correlations were weak and suggest that a
greater preference to use the Spanish language, rather than greater Spanish
language proficiency, are associated with better performance on the Grooved
Pegboard test.
Table 49
Laneuaee Domain
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Grooved Pegboard
(Dominant Hand)
-.31 ** -.36 **
_ 4 5 **
Grooved Pegboard
(Non-Dominant
Hand)
- . 2 1 -.36 ** -.46 **
r (78), * p < .0 5 ; **£<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
In comparing the Grooved Pegboard (D and ND) subtest to the CIS
Behavior/Familiarity variables, significant (weak) correlations reflected better
Grooved Pegboard (D and ND) performance with greater “Familiarity with
American Culture.” Only one other significant correlation was noted, which
associated greater “Familiarity with Latino Culture” to better Grooved Pegboard
performance (Dominant Hand only), but this correlation was the weakest on this
Table (Table 50). Therefore, greater Familiarity with American Culture was most
associated with improved Grooved Pegboard performance (Dominant and Non-
Dominant).
Table 50
Behavior/Familiaritv Domain
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Grooved
Pegboard
(Dominant
Hand)
. 3 4 **
-.26 * .04 . 1 2
Grooved
Pegboard
(Non-Dominant
Hand)
-.32 ** -.19 - . 0 0 .05
r (78), < .05; **p < .01.
A similar pattern was noted in comparing the Grooved Pegboard (D and
ND) to the CIS Values/Attitude variables (Table 51). All significant correlations
were weak, but it was noted that better Grooved Pegboard (D and ND)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
performance was most associated with greater identification with American values
(“Feminism”). One other significant correlation was noted: greater identification
with Latino values (“Respeto”) reflected better Grooved Pegboard (only D)
performance.
Table 51
Comparison o f NeSBHis-R Motor Domain Subtests bv Cultural Identity Scales.
Value/Attitude Domain
Perceived
Discrimination
“Respeto”
(Latino
Value)
Feminism
(American
Value)
Grooved Pegboard - . 2 1 -.35 ** -.33 **
(Dominant Hand)
Grooved Pegboard -.18 -.19 -.31 **
(Non-Dominant
Hand)
I (78), *u< -05; **£<.01.
When compared to the Marin variables, the Grooved Pegboard (D and ND)
was significantly correlated with both the “Media” and the “Ethnic Social
Relations” variables. The correlations were in the weak to moderate range and
suggest that Grooved Pegboard scores (D and ND) improved when the current
sample endorsed higher exposure to the English language (Table 52).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
T a b le 52
C o m p a riso n o f N e S B H is -R M o to r D o m a in S u b te sts bv M a rin V ariab les
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Grooved Pegboard -.17 -.42 ** -.34 **
(Dominant Hand)
Grooved Pegboard - . 2 0 -.45 ** -.32 **
(Non-Dominant
Hand)
r(7B), *£<.05; **£<.01.
A review of the culture measure clusters compared to performance on the
Grooved Pegboard (D and ND) revealed significant correlations on every cluster
(Table 53). In other words, Grooved Pegboard (D and ND) performance improved
as both English/American and Spanish/Latino identification increased. The
relationships were in the weak to moderate range and it is important to note the
strongest and most consistent correlations were reflected with higher identification
with English/American language and culture.
Table 53
and Marin Clusters
Cultural
Identity Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin Cluster
(Total)
Grooved Pegboard
(Dominant Hand)
-.40 ** -.38 ** -.33 ** -.34 **
Grooved Pegboard
(Non-Dominant
Hand)
-.40 ** -.24 * -.29 ** -.36 **
r (78), *£<.05; **£<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
The most consistent theme that was noted upon review of the NeSBHis-R
Motor Domain was that stronger Grooved Pegboard (D and ND) performance was
significantly correlated with higher English language proficiency, higher
identification with American culture and values, and higher exposure to the
English language.
In sum, variable support was noted with reference to the Motor Domain
and the stated hypotheses. Hypothesis B was confirmed because both the Marin
and CIS showed significant correlations on both the Language and Behavior
variables. Hypothesis C was partially confirmed because while English
proficiency showed the strongest correlations, there was not a notable difference
between the Behavior and Values domains. Finally, Hypothesis D was confirmed
because the CIS yielded significant correlations on both the English/American and
Spanish/Latino domains while the Marin did not reveal positive Spanish/Latino
correlations.
A summary of Research Question 4 indicates that Culture was significantly
and positively correlated with all the NeSBHis-R Domains. The Memory Domain
showed the smallest relationship with Culture Domains (e.g., only CIS Spanish
Proficiency on AVLT Long-Term Recall), the Reasoning, EF-AC, Visual-Spatial,
and M otor Domains showed the strongest relationships with Culture Domains
(e.g., language, behavior, and values), and the Language Domain was positively
correlated with two of the Culture Domains (e.g., language and behavior). Hence,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
Hypothesis B was confirmed on all but one NeSBHis-R Domain (Memory) as both
the CIS and the Marin yielded significant Language and Behavior variables.
Hypothesis C showed the strongest support on three NeSBHis-R Domains
(Reasoning, Language, EF-AC); on these Domains, the strongest correlations were
noted on the Culture “Language” Domain, followed by lower correlations on the
Culture “Behavior” Domain, and little or no influence was noted on the Culture
“Values” Domain. On these Domains, it was noted that the Culture “Language”
Domain loaded the strongest, was followed by lower correlations on the Culture
“Behavior” Domain, and little or no influence was noted on the Culture “Values”
Domain. The Culture Language Domain was also the highest on the Motor and
Visual-Spatial Domains. Hence, while only Spanish/Latino variables were
significant on the NeSBHis-R Language Domain, both Spanish/Latino and
English/American variables were significant on four NeSBHis-R Domains
(Reasoning, EF-AC, Visual-Spatial, and Motor). Also, while all these domains
were influenced by both cultures (English/American and Spanish/Latino), the
Reasoning, Visual-Spatial, and Motor Domains showed preference for the
English/American variables. It was also noted that the EF-AC Domain reflected
an English/American advantage on nonverbal tasks and a Spanish/Latino
advantage on verbal tasks.
Finally, Hypothesis D showed strong support as the CIS multidimensional
model was able to identify positive contributions from both Spanish/Latino and
English/American Culture Domains on four of the NeSBHis-R Domains
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
(Reasoning, EF-AC, Visual-Spatial, and Motor). It was also noted that on the
Memory Domain, only the CIS was able to identify the Spanish Proficiency as a
contributing factor on Long-Term recall (AVLT). Hence, while the Marin’s
unidimensional model was able to identify the significant contribution of the
Spanish language on the NeSBHis-R Language Domain, it only yielded significant
other English/American Culture influences on other NeSBHis-R Domains. Hence,
although the unidimensional model is able to identify the significant contribution
of English/American culture on NeSBHis-R subtests, one is left to assume that
then Spanish/Latino contributions are not present. The risk of this interpretation is
that the significant relationship could be explained as an English/American
cultural confound when in fact the contribution may be best explained as both
English/American and Spanish/Latino “language and culture” confounds.
Research Question 5 - Influence of Culture Variables on
NeSBHis-R Domain Performance While Controlling for
Gender. Age, and Total Years of Education: To What
Extent do Culture Variables Still Explain a Significant
Amount o f the Variance on NeSBHis-R Subtest Scores
Following the Control of Gender. Age, and Total Years of
Education?
Hypothesis. As a follow-up to Research Questions 1, 2, and 4, it was
hypothesized that culture (CIS and Marin) would still be significantly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
correlated with NeSBHis-R domain performance after controlling for gender, age,
and total years of education.
Pearson Product Moment Correlations were run to compare performance
on each NeSBHis-R Domain to each culture (CIS and Marin) variable and cluster,
while select demographics (gender, age, and total years o f education) were
partialed out. The results of these analyses are compared to the findings o f the
previous Research Question (Research Question 4). Results which were specific
to the NeSBHis-R Reasoning Domain are presented first.
The first series o f correlations focused on the CIS Language Domain and
showed that following a control of gender, age, and total years of education, the
CIS “Spanish Proficiency” variable was no longer significant, but “Spanish
Language Preference” and “English Proficiency” remained significant (Table 54).
While these correlations were weak, they still suggest that greater preference for
the Spanish language and higher English language proficiency were associated
with better performance on the Ravens subtest, and that stronger English
Proficiency was most related to higher scores on the Ravens.
Table 54
Comparison o f NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales. Laneuaee Domain. Controlling for Gender. Aee. and Total Years of
Education
Spanish Spanish Language English
Proficiency Preference Proficiency
Ravens Total -.01 .26 ** .37 **
Partial r (74), < .05; **p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
The second series o f correlations focused on the Ravens subtest and the
CIS Behavior/Familiarity Domain variables. These results showed that with the
additional control of gender, age, and total years of education, none of the CIS
Behavior/Familiarity Domain variables were significantly correlated with
performance on the Ravens subtest (Table 55). This result differs from the
significant correlations that were noted when answering Research Question 4:
without the control, “Familiarity with American Culture,” Familiarity with Latino
Culture,” and “Preferred Latino Affiliation” were all significantly correlated with
the Ravens subtest.
Table 55
Comparison of NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales. Behavior/Familiarity Domain. Controlling for Gender. Age, and Total
Years o f Education
Familiarity with Familiarity Preferred
American with Latino Latino Latino
____________________Culture_______ Culture_____Activism____ Affiliation
Ravens Total__________ .07___________ .07_________ .00__________ .08______
Partial r (74), *p < .05; **j> < .01.
The third series o f correlations focused on the CIS Values/Attitude Domain
variables as compared to Ravens subtest performance. The results showed that
none o f the correlations were significant (Table 56), which differs from the
significant relationship that was noted in Research Question 4: both American and
Latino values were correlated with Ravens test performance.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
Table 56
Comparison o f NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales. Value/Attitude Domain. Controlling for Gencer. Age, and Total Years of
Education
Perceived “Respeto” Feminism
_______________ Discrimination (Latino Value) (American Value)
Ravens Total________ .00_____________ T1______________ .05_______
Partial r (74), *£<.05; **£< .01.
Performance on the Ravens subtest was also compared to the Marin
variables. It was noted that while the “Language Use/Ethnic Loyalty” and “Ethnic
Social Relations” variables remained significant when correlated with Ravens test
performance, the “Media” variable did not (Table 57). Therefore, with the control
o f gender, age, and total years o f education, statistical results showed that
increased use of the English language and greater exposure to the English
language, socially, was correlated with better performance on the Ravens subtest.
These correlations were in the weak to moderate range.
Table 57
Comparison of NeSBHis-R Reasoning Domain Subtest bv Marin Variables.
Controlling: for Gender. Aee. and Total Years o f Education
Language Use/ Ethnic Social
Ethnic Loyalty Media Relations
Ravens Total .23 ** .21 .24 *
Partial r (74), *p < .05; **£ < .01.
The results from the correlations which compared performance on the
Ravens subtest to both the CIS and Marin clusters showed that only the Marin
cluster remained significant when the control of gender, age, and total years of
education was added (Table 58). Without the control (e.g., Research Question 4),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
all the clusters were significant. Thus, a significant correlation between the
Ravens and the Marin cluster suggests that as English language use and exposure
increases, so does performance on the Ravens subtest; however, it is important to
note that while significant, this correlation was weak.
Table 58
Comparison o f NeSBHis-R Reasoning Domain Subtest bv Cultural Identity
Scales and Marin Clusters. Controlling for Gender. Age, and Total Years of
Education
Cultural Cultural Cultural
Identity Scales Identity Identity Scales
English/ Scales Spanish/Latino Marin
American Spanish/Latino (All Variables) Cluster
Cluster Cluster Cluster (Total)
Ravens Total . 2 0 . 1 1 .18 .28 **
Partial r (74), *p < .05; **p < .0 1 .
A summary of the results which explain the relationship between the
NeSBHis-R Reasoning Domain and Culture (CIS and Marin), while controlling
for gender, age, and total years of education, suggests that only increased English
proficiency and increased exposure to/use of the English language remains
significantly correlated with better performance on the Ravens subtest (Table 58).
Without the control factor, Spanish proficiency was not significant but preference
for the Spanish language was significantly correlated with improved performance
on the Ravens subtest.
The NeSBHis-R Language Domain will now be compared to culture (CIS
and Marin) with the additional control o f gender, age, and total years of education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The first series o f correlations focuses on the CIS Language Domain variables and
these results showed that the “Spanish Proficiency” variable remained significantly
correlated with all the NeSBHis-R Language subtests. In addition, compared to
Research Question 4, the “Spanish Language Preference” variable was now
negatively correlated with two of the Verbal Fluency subtests (Animals and
Fruits). In other words, as Spanish language preference increased, performance on
these Verbal Fluency subtests declined; these correlations were weak (Table 59).
Table 59
Scales. Laneuaee Domain. Controlline for Gender. Aee. and Total Years of
Education
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Ponton-Satz Boston
Naming Test (Total)
.51** -.18 -.11
Verbal Fluency
(FAS Total)
.35 ** - . 1 0 . 0 0
Verbal Fluency
(Animals)
.38 ** -.24 * - . 2 0
Verbal Fluency
(Fruits)
.23 ** -.24 * -.16
Partial r (74), *p < .05; **g < .01.
A comparison of the NeSBHis-R Language Domain and the CIS Behavior/
Attitude variables, with the additional control of gender, age, and total years of
education, showed that the “Familiarity with Latino Culture” variable was still
significantly correlated with three of the four NeSBHis-R Language subtests: PS-
BNT, Verbal Fluency (FAS Total), and Verbal Fluency (Fruits). The Verbal
permission of the copyright owner. Further reproduction prohibited without permission.
121
Fluency (Animals) subtest remained significant, but on a different CIS variable:
“Familiarity with American Culture” showed a negative correlation with the
Verbal Fluency (Animals) subtest, which suggests that as familiarity with
American culture increases, performance on the Verbal Fluency (Animals) subtest
declines (Table 60). The previous Research Question (Research Question 3)
showed that increased familiarity with Latino culture, improved performance on
the Verbal Fluency (Animals) subtests. However, the overall results suggest that
while greater knowledge o f the Latino culture may improve performance on the
Verbal Fluency (Animals) subtest, less knowledge of the American culture
improves the current samples’ ability to access words associated with the semantic
category “Animals.” All correlations on Table 60 were weak.
Table 60
Scales. Behavior/Familiaritv Domain. Controlline for Gender. Ace, and Total
Years o f Education
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Ponton-Satz
Boston Naming
Test
(Total)
- . 1 0 .29 * -.09 - . 0 2
Verbal Fluency
(FAS Total)
-.04 .23 * -.13 .16
Verbal Fluency
(Animals)
-.28 * .17 .14 . 0 2
Verbal Fluency
(Fruits)
-.09 .31 * - . 0 1 -.08
Partial r (74), *g < .05; **p < .01.
with permission of the copyright owner. Further reproduction prohibited without permission.
122
A review o f the correlations which compared the NeSBHis-R Language
Domain to the CIS Values/Attitude variables showed that when controlling for
gender, age, and total years o f education, only one additional correlation was
significant: the “Respeto” variable showed a weak and negative correlation with
the Verbal Fluency (Animals) subtest (Table 61). In other words, as identification
with the Latino value (“Respeto”) increased, performance on the Verbal Fluency
(Animals) subtest decreased. Prior to the control, none o f the Value Attitude
Domain correlations were significant.
Table 61
Comparison o f NeSBHis-R Language Domain Subtests bv Cultural Identity
Education
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Ponton-Satz Boston
Naming Test (Total)
-.05 -.16 -.03
Verbal Fluency
(FAS Total)
. 1 1 . 0 0 .06
Verbal Fluency
(Animals)
- . 0 1 -.24 * -.03
Verbal Fluency
(Fruits)
-.18 -.19 - . 2 1
Partial r (74), *p < .05; **p < .01.
Comparisons between the Marin variables and performance on the
NeSBHis-R Language Domain showed that the additional control for gender, age,
and total years of education yielded similar results as correlations which did not
include the control. For example, only the “Language Use/Ethnic Loyalty”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
variable was significantly correlated with all the NeSBHis-R Language subtests.
The correlations were negative and ranged from the weak to moderate range.
Also, with the control factor, one additional NeSBHis-R Language subtest was
significant. The Verbal Fluency (Fruits) subtest showed a significant correlation.
In sum, less English language use (or greater Spanish language use) was correlated
with better performance on the NeSBHis-R Language subtests (Table 62).
Table 62
Comparison of NeSBHis-R Language Domain Subtests bv Marin Variables.
Controlling for Gender. Age, and Total Years of Education
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Ponton-Satz Boston
Naming Test (Total)
-.40 ** -.07 - . 0 1
Verbal Fluency
(FAS Total)
-.05 - . 0 1 - . 1 0
Verbal Fluency
(Animals)
-.34 ** -.13 -.13
Verbal Fluency
(Fruits)
-.28 * -.08 -.05
Partial r (74), *p < .05; **p < .01.
A review of the culture clusters (CIS and Marin) and the NeSBHis-R
Language Domain, with and without the control factor, showed variability.
Without the control factor, only the Spanish/Latino cluster was positively
correlated with all the NeSBHis-R Language subtests. With the control factor,
only one Verbal Fluency subtest (FAS Total) showed significance. In addition,
with the control factor, the Marin cluster was negatively correlated with the PS-
permission of the copyright owner. Further reproduction prohibited without permission.
124
BNT and one of the Verbal Fluency subtests (Animals), suggesting that greater
Spanish language use and exposure improved performance on these subtests
(Table 63).
Table 63
Comparison o f NeSBHis-R Language Domain Subtests bv Cultural Identity Scales
and Marin Clusters. Controlling for Gender. Age, and Total Years of Education
Cultural
Identity
Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural
Identity Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Ponton-Satz
Boston Naming
Test (Total)
-.14 . 2 0 -.05 -.26 *
Verbal Fluency
(FAS Total)
.05 .29 * .16 -.06
Verbal Fluency
(Animals)
-.19 . 1 1 -.05 -.28 *
Verbal Fluency
(Fruits)
- . 2 2 * . 1 0 -.15 -.19
Partial r (74), *p < .05; **p < .01.
In sum, the addition of the control factor when comparing the NeSBHis-R
Language Domain to culture (CIS and Marin) showed that Language subtests
remained significantly correlated with increased Spanish language proficiency,
increased identification with Latino culture, and decreased English language use.
The NeSBHis-R Executive Function/Attention-Concentration (EF-AC)
Domain is now compared to culture (CIS and Marin) with the additional control of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
gender, age, and total years of education. The following results are compared to
findings which did not include the control factor (Research Question 3).
First, the NeSBHis-R EF-AC Domain subtests were compared to the CIS
Language Domain variables (Table 64). The results showed that with the
additional control factor, “Spanish Proficiency” was still significantly correlated
with the Stroop (Color, Word, and Interference). Two o f the Digit Span subtests
also remained significantly correlated with “Spanish Proficiency”— Total and
Forward. The third Digit Span subtest (Backward) was no longer significantly
correlated with “Spanish Proficiency” nor “English Proficiency.” The Digit
Symbol subtest was still significantly correlated with “Spanish Language
Preference” and “English Proficiency” and the Color Trails 1 subtest was now
only significantly correlated with “Spanish Language Preference.” Without the
control factor greater Spanish Proficiency influenced higher Stroop scores, and
greater Spanish Proficiency showed higher Digit Span scores. Also, English
Proficiency and Spanish Language Preference, but not Spanish Proficiency,
increased Digit Symbol scores. Performance on Color Trails 1 improved as
Spanish Language Preference increased. In general, Color Trails 1 and 2 were not
significantly correlated with the CIS Language Domain.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
T a b le 64
Subtests bv Cultural Identitv Scales Laneuaee Domain. Controlling for Gender.
Aee. and Total Years o f Education
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Stroop Color -.38 ** .19 .16
Stroop Word -.30 * . 1 2 .07
Stroop Interference -.39 ** .09 . 0 2
Digit Span Total .30 * . 0 2 - . 0 2
Digit Span Forward .30 * - . 0 1 -.11
Digit Span Backward .2 1 .06 .08
Digit Symbol .04 .31 * .31 *
Color Trails 1 .08 -.26 * - . 2 1
Color Trails 2 .04 -.15 -.18
Partial r (74), *p < .05; **p < .01.
Second, the NeSBHis-R EF-AC Domain subtests were compared to the
CIS Behavior/Familiarity variables. With the control factor, it was noted that
“Familiarity with Latino Culture” and “Familiarity with American Culture” was no
longer significantly correlated with the following EF-AC subtests: all Stroop
subtests (“Latino”), Digit Symbol and both Color Trails subtests (“American”).
The Digit Span subtests (Total, Forward and Backward) remained significantly
correlated with the “Latino Activism” variable and the Stroop (Interference)
remained significantly correlated with the “Preferred Latino Affiliation” variable.
All significant correlations were in the weak range (Table 65).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
T a b le 65
Subtests bv Cultural Identity Scales. Behavior/Familiaritv Domain. Controlling for
Gender. Aee. and Total Years of Education
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Stroop Color .23 * - . 2 2 -.05 .14
Stroop Word .07 - . 2 1 .06 .24 *
Stroop
Interference
.13 -.19 -.08 .26 *
Digit Span
Total
-.16 .06 .35 ** .15
Digit Span
Forward
- . 2 1 .04 .34 ** . 1 0
Digit Span
Backward
-.09 .06 .29 * .16
Digit Symbol .15 -.05 .09 - . 0 1
Color Trails 1 -.13 . 0 0 - . 1 1 .04
Color Trails 2 -.05 - . 1 0 -.15 - . 0 0
Partial r (74), *g < .05; * * 2 < .01.
Third, the NeSBHis EF-AC subtests were compared to the CIS
Values/Attitude variables. It was noted that, with the control factor, none of the
correlations were significant as observed on Table 6 6 . Without the control factor,
significant correlations had been noted between the American value (“Feminism”)
and the Stroop (Color, Word, and Interference) subtests, the Digit Symbol subtest,
and the Color Trails (1 and 2) subtests. The Latino value (“Respeto”) had also
shown a significant correlation with the Digit Symbol subtest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
T a b le 66
Subtests bv Cultural Identitv Scales. Value/Attitude Domain. Controlling for
Gender. Age. and Total Years of Education
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Stroop Color . 0 2 -.03 -.13
Stroop Word .09 -.04 - . 1 2
Stroop Interference . 1 0 . 0 2 -.13
Digit Span
Total
.0 1 -.06 -.18
Digit Span
Forward
.07 -.08 - . 2 0
Digit Span
Backward
-.05 -.04 -.11
Digit Symbol .03 .13 - . 0 0
Color Trails 1 -.17 - . 0 1 -.07
Color Trails 2 -.06 .06 -.09
Partial r (74), *j> < .05; **p < .0 1 .
A comparison o f the Marin variables to the NeSBHis-R EF-AC variables
also showed that the control factor minimized the level o f statistically significant
correlations between these variables (Table 67). For instance, the “Media” and
“Ethnic Social Relations” variables were no longer significantly correlated with
the Digit Symbol and Color Trails (1 and 2) subtests. Moreover, the “Language
Use/Ethnic Loyalty” variable was now only significantly correlated with the Digit
Symbol subtest and not the Color Trails (1 and 2) subtests. It was noted, however,
that the Stroop (Color) subtest remained significantly correlated with the
“Language Use/Ethnic Loyalty” variable. It is important to note that the variables
that remained significant on the Marin, as compared to the NeSBHis-R subtests,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
were positive correlations, meaning that as English language use and exposure
increased, so did performance on the identified NeSBHis-R subtests.
T a b le 67
Subtests bv Marin Variables. Controllinc for Gender. Aee. and Total Years of
Education
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Stroop Color .33 * .06 .18
Stroop Word .13 .03 .09
Stroop Interference .18 .18 . 0 2
Digit Span
Total
- . 1 1 - . 0 0 .05
Digit Span
Forward
-.13 - . 0 2 .03
Digit Span
Backward
-.07 . 0 0 .07
Digit Symbol .24 * . 2 1 .09
Color Trails 1 -.18 -.17 - . 2 1
Color Trails 2 - . 0 1 -.06 .08
Partial r (74), *p < .05; **p < .01.
A final comparison was made between the culture clusters and the
NeSBHis-R EF-AC subtests with the additional control factor (Table 6 8 ). Most
notable, was that much less variance was associated with the English/American
and Spanish/Latino clusters as compared to NeSBHis-R subtests. Only the Color
Trail 1 subtest (better scores) was still significantly correlated with all culture
measure clusters. The Digit Span (Total) subtest continued to show a positive
relationship with the Spanish/Latino cluster and the Stroop (Color, Word, and
Interference) subtests were negatively correlated with the Spanish/Latino CIS
Cluster.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
Table 6 8
Comparison o f NeSBHis-R Executive Function/Attention-Concentration Domain
Subtests bv Cultural Identity Scales and Marin Clusters. Controlling for Gender.
Age, and Total Years of Education
Cultural
Identity
Scales
English/
American
Domain
Cultural
Identity
Scales
Spanish/Latino
Domain
Cultural
Identity Scales
Spanish/Latino
(All Variables)
Marin
Total
Stroop Color .13 -.26* - . 0 2 .26 *
Stroop Word - . 0 2 -.27* - . 0 2 - . 1 1
Stroop
Interference
- . 0 2 -.28 * -.07 -.17
Digit Span
Total
- . 1 0 .16 .23 * -.05
Digit Span
Forward
-.16 .13 .2 1 -.07
Digit Span
Backward
- . 0 1 .13 .19 - . 0 2
Digit Symbol . 2 2 .09 .2 1 .24 *
Color Trails 1 -.24 * -.06 -.26 * - . 2 2 *
Color Trails 2 - . 2 0 -.08 - . 2 0 - . 0 0
Partial r (74), *p < .05; **p < .01.
In sum, it was noted that relative to the NeSBHis-R EF-AC subtests and
the culture variables (CIS and Marin), the added control factor greatly limited the
amount o f variance that was explained by culture in the interpretation of NeSBHis-
R subtest performance. One of the main themes include less influence of English
proficiency on a nonverbal subtest (e.g., Color Trails). Moreover, Spanish
language proficiency continued to correlate significantly with the Stroop subtests
(Color, Word, and Interference), suggesting that stronger Spanish language
proficiency improved performance on the Stroop subtest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
The NeSBHis-R Memory Domain is now compared to culture (CIS and
Marin) with the additional control for gender, age, and total years o f education.
Because culture showed minimal statistical significance before and after the
control factor as compared to the NeSBHis-R Memory Domain, a summary of
Tables 69-73 is provided (Tables 69-73). Before the control factor (Research
Question 3), only the AVLT (Long-Term Recall) subtest showed a significant and
positive correlation with the CIS “Spanish Proficiency” variable. All other
correlations did not show statistical significance. With the control factor, it was
noted that the AVLT (Long-Term Recall) subtest was no longer significantly
correlated with the CIS “Spanish Proficiency” variable. Instead, the Rey-
Osterreith(Delay) subtest showed a statistically significant and negative correlation
with the CIS “Familiarity with American Culture” variable and the
“English/American” cluster, suggesting that as familiarity with American culture
increased, performance on the Rey-Osterreith(Delay) decreased. However, these
correlations were in the weak range and should, therefore, be interpreted with
caution.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
Table 69
Comparison o f NeSBHis-R Memory Domain Subtests bv Cultural Identity Scales.
Language Domain. Controlling for Gender. Age, and Total Years of Education
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
AVLT .09
(Total: Lists 1-5)
. 0 0 -.15
AVLT .12
(Short-Term Recall)
- . 0 0 -.06
AVLT .22
(Long-Term Recall)
-.05 -.11
AVLT -.14
(Recognition)
-.09 -.11
Rey-Osterreith -.02
(Delay)
-.09 -.09
Partial r (74), < .05; **p < .01.
Table 70
Comparison o f NeSBHis-R Memorv Domain Subtests bv Cultural Identity Scab
Behavior/Familiaritv Domain. Controlling for Gender. Age. and Total Years of
Education
Familiarity with
American
Culture
Familiarity
with Latino Latino
Culture Activism
Preferred
Latino
Affiliation
AVLT -.14
(Total:
Lists 1-5)
-.05 .03 - . 2 0
AVLT -.14
(Short-Term
Recall)
. 0 1 .14 -.15
AVLT -.22
(Long-Term
Recall)
.04 .13 - . 2 0
AVLT -.07
(Recognition)
- . 0 2 .06 -.07
Rey-Osterreith -.33 **
(Delay)
- . 0 1 .08 -.14
Partial r (74), < .05; **g < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
T a b le 71
Value/Attitude Domain. Controlling for Gender. Age. and Total Years of
Education
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
AVLT
(Total:
Lists 1-5)
-.09 -.03 -.05
AVLT
(Short-Term Recall)
- . 0 2 -.04 -.09
AVLT
(Long-Term Recall)
-.09 - . 1 1 -.09
AVLT
(Recognition)
. 0 2 . 2 0 .14
Rey-Osterreith
(Delay)
.08 -.05 - . 1 1
Partial r (74), *jj < .05; **p < .01.
Table 72
Comparison o f NeSBHis-R Memory Domain Subtests bv Marin Variables.
Controlling for Gender. Age, and Total Years o f Education
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
AVLT
(Total:
Lists 1-5)
-.07 . 0 2 -.15
AVLT
(Short-Term Recall)
- . 1 0 .07 -.13
AVLT
(Long-Term Recall)
- . 2 0 . 0 2 - . 2 0
AVLT
(Recognition)
-.04 - . 0 2 -.13
Rey-Osterreith
(Delay)
-.17 - . 0 1 - . 0 2
Partial r (74), < .05; **£< .0 1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
T a b le 73
and Marin Clusters. Controlling for Gender. Aae. and Total Years o f Education
Cultural Identity
Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
AVLT
(Total:
Lists 1-5)
-.08 .04 .0 0 -.08
AVLT
(Short-Term
Recall)
-.05 .09 .1 0 -.07
AVLT
(Long-Term
Recall)
- . 1 2 .08 .0 2 -.17
AVLT
(Recognition)
.0 2 .09 .04 -.07
Rey-Osterreith
(Delay)
-.23 * -.04 -.03 - .1 1
Partial r (74), *g < .05; **g < .01.
Similar to the NeSBHis-R Memory Domain, when the control factor was
added, the Visual-Spatial Domain showed minimal statistical significance when
compared to culture (CIS and Marin). Therefore, a summary of tables 74-78 is
provided (Tables 74-78). Before the control factor was applied, the Block Design
subtest showed statistically significant and positive correlations with all culture
clusters and most culture variables. While these results suggested that
performance on this subtest improved with use of both languages and exposure to
both cultures, the American/English influence was slightly higher. A more
balanced influence by both cultures and languages was noted on the Rey-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
Osterreith (Copy) subtest. However, when the control factor was added, none of
the correlations were statistically significant.
Table 74
Scales Laneuaee Domain. Controlline for Gender. Aee. and Total Years of
Education
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Block Design
(Total)
. 1 1 -.03 .13
Rey-Osterreith
(Copy)
.16 -.07 .04
Partial r (74), *p < .05; **g < .01.
Table 75
Comparison of NeSBHis-R Visual-Spatial Domain Subtests bv Culture Identity
Scales Behavior/Familiarity Domain. Controlling for Gender. Age, and Total
Years o f Education
Familiarity with
American
Culture
Familiarity
with Latino
Culture
Latino
Activism
Preferred
Latino
Affiliation
Block Design
(Total)
-.11 -.05 .08 .19
Rey-Osterreith
(Copy)
- . 1 2 .06 -.07 . 0 1
Partial r (74), *p < .05; **p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
Table 76
Comparison o f NeSBHis-R Visual-Spatial Domain Subtests bv Culture Identity
Scales Value/Attitude Domain. Controlling for Gender. Age, and Total Years of
Education
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Block Design
(Total)
.17 - . 0 0 -.05
Rey-Osterreith .05 -.05 -.05
(Copy)
Partial r (74), *p < .05; **p < .01.
Table 77
Comparison o f NeSBHis-R Visual-Spatial Domain Subtests bv Marin Variables.
Controlling for Gender. Age, and Total Years o f Education
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Block Design
(Total)
.05 -.06 .08
Rey-Osterreith - . 2 1 -.03 .04
(Copy)
Partial r (74), *p < .05; **g < .01.
Table 78
Comparison o f NeSBHis-R Visual-Spatial Domain Subtests bv Cultural Identity
Scales and Marin Clusters. Controlling for Gender. Age, and Total Years of
Education
Cultural Identity
Scales
English/
American
Cluster
Cultural Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Block Design
(Total)
. 0 2 .05 .15 .04
Rey-Osterreith
(Copy)
-.08 .05 - . 0 2 - . 1 2
Partial r (74), *p < .05; **p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
The NeSBHis-R Motor Domain, as the Visual-Spatial and Memory
Domains, showed minimal statistical significance, upon review of the correlations
with culture, when the control variables were added to the model. A review of
Tables 79-83 shows that the control factor greatly reduced the statistical
significance o f the correlations between the Grooved Pegboard (D and ND) and
culture (CIS and Marin). Prior to the control, the Grooved Pegboard subtest
demonstrated statistically significant correlations with most culture variables, but
was most influenced by the English language and American culture variables.
Higher scores on these two culture variables improved performance on the
Grooved Pegboard subtest. Nevertheless, with the control factor, the Grooved
Pegboard subtest only showed statistically significant correlations on the
“Preferred Latino Affiliation” variable (Dominant hand), the “English Proficiency”
variable (Non-Dominant hand), and the “Media” variable (Non-Dominant hand).
Thus, with the control factor, the findings associated with the NeSBHis-R Motor
Domain and culture do not support a statistically significant relationship between
these variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
Table 79
Comparison o f NeSBHis-R Motor Domain Subtest bv Cultural Identity Scales
Language Domain. Controlling for Gender. Age, and Total Years of Education
Spanish
Proficiency
Spanish Language
Preference
English
Proficiency
Grooved Pegboard
(Dominant Hand)
- . 1 2 -.15 - . 2 1
Grooved Pegboard
(Non-Dominant
Hand)
-.07 -.19 -.25 *
Partial r (74), *p < .05; **p < .01.
Table 80
Comparison o f NeSBHis-R Motor Domain Subtest bv Cultural Identity Scales.
Behavior/Familiarity Domain. Controlling for Gender. Age, and Total Years of
Education
Familiarity with Familiarity Preferred
American with Latino Latino Latino
Culture Culture Activism Affiliation
Grooved
Pegboard
(Dominant
Hand)
-.06 -.06 .09 .28 *
Grooved
Pegboard
(Non-
Dominant
Hand)
- . 1 0 -.06 - . 0 1 .17
Partial r (74), *p < .05; **p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
T a b le 81
C o m p a riso n o f N e S B H is-R M o to r D o m a in S u b te st b v C u ltu ral Id e n tity S c a le s
Value/Attitude Domain. Controlling for Gender. Age, and Total Years of
Education
Perceived
Discrimination
“Respeto”
(Latino Value)
Feminism
(American Value)
Grooved Pegboard
(Dominant Hand)
- . 1 2 -.19 -.07
Grooved Pegboard
(Non-Dominant
Hand)
-.06 -.04 -.06
Partial r (74), *p < .05; **p < .01.
Table 82
Comparison o f NeSBHis-R Motor Domain Subtest bv Marin Variables.
Controlling for Gender. Age, and Total Years o f Education
Language Use/
Ethnic Loyalty Media
Ethnic Social
Relations
Grooved Pegboard - . 0 1 -.19 -.16
(Dominant Hand)
Grooved Pegboard -.06 -.26 * -.17
(Non-Dominant
Hand)
Partial r (74), *p < .05; **p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
T a b le 83
C o m p a riso n o f N e S B H is -R M o to r D o m ain S u b te st b v C u ltu ra l Id en tity S cales and
M a rin C lu sters. C o n tro llin g f o r G en d er. A ge, an d T o ta l Y e a rs o f E d u catio n
Cultural
Identity
Scales
English/
American
Cluster
Cultural
Identity
Scales
Spanish/Latino
Cluster
Cultural Identity
Scales
Spanish/Latino
(All Variables)
Cluster
Marin
Cluster
(Total)
Grooved Pegboard
(Dominant Hand)
- . 1 2 -.17 - . 1 0 - .1 2
Grooved Pegboard
(Non-Dominant
Hand)
-.17 -.09 - .1 1 -.18
Partial r (74), * jj < .05; * * p < .01.
A summary o f the findings associated with Research Question 5
demonstrates that the addition of the demographic controls (gender, age, and
education) significantly compromised the influence o f Culture on NeSBHis-R
Domains. Prior to the addition of demographic controls, all the NeSBHis-R
Domains were confounded by Culture. However, following the control for gender,
age, and education, only three NeSBHis-R Domains remained significant:
Nonverbal Reasoning, Language, and EF-AC. Among these NeSBHis-R
Domains, it is notable that the Language Domain o f Culture continued to show the
strongest relationship with these neuropsychological tests. Spanish proficiency
was most associated with NeSBHis-R subtests that require oral output (e.g.,
Stroop, PS-BNT, Verbal Fluency, and Digit Span) and English proficiency and use
most influenced nonverbal subtests (e.g., Ravens, Digit Symbol, and Color Trails).
The Behavior Domain of Culture showed less support, and the Values Domain was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
insignificant. Relative to the mulitidimensional and unidimensional models, it was
noted that the CIS continued to provide additional bicultural information relative
to the identified NeSBHis-R Domains.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
CHAPTER 4
DISCUSSION
In this chapter, the results o f the present study are summarized and
interpreted. Following a brief overview o f the entire study, the five central
research questions are discussed in three sections under the Interpretation o f
Results. Section one focuses on the NeSBHis-R as compared to three
demographic variables (Gender, Age, and Education), as well as Research
Question I and its accompanying hypothesis.
The second section focuses on (a) the operationalization of Objective
Language Categories; and (b) the NeSBHis-R as compared to the five Objective
Language Categories. The discussion in this section combines results from
Research Questions 2 and 3 (and accompanying hypotheses). An interpretation of
the findings which compared NeSBHis-R Domains to the Objective Language
Category variable are presented (Research Question 2) while integrating
subsequent findings from Gender, Age, and Education controls (Research
Question 3).
The third section focuses on (a) a reliability analysis o f culture measures
(CIS and Marin); and (b) the NeSBHis-R as compared to CIS and Marin
variables/clusters. The discussion in this section combines results from Research
Questions 4 and 5 (and accompanying hypotheses). An interpretation of
NeSBHis-R and Culture (CIS and Marin) comparisons are presented (Research
Question 4) while integrating subsequent findings from Gender, Age, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
Education controls (Research Question 5). Finally, theoretical and clinical
implications are discussed, as well as limitations of the study and directions for
future research.
Introduction
The main objective o f the current study was to further the process of
normative data collection among the Latino/a community and to provide improved
administration and interpretation procedures when evaluating neuropsychological
performance with a Spanish language test battery. Unfortunately, despite the rapid
growth o f psychometrics during the last two decades, appropriate tests and norms
for use with Latino/a clients is still very limited. Although some psychometric
batteries are currently available in Spanish (e.g., Wooodcock Munoz, Bateria-R;
Clinical Evaluation o f Language Fundamentals), only two Spanish language
neuropsychological batteries exist to date. The NeSBHis is one of these. In spite
of these limitations, the need for neuropsychological assessment of Hispanics in
both academic, clinical, and research settings continues to grow.
The goals o f the current study were threefold. First, to advance normative
data collection for use with Hispanics by presenting a more comprehensive
NeSBHis; hence, the NeSBHis-R. This goal was partially accomplished. The
current sample was not large enough to stratify the results by Age, Gender, and
Education; therefore, the current normative data charts should be interpreted with
caution. Nevertheless, the addition of two subtests and two culture measures, as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
well as the consideration of bilingual and monolingual differences, added a
significant contribution to the original NeSBHis.
Consequently, the second goal of the current study was to further
understand the role of bilingual and monolingual language skills on the NeSBHis-
R. Existing theories suggest that different levels of bilingualism can influence the
process of cognitive benefits and deficits (Threshold Theory: Ricciardelli, 1992);
therefore, the current study attempted to test this theory. This effort required the
operationalization of the different levels o f language which are notable among the
Hispanic community. Following an objective assessment of language, the current
sample yielded five distinct language categories which were confirmed when
correlated with subjective language ratings. The five Objective Language
Categories (Objective Language Category variable) showed some support for the
Threshold Theory. However, since four o f the five categories were assessed as
having “Strong” linguistic abilities, the comparisons between the identified
bilingual groups could not test the fiill scope of the Threshold Theory. In spite of
that, significant differences were noted between bilinguals, between bilinguals and
monolinguals, and between monolingual groups. Hence, while the Objective
Language Categories explained a significant amount of variance on all the
NeSBHis-R Domains, the most salient findings were associated with the
Nonverbal Reasoning, Language, Memory, and EF-AC Domains.
The third goal of this study was to further the understanding o f Culture as a
moderator variable. Two measures of culture were identified, the Marin and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
CIS, and they proved to be reliable as compared to the current sample. Next, the
theoretical and clinical implications o f Culture variables, as compared to the
NeSBHis-R Domains, were tested. While the literature differentiates between
multidimensional and unidimensional measures o f culture, these two models have
not been compared empirically. Therefore, the current study offers an initial
attempt to understand how these two culture models present in a clinical setting, in
this case, neuropsychological assessment. The current study offers support for two
distinct culture models in that only the multidimensional CIS measure revealed
bicultural findings. The unidimensional Marin model was less sensitive in its
ability to distinguish between the two cultures simultaneously. Nonetheless, both
the CIS and Marin showed similar significance when compared to NeSBHis-R
Domains. Specifically, among the Culture variables (e.g., Language, Behavior,
and Values), Language showed the strongest relationship with NeSBHis-R
subtests, Behavior followed, and Values showed the least significance. Overall,
Culture contributed to all NeSBHis-R Domains, but was most significantly
correlated with the Nonverbal Reasoning, Language, and EF-AC Domains.
In sum, the current study supports Culture and Language as important
moderator variables that should be considered in the neuropsychological
assessment of Hispanics. These findings were established in the context of
controlling for five Hispanic risk factors: (a) Occupational/Economic Stress, (b)
Parental Stress, (c) Family/Cultural Stress, (d) Marital Stress, and (e) Immigration
Stress. Hence, the focus o f Language (Objective and Subjective), Culture, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
Hispanic risk factors on neuropsychological performance offers a significant
contribution to the existing body o f knowledge. Important theoretical and clinical
implications are indicated, keeping in mind the limits of generalizability. For
instance, the current sample was most represented by first-generation, Mexican
individuals who were of middle to middle-lower social economic standing.
Interpretation of Results
NesBHis-R as Compared to Three Demographic
Variables fGender. Age, and Education) /Research Question O
Research Question 1 - Demographic Differences on NeSBHis-R Domain
Performance: To What Extent Will Demographic Differences Among the Current
Sample Influence their Performance on NeSBHis-R Subtests? This research
question asked whether there were demographic differences (Gender, Age, and
Education) on NeSBHis-R subtest scores and respective NeSBHis-R Domains
(Nonverbal Reasoning, Language, Executive Function/Attention Concentration,
Memory, Visual-Spatial; Motor). The interpretation of these results includes a
comparison between the original NeSBHis and the NeSBHis-R findings because
both batteries considered the potential confound of Age, Gender, and Education on
test performance. However, it is important to note that while the NeSBHis’ larger
sample (n=300) allowed for comparison of education and age by groups, this
approach was not possible with the smaller sample of the current NeSBHis-R
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
sample (N=78). Therefore, as mentioned in previous chapters, Age and Education
variables in the current study, are treated as continuous independent variables.
Nevertheless, it was still considered useful to compare the two batteries on the
identified demographic variables as this is the first attempt to provide additional
support for the NeSBHis as a valid measure o f neuropsychological functioning
with Hispanic populations. It was also important to provide a baseline for the most
notable demographic confounds to neuropsychological test performance so that
these findings could be considered in the interpretation o f Objective Language
Categories and Culture (CIS and Marin) as compared to NeSBHis-R performance..
Hypothesis. The first hypothesis was that Age, Gender, and Total Years of
Education would all be correlated with neuropsychological subtest scores. First, it
was predicted that there would be a significant difference between men and
women on neuropsychological test performance. However, upon review of the six
NeSBHis-R domains, minimal support was found for the current hypothesis as it
was noted that only the Reasoning, EF-AC, and Visual-Spatial Domains showed a
significant difference between men and women. The results favored men to
women by approximately five points on the Ravens subtest and by one point on
the Digit Span Forward and Rey-0 (Copy) subtests. While these gender
discrepancies were small, it is important to note that this result differs from the
original NeSBHis study in that significant gender differences on the Ravens
(Total) and Rey-O (Copy) subtests were not noted in that study. Another
discrepancy between the original and revised NeSBHis batteries was that the
with permission of the copyright owner. Further reproduction prohibited without permission.
148
original NeSBHis significantly favored men to women on the PS-BNT (Language
Domain) while the NeSBHis-R did not support this finding. However, a
qualitative review of the NeSBHis and NeSBHis-R showed that there was a
general tendency across neuropsychological domain performance which supported
higher scores for men than women. Therefore, the current findings suggest that
gender differences should still be considered by clinicians and researchers when
assessing for neuropsychological ability among the Latino/a population.
The second prediction associated with Hypothesis A was that education
would be positively correlated with neuropsychological test performance. In other
words, as total years of education increased so would neuropsychological test
performance. Results showed strong support for this prediction in that greater
years o f education improved performance on all NeSBHis-R domains with the
exception of one (Memory Domain) subtest: the Auditory Verbal Learning Test
(AVLT: all insignificant). This finding is largely consistent with the original
NeSBHis study. All NeSBHis subtests were significantly influenced by education
except the AVLT (short-term recall). Hence, while the AVLT subtest appears to
be less influenced by exposure to education when assessing for short-term learning
skills (working memory), clinicians should still consider how greater exposure to
academic settings likely improves the capacity for initial learning and long-term
memory retrieval. More importantly, however, higher years of education
consistently improved performance on all other subtests from both the NeSBHis
and NeSBHis-R, which shows strong support for Hypothesis A. In other words,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
there is strong support that both verbal (reading and reciting) and nonverbal
(writing) exposure, academically, has a positive influence on the development of
stronger neuropsychological skills.
The third prediction associated with Hypothesis A stated that age would be
negatively correlated with neuropsychological test performance. The expectation
was that neuropsychological test performance would decrease as age increased.
Statistical comparisons between age and neuropsychological test performance
yielded overall consistent results between the NeSBHis and the NeSBHis-R. For
example, both batteries showed significant inverse relationships with level o f
performance on nonverbal output subtests (e.g., Ravens Total, Digit Symbol, Color
Trails 1 and 2, Rey-0 Delay, Block Design, Grooved Pegboard D and ND). In
other words, as age increased, performance on the Nonverbal Reasoning, EF-AC,
Visual-Spatial, Motor, and Memory Domains decreased. It was also noted that the
EF-AC and Memory Domains showed some discrepancies. Both NeSBHis
batteries showed that age does not significantly influence verbal output EF-AC and
Memory subtest scores (e.g., Digit Span and AVLT subtests). In fact, the addition
of the Stroop subtest to the NeSBHis-R was also consistent with this finding.
Actually, the notion that age does not negatively affect language skills and related
domains (e.g., Memory and EF-AC) is further supported by the insignificant
relationship between age and the Language Domain of the NeSBHis-R. A similar
finding was noted on the NeSBHis-R’s Language Domain with the exception of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
the PS-BNT. This language subtest showed a significant relationship with age on
the NeSBHis but not on the NeSBHis-R.
Thus, the overall age-related findings suggest that fine motor, and
constructional, speed of processing skills show the most significant decline with
the aging process. Moreover, the aging process also reduces other skills (e.g.,
memory, attention, concentration, and executive functions) if the task requires
additional rapid, visual-motor agility. Actually, with reference to memory skills, it
was notable that while age did not significantly reduce performance on the Rey-O
Copy, it did significantly reduce the ability to recall this visually presented
information from memory. This finding was consistent on both the NeSBHis and
NeSBHis-R.
In sum, with reference to the first research question there is considerable
support for the NeSBHis as a valid measurement of neuropsychological
functioning among Hispanics as confirmed by overall consistent Age, Gender, and
Education findings on both the NeSBHis and the NeSBHis-R. A comparison of
the two batteries showed some discrepancies on the PS-BNT and the AVLT
subtests when compared to demographic variables. These discrepancies could
present secondary to the different sample sizes of the two batteries, but more
importantly, it speaks to the need to also be sensitive to education, age, and
gender effects with these subtests. Overall, higher education showed the greatest
influence over strong neuropsychological skills, the aging process decreased
nonverbal skills, and gender differences were generally insignificant.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
NeSBHis-R as Compared to Objective Language Categories
^Research Questions 2 and 3 ~ )
The interpretation o f Research Questions 2 and 3 are integrated in the
following discussion; therefore, the research questions are stated first along with
their respective objectives. Then, the hypotheses and findings for Research
Question 2 are discussed; the hypotheses associated with Research Question 3 will
be integrated in this discussion.
Research Question 2 - Influence of Objective Language Categories on
NeSBHis-R Domain Performance: To What Extent Will Distinct Objective
Language Skills Between Spanish-Enelish Bilinguals and Spanish Monolinguals
Explain NeSBHis-R Subtest Scores? This question had two objectives. Since the
original NeSBHis study was criticized for not addressing the bilingual versus
monolingual influence of their sample (Cervantes and Acosta, 1992), the current
study sought to (a) operationalize bilingual and monolingual variables (Hypothesis
1); and (b) assess the influence that the operational categories could have on
neuropsychological functioning (Hypotheses B through E).
Research Question 3 - Influence of Objective Language Categories on
NeSBHis-R Domain Performance While Controlling for Gender. Age, and Total
Years of Education: To What Extent Does the Objective Language Category
Variable Still Explain a Significant Amount o f the Variance on NeSBHis-R
Subtest Scores Following the Control of Gender. Age, and Total Years of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Education? The objective of this research question was to further explain findings
associated with research questions one and two. Since Research Question 1
showed that each NeSBHis-R Domain was influenced by one or more
demographic variables (Age, Gender, and Education), it was important to consider
how much of the variance associated with NeSBHis-R scores could still be
explained by the Objective Language Category after controlling for significant
demographic data. Thus, the hypothesis which was associated with Research
Question 3 predicted that the Objective Language Category variable would still
explain a significant amount o f variance associated with NeSBHis-R Domain
performance after controlling for gender, age, and total years o f education.
Research Question 3 will now be discussed in the context o f Research
Question 2.
Research Question 2, Hypothesis A. It was hypothesized that bilingual and
monolingual language proficiency could be operationalized to fit the five
Objective Language Categories that were identified upon review of the current
sample’s Woodcock-Johnson Language Proficiency Battery - Revised (WJLPB-R)
Standard Scores. When compared to six subjective language ratings, two by the
examiner and four by the examinee, the five Objective Language Categories were
confirmed and therefore support the following definitions:
1. Strong Balanced Bilingual: English and Spanish WJLPB-R Standard
Scores are above 90 and fall within the same standard deviation.
permission of the copyright owner. Further reproduction prohibited without permission.
153
2. Strong English Dominant Bilingual: English and Spanish WJLPB-R
Standard Scores are above 90 but English proficiency Standard Scores fall one
standard deviation above Spanish proficiency Standard Scores.
3. Strong Spanish Dominant Bilingual: Spanish and English WJLPB-R
Standard Scores are above 90 but Spanish proficiency Standard Scores fall one
standard deviation above the English proficiency Standard Scores.
4. Strong Spanish Monolingual: Spanish WJLPB-R Standard Scores are
above 90 and English WJLPB-R Standard Scores are below 79.
5. Weak Spanish Monolingual: Spanish WJLPB-R Standard Scores are
below 89 and English WJLPB-R Standard Scores are below 79.
Specifically, subjective language proficiency ratings by both the examiners
and examinees were consistent with the five Objective Language Category
definitions.
While bilingual research has received considerable attention among
pediatric populations (e.g., Bialystok, 1986, 1988, 1998 & 1999), less research is
available with adult populations, and no empirical support for adult or child
Latino/a bilingual and monolingual differences were found which are specific to
the study of neuropsychology. Therefore, the current effort to operationalize
bilingual and monolingual skills among an adult Latino/a sample offers a
considerable advancement to the study of language proficiency in
neuropsychological assessment. Other notable findings associated with the current
study’s comparison o f objective and subjective language proficiency involve
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
differences between examiner and examinee ratings, and differences between
subjective language measures.
With respect to examinee and examiner differences, both groups provided
consistent assessment of English language proficiency when compared to the
objective scores (WJLPB-R). The ability to subjectively identify strong Spanish
language proficiency, as compared to WJLPB-R Standard Scores, was equal
between groups; however, only examinees were able to subjectively rate the
distinction between strong and weak Spanish language proficiency. This finding
suggests that examiners were not able to subjectively distinguish between strong
and weak Spanish language proficiency of examinees. It is likely that the
examiners observed strong conversational Spanish language skills among the
examinees which inflated the examiner’s subjective language assessment of the
subjects. Two observations were noted from this finding. First, examiners need to
be sensitive to the fact that conversational skills are not necessarily representative
of more formal objective language proficiency assessment. Therefore, caution
should be exercised when comparing fluid conversational skills to more formal
and structured language skills that are assessed with objective measures. Second,
this finding also supports the practice o f requesting the examinee’s subjective
rating of their Spanish proficiency prior to engaging in formal (objective) language
assessment. In fact, the method that is chosen to obtain such information from the
examinee also proved to be an important issue.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
The current study applied two methods of subjective language assessment:
four 5-point Likert scales were developed for the current study (English and
Spanish, Examinee and Examiner) as well as two variables from the Cultural
Identity Scales (English and Spanish). The CIS is a valid and reliable culture
instrument. While all six variables showed statistical support for the Objective
Language Category variable, upon review of the post-hoc tests, only the CIS
identified significant differences between all five Objective Language Categories.
Hence, not only is it important to obtain subjective language assessment from
examinees, but it is also important to use a valid and reliable measure for such an
assessment. Although the administration of a neuropsychological battery can be
very time consuming, it is important to take the additional time to assess language
subjectively as well as objectively.
Since Hypothesis A addressed issues specific to the operationalization of
the Objective Language Categories, the remaining hypotheses (B through E) will
review the findings which compared the Objective Language Categories to the
NeSBHis-R Domains. The additional control for Gender, Age, and Education in
this model is integrated into the discussion.
Research Question 2, Hypothesis B. The second hypothesis for this
question was that the Objective Language Category variable would explain a
significant amount of variance on neuropsychological domain/subtest
performance. The results showed that all six NeSBHis-R Domains were
significantly influenced by Objective Language Categories. However, when a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
control for Gender, Age, and Education was added to the model (Research
Question 3), the Objective Language Category variable only explained a
significant amount o f variance on four of the six NeSBHis-R Domains:
Reasoning, Language, EF-AC, and Memory. A review and interpretation of these
findings will now be presented.
The Ravens subtest is considered to be a test of nonverbal intelligence
because it is related to Spearman’s g (Ponton et al., 1996) and because it does not
require language output (e.g., oral, reading, and written). However, objective
language skills associated with the Objective Language Category variable showed
a considerable effect on higher Ravens total scores. The Ravens subtest was also
significantly related to education (Research Question 1), so it can be inferred that
while language output (e.g., oral, reading, and written) is not required by this test,
the process of completing diagrammatic puzzles was facilitated with the current
sample, when higher-order complex language skills were present. Moreover, the
ability to formulate language in a more complex fashion appears to influence other
cognitive skills such as complex perceptual organizational reasoning, suggesting
that processing skills that are associated with oral language and reading (e.g.,
using letters to create words, using words to create sentences, etc.) likely improves
nonverbal tasks which rely on similar planning and organization skills (e.g.,
relying on details of a puzzle to complete a puzzle). One could also argue that
nonverbal tasks include a verbal component which is applied in a perceptual
organizational fashion. Hence, although nonverbal tasks such as the Ravens, may
permission of the copyright owner. Further reproduction prohibited without permission.
157
not include the oral component of language, other problem-solving skills are
enhanced by having access to reading and writing skills. One could say that
complex (higher order) processing skills are similar in language and perceptual
organizational (Ravens) abilities. Thus, while the Ravens does not include an oral
language component, it is likely that performance on this test is still influenced by
language proficiency skills. It is likely that repetition and development of verbal
and nonverbal skills in academic settings facilitate an individual’s growth in the
areas of cognitive reasoning and processing skills.
Moreover, when comparisons between the Ravens total score and the
Objective Language Categories included a control for Gender, Age, and
Education, the statistical correlations remained significant (Research Question 3).
Hence, caution should be taken when interpreting the Ravens subtest as a measure
of nonverbal intelligence because the skills required for language production also
benefit general reasoning abilities.
On a similar note, the two subtests which represent the Visual-Spatial
domain were significantly correlated with the Objective Language Category
variable. These subtests, “Block Design” and “Rey-O Copy,” also do not require
language (oral, reading, and written). Nevertheless, strong language proficiency
proved to be an asset relative to higher scores on these NeSBHis-R subtests. As
with the Nonverbal Reasoning Domain, one must consider how the added benefit
of linguistic development can manifest as advanced competence in multiple
cognitive domains. As with the Ravens subtest, it can be inferred that strong
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
language skills allow for general cognitive benefits, which in the case o f the
visual-spatial domain, present as stronger perceptual-organizational skills that
require higher-order planning and eye-hand coordination. However, based on the
findings following the Gender, Age, and Education controls (Research Question
3), one can conclude that while language may account for some improvement in
visual-spatial functioning, fine motor and perceptual-organizational skills are the
most salient contributors relative to improved skills Block Design and Rey-O
Copy performance. Neither the Block Design nor the Rey-O Copy subtests
remained significant following the control for demographic variables.
The influence of Objective Language Categories on the NeSBHis-R
Language Domain was very strong as it was noted that all but one subtest on this
domain (“Verbal Fluency, Fruits”) showed a significant effect. Moreover, the
same pattern was evidenced when the comparison of NeSBHis-R and Objective
Language Categories included a control for Gender, Age, and Education (Research
Question 3). Hence, on the PS-BNT and Verbal Fluency (FAS and Animals)
subtests, stronger language proficiency improves language production and
knowledge, supporting these subtests as measures of language abilities. The
“Verbal Fluency, Fruits” subtest, however, appears to be less influenced by
language development and more influenced by other variables such as culture
and/or environmental experiences. This possibility is explored in the following
section.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
Eight o f the nine EF-AC subtests showed significant effects as compared to
the Objective Language Category variable. The only test that was insignificant
was the “Digit Span Forward” subtest. Since the majority o f the EF-AC subtests
show improved scores when strong language proficiency is present, then the
interpretation o f these subtest scores (verbal and nonverbal) should not only
consider attention, concentration, and executive function skills, but also language
proficiency. On the other hand, the “Digit Span Forward” subtest is likely a
stronger measure o f attention, concentration, and working memory, and less
influenced by language proficiency.
However, when controls for Gender, Age, and Education were added to the
comparison o f the EF-AC Domain and the Objective Language Category variable
(Research Question 3), only four o f the nine EF-AC subtests were significant. The
Stroop (Color, Word, and Interference) and the Digit Span (Backward) subtests
remained significant, while the Digit Span (Total), Digit Symbol, and Color Trails
(1 and 2) were no longer significant. Digit Span (Forward) was still not
significant. Since the Stroop requires rapid naming skills these results suggest that
language proficiency, as assessed by the Objective Language Category variable,
should still be strongly considered in the interpretation of the Stroop. Although
one would expect less of an influence of objective language proficiency on the
Digit Span Backward subtest, because this subtest requires the manipulation of
numbers in working memory, it is important to also consider how an individual’s
access to complex language can assist with the cognitive flexibility required by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
this task. Overall, Objective Language Categories showed the strongest influence
on all the Stroop subtests and one Digit Span subtest (Backward). While one may
consider the potential influence of Objective Language Categories on Digit Span
(Total), Digit Symbol, and Color Trails (1 and 2), it is likely a secondary benefit as
compared to EF-AC skills. O f the nine EF-AC subtests, the Digit Forward
subtests is likely the least affected by Objective Language Categories.
A final note with respect to Hypothesis B includes mention o f the
NeSBHis-R Domains which showed little or no significant effect by the Objective
Language Category variable: Memory and M otor Domains. On the Memory
Domain, only the “AVLT Long-Term Recall” subtest was significant. The other
“AVLT” subtests and the “Rey-O Delay” subtest were insignificant, suggesting
these tasks are more likely assessing for memory skills rather than language skills.
Regarding the Motor Domain, a significant effect was noted when
compared to the Objective Language Categories; however, the post hoc tests were
not significant. Actually, when the control for Gender, Age, and Education was
added to these two models (Objective Language Categories by the Memory
Domain, and the Objective Language Categories by the Motor Domain) only the
AVLT Long-Term Recall subtest remained significant. It is possible that stronger
language proficiency skills facilitate long-term word retrieval. Also, it was noted
in the earlier discussion that the NeSBHis-R Language Domain showed a strong
relationship between word retrieval skills and the Objective Language Categories.
Overall, while these findings suggest that one should consider how strong
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
language proficiency could facilitate long-term retrieval of verbal information, the
general results show that verbal learning (over five trials), working memory, and
recognition on the “AVLT” subtests, and long-term retrieval o f a complex figure
were not influenced (negatively nor positively) by language proficiency skills. It is
also less likely that performance on the Grooved Pegboard subtest is confounded
by language proficiency skills.
In summary, while the results showed that all six NeSBHis-R Domains
were significantly influenced by Objective Language Categories (Research
Question 2), only four remained significant when a control for Gender, Age, and
Education was added to the model (Research Question 3). Specifically, the
Objective Language Category variable showed a consistent contribution on the
Reasoning, Language, EF-AC, and Memory Domains (before and after control
variables were added). Although strongest support for the influence of Objective
Language Categories on NeSBHis-R Domains was noted on subtests which
include a language com ponent, it was also notable that the Language Categories
offered a strong contribution to improved Ravens total scores. Upon review of the
findings before and after demographic controls, it is clear that subtests which
include an oral component should always consider language proficiency
(monolingual and bilingual) in the interpretation of the test results, and that a
similar approach should be taken when using the Ravens test. Finally, while other
nonverbal subtests showed variable results relative to the Objective Language
Categories (before and after control variables were added), consideration of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
possible language confounds should still be considered with higher-order
reasoning tasks. The current findings suggest that language serves as a cognitive
benefit when confronted with general problem-solving and reasoning tasks.
As a follow up to these findings, a discussion o f specific differences
between the five Objective Language Categories will now be presented
(Hypotheses C through E).
Research Question 2, Hypothesis C. The third hypothesis addresses the
post hoc tests which followed the results from Hypothesis B. Hence, Hypothesis
C predicted that “Strong” Balanced Bilinguals would perform significantly better
on neuro-psychological subtests than “Strong” English Dominant Bilinguals and
“Strong” Spanish Dominant Bilinguals. This prediction was based, in part, on the
“Threshold Theory” (Ricciardelli, 1992) and attempts to gain a further
understanding of how different levels of bilingualism, in this case different
bilingual categories, affect neuropsychological performance. Although the current
sample only confirmed “Strong” bilingual proficiency categories (“Strong”
language proficiency in both English and Spanish), it was expected that balanced
bilinguals would do better than unbalanced bilinguals on neuropsychological tests.
The findings, in general, were variable but showed some support for the
stated Hypothesis C. For example, Strong Balanced Bilinguals and Strong
Spanish Dominant Bilinguals did not perform significantly better than one another
on any NeSBHis-R subtest. Actually, keeping in mind that both of these bilingual
groups showed strong Spanish language proficiency, the fact that the Strong
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
Balanced Bilingual group had equally strong language proficiency skills in English
and Spanish, did not appear to improve or worsen their performance on NeSBHis-
R subtests. The contrary can also be stated, that the stronger Spanish Language
skills of the Strong Spanish Dominant Bilingual group did not appear to improve
or worsen performance on NeSBHis-R subtests when compared to the Strong
Balanced Bilingual group. Nevertheless, it is important to note that although
significant differences were not noted between these two groups, the Strong
Spanish Dominant Bilingual groups consistently showed higher scores than Strong
Balanced Bilinguals on the NeSBHis-R Language Domain which provided further
support for the language category distinctions.
On the other hand, the hypothesis was supported when significant
differences were noted between the Strong English Dominant Bilingual group and
the other bilingual groups (Strong Spanish Dominant/Strong Balanced) on the
Language, EF-AC, and Memory Domains. No differences between the bilingual
groups were noted on the Visual-Spatial, Reasoning, and Motor Domains.
Specifically, on the NeSBHis-R Language Domain, the Strong Balanced Bilingual
group and the Strong Spanish Dominant Bilingual group performed better than the
Strong English Dominant Bilingual group on the “PS-BNT” and the “Verbal
Fluency, Animals” subtests. This finding suggests that despite the fact that all
three bilingual groups showed strong Spanish language proficiency, stronger
English language skills did not facilitate expressive vocabulary skills to the extent
that such skills were available to bilinguals who have stronger or equal Spanish
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
language skills. This finding applies to the ability to name familiar objects (PS-
BNT), animals (Verbal Fluency, Animals), and words following letter prompts
(Verbal Fluency — FAS, only as compared to Strong Spanish Dominant Bilinguals)
but not to the naming o f fruits (Verbal Fluency, Fruits). It is likely that Strong
Balanced Bilinguals and Strong Spanish Dominant Bilinguals, as compared to
Strong English Dominant Bilinguals, develop stronger language processing skills
(e.g., planning and organization) and/or have greater access to verbal verbal
stimuli (familiar objects and animals) in their immediate environment. It is also
noteworthy that Objective Language Categories remained significant as compared
to the NeSBHis-R Language Domain (PS-BNT, Verbal Fluency, FAS and
Animals) following the control for Gender, Age, and Education. Since the Verbal
Fluency “Fruits” subtest remained insignificant, the issue o f environment or
culture as confounds is plausible as this subtest was significantly influenced by the
education variable (Research Question 1).
Strong Balanced Bilinguals and Strong Spanish Dominant Bilinguals also
outperformed Strong English Dominant Bilinguals on two of the “Stroop”
subtests: rapid naming o f “colors” and rapid naming of “words” (NeSBHis-R, EF-
AC subtest). In addition, when an executive function factor is added (Stroop
Interference), Strong Balanced Bilinguals performed significantly better than
Strong English Dominant Bilinguals. Also, though not significant, the Strong
Spanish Dominant Bilingual group demonstrated a better score on this test
(approximate 25 second advantage). On a further note, The Stroop (Color, Word,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
and Interference) remained significant following the control for Gender, Age, and
Education (Research Question 3).
Bilingual groups did not significantly differ on any other EF-AC subtest,
but it is notable that, although not statistically significant, Strong Balanced and
Spanish Dominant Bilingual groups had better scores on the “Digit Span” subtests
when compared to the English Dominant Bilingual group. This finding is
presented because it shows some support for the theme that was noted on the EF-
AC domain: Strong English Dominant Bilinguals showed better scores (not
significant) on the nonverbal (no oral output) EF-AC subtests (e.g., Digit Symbol
and Color Trails 1 and 2) when compared to the Strong Balanced and Spanish
Dominant Bilingual groups. In sum, while Strong Balanced and Strong Spanish
Dominant Bilinguals outperformed Strong English Dominant Bilinguals on EF-AC
subtests which require verbal “oral” output, the contrary was true with reference to
nonverbal (no oral output) EF-AC subtests. It is important to consider, therefore,
the influence of Spanish language skills on verbal output EF-AC subtests as these
tests show that they are not pure measures o f EF-AC. They are confounded by the
level o f bilingual proficiency, in this case equal or stronger Spanish language
skills. Similarly, nonverbal EF-AC subtests are confounded by bilingual skills,
and specifically, suggest that the “Digit Symbol” and “Color Trails” subtests may,
in part, require skills that are consistent with more developed English language
skills. In fact, stronger English language skills may not only be associated with
better fine motor (e.g., writing) and speed of processing skills, but also better
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
attention, concentration, and executive function skills. One must also consider,
however, the potential cultural confound of these subtests.
Another finding specific to Hypothesis C was noted on the NeSBHis-R
Memory Domain. While only one o f the five subtests was significant on this
domain when compared to the Objective Language Category variable (AVLT
Long-Term Recall), the Strong Spanish Dominant Bilingual group showed
consistently higher scores than Strong Balanced and Strong English Dominant
Bilingual groups on all the Memory Domain subtests. These results suggest that
Strong Spanish Dominant Bilinguals can retrieve more verbal information (Long
term) than Strong English Dominant Bilinguals (significant finding) and that
Strong Spanish Dominant Bilinguals also perform better than their bilingual
counterparts on both verbal and nonverbal recall (not significant, but notable).
Therefore, it is important to consider how stronger Spanish language skills among
Strong Bilingual groups may confound learning and memory subtests. Moreover,
the AVLT Long-Term Recall subtest remained significant following the control
for Gender, Age, and Education. Hence, the Spanish and English Dominant
Bilingual difference on Longer Term Recall of the AVLT warrants further
attention.
Hence, the current study offers some support for the Threshold Theory
(Ricciardelli, 1992). Among Strong Bilingual groups, Spanish verbal output is
better for individuals whose Spanish is equal or better than their English; however,
when English becomes the stronger language, Spanish verbal skills (including
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
long-term recall o f verbal information) become slightly compromised. On the
other hand, stronger English language skills, among strong bilinguals, appears to
allow for cognitive benefits on nonverbal tasks. While this was not a significant
finding, it was a notable trend. The Threshold Theory was also supported in that
small differences were noted between the three bilingual groups. Given that all
three bilingual groups demonstrated “Strong” (e.g., at least average) language
proficiency in both English and Spanish suggests that strong bilingualism is a
cognitive benefit. The extent to which culture confounds these findings, however,
also needs to be considered.
Research Question 2, Hypothesis D. The fourth hypothesis for this
question was that all bilingual categories (Strong Balanced and Strong
Unbalanced) would perform significantly better on neuropsychological subtests
than Spanish Monolinguals (Strong). Hypothesis D showed variable confirmation
on three of the six NeSBHis-R Domains: Language, EF-AC, and Visual-Spatial.
Also, although the Reasoning, Memory, and Motor Domains were not significant,
some trends were noted.
To begin, on the Language Domain, it was noted that Strong Spanish
Monolinguals performed significantly better than Strong English Dominant
Bilinguals on three of the four subtests. The “Verbal Fluency, Fruits” subtest was
not significant as both groups recited the same number o f items (14). Based on
other findings which pertain to the NeSBHis-R “Verbal-Fluency, Fruits” subtest,
cultural/environmental confounds should be considered. Despite the fact that both
with permission of the copyright owner. Further reproduction prohibited without permission.
168
groups qualified for Strong Spanish language skills, in this case, the Bilingual
component did not serve as a cognitive benefit. In fact, as with other findings
(Hypothesis C), it appears that stronger English language skills among Strong
Bilinguals, may limit specific Spanish language output skills (word retrieval and
knowledge). No other significant differences nor trends were noted in comparing
Strong Bilinguals to Strong Spanish Monolinguals, therefore, Hypothesis D was
not confirmed on the Language Domain. In fact, the opposite was revealed.
Strong Spanish Monolingual subjects did better than Strong English Dominant
Bilinguals. The current Language Domain findings are further supported after the
control o f Gender, Age, and Education (Research Question 3).
Conversely, on the EF-AC Domain, Hypothesis D was confirmed relative
to the “Color Trails 1” subtest. It was noted that all three bilingual groups
performed significantly better on this subtest when compared to the Strong
Spanish Monolingual group. This test relies on attention, concentration, fine
motor speed, and visual tracking skills, which suggest that these abilities are
enhanced when strong bilingual skills are present. However, this finding also must
be interpreted with caution as the Color Trails subtests (1 and 2) were no longer
significant following the control for Gender, Age, and Education. Partial support
for bilingual skills over monolingual skills was evidenced relative to Strong
Spanish Dominant Bilingual subjects and Strong Spanish Monolingual subjects.
The former group performed significantly better on two o f the “Digit Span”
subtests (Backward and Total); moreover, the Digit Span Backward subtest
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
remained significant before and after the control for demographic variables
(Research Question 3: Gender, Age, and Education). Also, while not significant,
it is notable that the Strong Balanced Bilingual group also showed higher scores
relative to Strong Spanish Monolinguals on these subtests. Strong Spanish
Monolinguals and Strong English Dominant Bilinguals performed the same on
these subtests. One must consider, given these findings, that the additional
bilingual ability, with Spanish as equal or stronger, can improve an individual’s
ability to maintain cognitive flexibility when manipulating numbers in working
memory.
Other EF-AC Domain results which were notable and which provide
further support for findings that were identified in the discussion of Hypothesis C
also deserve mention. For example, Strong English Dominant Bilinguals continue
to show stronger EF-AC skills when the task focuses on nonverbal skills. This
group performed significantly better than the Strong Spanish Monolingual group
on the “Color Trails 1 and 2” subtests. Also, the Strong Balanced Bilingual group
performed better than the Strong Spanish Monolingual group on the “Digit
Symbol” and “Color Trails l ”subtests. Yet, when the control for Gender, Age, and
Education was added, the “Color Trials” (1 and 2) and the Digit Symbol subtests
were no longer significant. Other findings which were not significant on the EF-
AC Domain, but which merit mention include the fact that all bilingual groups
performed better than the Strong Spanish Monolingual group on both the “Color
Trails 1 and 2” and “Digit Symbol” subtests. Also Strong Balanced and Spanish
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
Dominant Bilinguals performed better than Strong Spanish Monolinguals on all
Digit Span subtests. Hence, while there was some variability among the EF-AC
Domain in comparing strong bilingual and monolingual subtest performance, the
trends showed preference for bilingual skills over monolingual skills relative to
EF-AC performance. Also, stronger English language skills, among Bilinguals as
compared to Strong Spanish Monolingual language skills, suggest a cognitive
benefit for this Bilingual group on nonverbal EF-AC subtests.
On the Visual-Spatial Domain, there was also support for bilingual skills
over monolingual skills. For instance, Strong Spanish Dominant Bilinguals
performed significantly better than Strong Spanish Monolinguals on the Block
Design subtest; however, the Block Design subtest was no longer significant
following a control for Gender, Age, and Education. Also, while not significant, it
was also noted that all three bilingual groups performed better on this subtest when
compared to the Strong Spanish Monolingual group. Finally, the Memory, Motor,
and Nonverbal Reasoning Domains were not significant as compared to strong
bilingual and monolingual differences. However, it is important to note that there
was indeed a five to six point discrepancy between the bilingual groups and the
Strong Spanish Monolingual group which favored bilinguals on the “Ravens”
subtest. Also, both dominant and non-dominant fine motor skills on the Grooved
Pegboard showed preference for bilingual groups over the Strong Spanish
Monolingual group. Performance was comparable between bilinguals and Strong
Spanish Monolinguals on the Memory Domain and the “Rey- O Copy” subtest.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
In sum, strong bilinguals outperformed Strong Spanish Monolinguals on all
NeSBHis-R Domains, with the exception o f the Language and Memory Domains.
On the former, Strong Spanish Monolinguals demonstrated significantly higher
scores than bilinguals, and on the Memory Domain, no differences were noted.
The strongest support was noted on the following subtests because they were
significant before and after the control variables were added: PS-BNT; Verbal
Fluency (FAS), Verbal Fluency (Animals), and Block Design.
Research Question 2, Hypothesis E. The fifth hypothesis for this question
was that all “Strong” categories (bilingual and monolingual) would perform
significantly better on neuropsychological subtests than “Weak” Spanish
Monolinguals. This hypothesis was confirmed on all NeSBHis-R Domains except
the Memory Domain. On the Reasoning Domain, all “Strong” bilingual groups
and the “Strong” Spanish Monolingual group demonstrated significantly higher
scores on the Ravens subtest. Again, as noted in earlier discussion, it is notable
that the “Ravens” subtest showed considerable influence by language proficiency
and education level. This test is considered a test o f nonverbal intelligence, yet it
has repeatedly shown a strong language component. In fact, “Weak” Spanish
Monolinguals, as compared to either Objective Language Categories, showed up
to a 20-point discrepancy on this subtest. Also, comparisons between Ravens and
the Objective Language Category variable remained significant following the
control for Gender, Age, and Education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
On the Language Domain, some variability was noted among “Strong”
language groups as compared to the “Weak” language group. Significant findings
were noted when comparing Strong Spanish Dominant Bilinguals to Weak
Spanish Monolinguals on three of the four Language Domain subtests (all but the
AVLT, Fruits); the former showed consistently higher scores. Also, Strong
Balanced Bilinguals and Strong Spanish Monolinguals did significantly better than
Weak Spanish Monolinguals on the Verbal Fluency (FAS) subtest. These
Language Domain subtests also remained significant following the control of
Gender, Age, and Education. Furthermore, though not significant, it is important
to note that Weak Spanish Monolinguals showed consistently lower scores on the
Language Domain when compared to all “Strong” categories (Strong Spanish
Dominant Bilingual, Balanced Bilingual, and Spanish Monolingual).
On the EF-AC Domain, significant findings were noted on all subtests
except the “Stroop (Word and Interference)” and “Digit Span (Forward)” subtests.
Strong Spanish Monolinguals only performed significantly better than Weak
Spanish Monolinguals on one EF-AC subtest (“Stroop Word”); however, a
qualitative review of the results showed that the former generally obtained higher
scores. Among the strong bilingual groups, the English Dominant Bilingual group
significantly outperformed the Weak Spanish Monolingual group on all nonverbal
subtests (Digit Symbol and Color Trails 1 and 2). It was notable that Strong
Balanced and Strong Spanish Dominant Bilinguals also significantly outperformed
Weak Spanish Monolinguals on these subtests. However, these subtests did not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
remain significant following the control for Gender, Age, and Education. On
verbal output EF-AC subtests, both Strong Balanced and Spanish Dominant
Bilinguals scored better than Weak Spanish Monolinguals on the “Digit Span
Backward” but only the Strong Spanish Dominant Bilingual group outperformed
the Weak Spanish Monolingual group on the “Digit Span Total” subtest. Finally,
Strong Balanced Bilinguals, Strong Spanish Dominant Bilinguals, and Strong
Spanish Monolingual groups scored better than Weak Spanish Monolinguals on
the “Stroop Word” subtest. The Digit Span Backward and the Stroop subtests also
remained significant following the control o f Gender, Age, and Education. In
general, Strong Bilingual skills present as a cognitive benefit on the NeSBHis-R
EF-AC Domain. Cognitive benefits/deficits were much less salient between
Strong and Weak Spanish Monolingual groups.
No significant differences between strong and weak language categories
were noted on the Memory Domain; however, on the Visual-Spatial Domain,
Strong Spanish Dominant Bilinguals performed significantly better than Weak
Spanish Monolinguals on the “Block Design” subtest. In fact, on this subtest and
the “Grooved Pegboard (D and ND) subtest, both monolingual groups showed
lower scores than the bilingual groups, which suggests that bilingual skills may
improve visual-spatial constructional skills and fine-motor speed and precision.
Overall, relative to Hypothesis E, it was notable that Weak Spanish
Monolingual subjects did not outperform any “Strong” language group on any
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
NeSBHis-R subtest. This finding supports the notion that strong language
proficiency serves as a cognitive benefit across neuropsychological domains, with
the exception o f the Memory Domain.
A summary o f this section shows that the five Objective Language
Categories were confirmed, and a discussion of subsequent research questions
(two and three) demonstrated that the Reasoning, Language, Memory, and EF-AC
Domains showed the most consistent influence by the Objective Language
Category variable before and after controls were added to the models. This
general finding suggests that while there is not a difference between Strong
Balanced and Strong Spanish Dominant Bilinguals, Strong English Dominant
Bilinguals produced significantly lower NeSBHis-R scores when compared to
these two groups. Also, bilingual groups generally showed stronger NeSBHis-R
scores when compared to the Strong Spanish Monolingual group, with the
exception o f the Language Domain. When the Weak Spanish Monolingual group
was included in the comparisons, it was notable that they did not outperform any
other group on any NeSBHis-R subtest/domain. It was also notable that the Strong
English Dominant Bilingual group outperformed the other groups on EF-AC
subtests which included a nonverbal component. Finally, bilingual skills presented
as a cognitive benefit on the EF-AC Domain.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
NeSBHis-R as Compared to Culture
("Research Questions 4 and 51
In the following discussion, the interpretation o f Research Questions 4 and
5 are combined; therefore, the research questions and objectives are stated first.
Then, the hypotheses and findings for Research Question 4 are discussed. The
findings associated with Research Question 5 will be integrated into the discussion
o f Research Question 4
Research Question 4 - Influence o f Culture Variables on NeSBHis-R
Domain Performance: To What Extent Does Culture (e.g.. Languaee
Use/Exposure. Behavior. Values’ ) Explain a Significant Amount o f the Variance on
NeSBHis-R Subtest Scores? This question had two objectives. Since the original
NeSBHis study only included the Marin as a measure o f culture (Ponton et al.,
1996) and since these results are not yet available for review, the current study
sought to (a) confirm the reliability o f both the Marin and the CIS measures with
the current sample (Hypothesis A); and (b) assess the influence that the Marin and
CIS variables/clusters could have on neuropsychological functioning (Hypothesis
D).
Research Question 5 - Influence of Culture Variables on NeSBHis-R
Domain Performance While Controlling for Gender. Age, and Total Years of
Education: To What Extent Do Culture Variables Still Explain a Significant
Amount of the Variance on NeSBHis-R Subtest Scores Following the Control of
Gender. Age, and Total Years o f Education? The objective of this research
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
176
question was to further explain findings associated with Research Questions 1 and
4. Since, Research Question 1 showed that each NeSBHis-R Domain was
influenced by one or more demographic variables (Age, Gender, and Education), it
was important to consider if there was still a statistically significant correlation
between NeSBHis-R performance and Culture (CIS and Marin) following a
control for significant demographic data. Hence, the hypothesis which was
associated with Research Question 5 predicted that culture (CIS and Marin) would
still be significantly correlated with NeSBHis-R domain performance following
the control for Gender, Age, and Total Years of Education.
Research Question 4, Hypothesis A. It was hypothesized that the CIS and
Marin would yield the same constructs with the current sample. Following
reliability analyses of each CIS variable, it was found that only one variable
produced a weak reliability coefficient: Latino Activism (.5702); however, when
this four-item variable was collapsed, the reliability showed marked improvement
(CIS Items 24 and 25: .7010; CIS Items 28 and 29: .7363). Hence, caution was
taken in the interpretation o f the CIS Latino Activism variable. The Marin, on the
other hand, demonstrated consistently strong reliability coefficients across the
three variables that make up this measure. Overall, it was determined that both the
CIS and the Marin were reliable measures as compared to the current sample.
Next, the remaining hypotheses associated with Research Question 4 are
presented (Hypotheses B, C, and D). Thereafter, the discussion integrates results
from these three hypotheses as well as the findings from research question five.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
Research Question 4, Hypothesis B. It was hypothesized that the CIS and
Marin measures would show similar results on comparable variables. For
example, the CIS Language Domain would show similar results when compared to
the Marin Language Use and Media Domains. Similarly, it was predicted that the
CIS Behavior Domain would show similar results when compared to the Marin
Ethnic Social Relations variable.
Research Question 4, Hypothesis C. It was hypothesized that the language
specific culture variables would yield the strongest correlations on NeSBHis-R
subtests (CIS Language, Marin Language Use/Ethnic Loyalty, Marin Media), that
behavior variables would follow (CIS Behavior, Marin Ethnic Social Relations),
and that the values domain from the CIS would show the least significance when
compared to performance on neuropsychological tests.
Research Question 4, Hypothesis D. It was hypothesized that the CIS
multidimensional model (separate scores for English/American and Spanish/Latino
variables) would be a more sensitive model (e.g., identify the bicultural elements)
than the Marin’s unidimensional model (one score: high or low acculturation)
when compared to performance on the NeSBHis-R Domains.
Hence, the following discussion o f hypotheses B, C, and D, will be
presented by NeSBHis-R Domains. The first domain of focus is Nonverbal
Reasoning. The results showed support for Research Question 4, Hypothesis B
when compared to the NeSBHis-R Reasoning Domain. For example, all CIS
(Language, Behavior, and Values) and Marin (Language and Behavior) variables
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
and clusters (Language and Behavior) showed statistically significant correlations
as compared to Ravens total scores, suggesting that both the CIS and the Marin
were equally sensitive to the measurement of Language and Behavior (Research
Question 4, Hypothesis B). However, when the control for Gender, Age, and
Education was added, some differences were noted (Research Question 5). For
example, while the Language variables remained significant on both the CIS and
the Marin, the CIS Behavior and Values variables were now insignificant. Hence,
the consistently positive correlations between the “Culture” Language variable and
Ravens subtest performance suggests that subjective language (e.g., proficiency
and use/exposure) significantly improves Ravens total scores, and that the CIS and
Marin were not consistent in their assessment o f Behavior. Hence, Research
Question 4, Hypothesis C was also confirmed because, following the control
(Research Question 5), Language findings were most consistent, Behavior was less
consistent, and Values showed the least significance. Research Question 4,
Hypothesis D was partially supported. Prior to the control, it was noted that CIS
Spanish/Latino variables demonstrated a significant relationship with Ravens,
while the Marin did not. Even after the control, it was noted that Spanish
Language Preference was still significant on the CIS whereas Spanish/Latino
influence was not noted on the Marin.
Overall, while the positive correlation between Spanish/Latino culture and
Ravens should be considered, the most important consideration focused on how
the English/ American culture variables showed the most consistent and strongest
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
positive relationship with higher Ravens total scores. These findings not only
provide support for the measure o f language on the CIS and the Marin, but also a
potential benefit of the multidimensional model. Most importantly, these findings
suggest that the Ravens subtest may be confounded by culture. Additionally, it is
likely that the Ravens taps skills that are most consistent with the English language
(oral, written, reading, and social exposure) and less associated with Spanish
language skills, regardless o f a preference for the Spanish language. There is not a
clear explanation of the way the specific linguistic differences between English
and Spanish contribute to Ravens performance. However, it can be inferred that
for Hispanics, exposure, use, and proficiency o f the English language develops
higher-order reasoning abilities which complement the skills needed to perform
successfully on the Ravens test.
The NeSBHis-R Language Domain will now be of focus. Overall, it was
noted that Spanish/Latino culture variables were most consistently correlated with
strong performance on the NeSBHis-R Language Domain. As one would expect
with the Spanish language assessment, English/American culture variables were
not significantly correlated with the NeSBHis-R Language Domain. These
findings were supported before and after the control variables were added to the
model. Specifically, Hypothesis B (Research Question 4) showed similar findings
before and after the control factor was applied. Both the CIS and the Marin
showed most consistency in the identification of stronger Spanish language
proficiency/use and NeSBHis-R Language subtests. The CIS, however, showed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180
greater sensitivity as all NeSBHis-R subtests were significantly correlated with the
CIS Language Domain and only the PS-BNT and Verbal Fluency “Animals”
subtests were significantly correlated with the Marin. Moreover, only the CIS
produced significant correlations with the Behavior Domain as compared to all
four NeSBHis-R Language subtests. The Marin did not produce any significance
secondary to the Behavior variable (Ethnic Social Relations). These findings were
supported before and after the control variables were added. Finally, with respect
to Hypothesis D (Research Question 4), there was minimal statistical support. For
instance, both the Multidimensional and Unidimensional models showed the same
results: Significant correlations were only noted between Spanish/Latino variables
and performance on the NeSBHis-R Language Domain. While this finding was
consistent following the inclusion of the control variables, differences were noted
between the CIS and the Marin clusters when the control variables were added. In
summary, it was noted that expected culture variables improved performance on
the NeSBHis-R Language domain (Spanish Language and Latino Behavior).
Focus will now be placed on the NeSBHis-R EF-AC Domain. Overall,
there was considerable support for all the hypotheses associated with Research
Question 4 (Hypotheses B, C, and D). Both the CIS and the Marin “Language”
variables showed that Digit Symbol was positively correlated with English
language (proficiency and use) and American Behavior. However, both culture
measures continued to support the Language findings after the control factor was
added (e.g., Gender, Age, and Education), but only the CIS Behavior variable
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
continued to show a significant correlation with the Digit Symbol subtest.
Therefore, Hypothesis B was only partially confirmed in that the Language
variables were consistent on both culture measures, but the Behavior findings were
not.
Hypothesis C was confirmed. The “Language” variables (CIS and Marin)
showed the most consistent correlations with NeSBHis-R EF-AC subtests before
and after the control variables were added. The Behavior variable was also
significantly correlated with EF-AC subtests; however less consistently, as it
showed greater discrepancies following the addition of the control variables.
Finally, the “Values” variable was the least consistent when correlated with the
EF-AC Domain. Although Values showed a considerable relationship to EF-AC
subtests prior to the addition o f control variables, none of these correlations
remained significant when controls were considered in the model.
Hypothesis D showed considerable support when applied to the EF-AC
Domain. For instance, before and after the control for Gender, Age, and
Education, the CIS produced significant correlations on both Spanish/Latino and
English/American variables, but the Marin only identified positive correlations
with English Language and Behavior.
In summary, relative to the EF-AC Domain, there was a great difference
between the number o f correlations which were significant and positive prior to
adding the control variables. Nevertheless, it is important to note that the
measurement of English Language as a culture variable was positively correlated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
with both the Digit Symbol and Color Trails 1 subtests (EF-AC) before and after
control variables were considered. Color Trails 2 was only significant prior to the
addition o f control variables. Spanish Language was positively correlated with all
EF-AC subtests prior to the control for demographics, and with all the Stroop
subtests and two of the Digit Span subtests (Total and Forward) when the control
variables were added. The Behavior variable (“Latino Activism”) was most
consistently correlated, positively, with the Digit Span subtests; however, this
variable had also shown a weak reliability coefficient when compared to the
current sample. Therefore, this finding should be interpreted with caution. While
the influence of Spanish language on Spanish language neuropsychological tests is
less surprising (Stroop and Digit Span), it is still important to highlight the
possibility that these EF-AC subtests are not just measuring attention,
concentration, and working memory skills, but also language proficiency. In
addition, it is important to consider how the nonverbal EF-AC subtests (e.g., Digit
Symbol and Color Trails) are likely confounded by American culture and English
language exposure/use/proficiency. It is conceivable that these subtests are
culturally biased in favor of American/English experiences. Finally, it is also
important to consider how the Multidimensional model was able to identify a
bicultural influence on EF-AC subtests.
Results from the NeSBHis-R Memory Domain will now be interpreted. In
general, both the A VET and Rey-O subtests were not confounded by culture (CIS
and Marin) and therefore Hypotheses B through D (Research Question 4) were not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
confirmed. Although the CIS Spanish Language Proficiency variable was
positively correlated with the AVLT Long-Term Recall subtest, providing some
support for Hypotheses C and D, it was a weak relationship and it was no longer
significant following the addition of control variables. While it is important to
consider the potential influence of Spanish language proficiency on the retrieval of
verbal information from memory, the current study suggests that learning and
memory are the strongest predictors of performance on these subtests.
The NeSBHis-R Visual-Spatial and Motor Domains showed a similar
pattern as the Memory Domain; therefore, the discussion of these last two
NeSBHis-R Domains will be presented conjointly. Although both the Visual-
Spatial and Motor Domains were considerably confounded by culture prior to the
addition of controls, none o f these correlations remained significant following the
inclusion of control variables (Gender, Age, and Education). Therefore,
Hypotheses B through D (Research Question 4) were not confirmed.
Nevertheless, the findings prior to the addition of controls merit mention.
For example, the three hypotheses were generally confirmed as compared to the
Visual-Spatial Domain and the Motor Domain. Three o f the subtests associated
with these domains were significantly correlated with English proficiency/
use/exposure (Block Design; Grooved Pegboard D and ND). The Rey-O Copy
subtest did not show particular preference for English language skills. Also, it was
noted that the CIS identified significant and positive correlations with both
English/American and Spanish/Latino variables on both the Visual-Spatial and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
Motor Domains, while the Marin did not. Although, the preference was consistent
for English/American on the above mentioned subtests, it is noteworthy that the
multidimensional model demonstrated greater sensitivity to bicultural influences
than did the unidimensional model.
A summary o f this final section shows that while the Culture variables
showed a significant relationship with all NeSBHis-R Domains, these confounds
were considerably reduced when other demographic variables were taken into
account (e.g., Gender, Age, and Education). Nevertheless, while the influence o f
Culture (Language, Behavior, and Values), must be considered with most
NeSBHis-R subtests, it is evident that greatest attention should be focused on the
Nonverbal Reasoning, Language, and EF-AC Domains. These NeSBHis-R
Domains, when compared to Culture among demographic controls, provided
support for all three hypotheses. First, variables which were associated with the
Cultural assessment of Language demonstrated the strongest relationship with the
identified NeSBHis-R Domains, Behavior followed, and Values were largely
insignificant (Hypothesis C). Second, although there was general consensus
between the two culture measures when identifying the influence of Language and
Behavior on NeSBHis-R subtests (Hypothesis B), the multidimensional CIS model
still identified a bicultural contribution to NeSBHis-R performance where the
unidimensional Marin model did not. Specifically, it was noted that
English/American variables were significantly correlated with the Ravens, Digit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185
Symbol, and Color Trails, whereas Spanish/Latino variables were most significant
when compared to the Stroop, Verbal Fluency, PS-BNT, and Digit Span subtests.
Theoretical Implications
The present study offered implications for both existing theories of
bilingualism and culture, and for the study of neuropsychological assessment of
Hispanics. Researchers agree that multilingual skills contribute in some way,
positively or negatively, to cognitive functioning (Bialystok, 1986, 1988,
1998,1999; Cervantes & Acosta, 1992; Diaz, 1985; Garcia, 1986; Geisinger, 1992;
Pialorsi, 1981; Ricciardelli, 1992). In fact, scholars have identified important
theories which try to explain the bilingual experience, cognitively. Some of these
include the Threshold Theory (Ricciardelli, 1992) and the First Language
Interference Theory (Pialorsi, 1981). The current study provides some support for
the Threshold Theory in a Hispanic sample.
While the current study yielded mostly “Strong” levels o f language
proficiency among the sample, significant differences were still identified between
the three bilingual categories. For example, although differences were not noted
between the Balanced Bilingual and Spanish Dominant Bilingual groups, these
groups did differ significantly when compared to English Dominant Bilinguals.
Indeed, it is important to note that the dependent variable was a Spanish language
neuropsychological battery. However, it is also notable that all bilingual groups
were identified as having at least an average level o f Spanish language proficiency.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
186
It is also important to note that the fact that less differences were identified
between bilingual groups in the current study (at least average proficiency in both
languages) supports the concept of the Threshold Theory, which states that
individuals need to attain a high degree of bilingualism in order to attain overall
linguistic advantages. It is also likely that greater neuropsychological differences
among bilinguals could be identified if the sample represented both strong and
weak levels of language proficiency across language categories.
The current study also provided considerable support for the use of culture
as a moderator variable. The importance of including measures o f culture in
clinical and research settings has been expressed repeatedly in the literature (Dana,
1993 & 1994); however, the empirical outcome of this approach is still very
limited, especially in the realm o f neuropsychological assessment of Hispanics.
Not only does this study offer a comprehensive analysis of cultural identity, but it
is also a first attempt to control for Hispanic stress factors within the larger
analysis of Language, Culture, and Neuropsychological skills. The current study
confirmed the importance of looking at several levels o f culture (Language,
Behavior, and Values) within the bicultural experience. Also, a comparison of two
theoretical and empirical models o f culture demonstrated strong support for (a)
Language and Behavior and “Culture” variables, and (b) for the multidomain,
multidimensional measurement o f culture.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
Clinical Implications
This study has implications for the neuropsychological assessment of
Hispanics. The normative data that was compiled from this study did not yield cell
sizes that were large enough for clinical application; therefore, this data should not
be used for statistical interpretation o f individual test scores in clinical settings.
However, all six NeSBHis-R Domains, as studied with the current sample, were
significantly influenced by both objective language skills, subjective language
proficiency, and culture. In fact, the results controlled for five Hispanic risk
factors: Occupational/Economic, Parental, Family/Cultural Conflict, and Marital,
Immigration.
The overall results have salient clinical implications which will now be
presented by NeSBHis-R Domain. Focus is placed on outlining the language and
culture findings which are deemed important to the interpretation phase of
neuropsychological assessment among Hispanics.
The NeSBHis-R Nonverbal Reasoning Domain is comprised o f one
subtest: the Ravens Progressive Matrices, a test of both visuoperception and
abstract reasoning (Lezak, 1983). The clinical implications for use o f the Ravens
with a Hispanic client is that as total years of education increase, performance on
this subtest will improve. This finding supports other research (Lezak, 1983).
Age and gender are less influential though men did show a 5-point advantage to
women. If one considers the strong influence o f education on this subtest, it is not
surprising that language skills also improve performance on the Ravens. For
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
instance, “Strong Bilingual” groups will likely yield higher total scores when
compared to both “Strong and Weak Spanish Monolingual” groups, and “Weak
Spanish Monolinguals” (e.g., below average level of skill) are the most sensitive to
lower total Ravens scores. Moreover, this abstract reasoning task, which does not
require verbal output, is also vulnerable to cultural factors. The current sample
endorsed a preference for speaking Spanish, advanced English Proficiency still
yielded improved scores on this test. These results suggest that language
processing skills are similar to the abstract reasoning skills that are required for
successful performance on the Ravens. Also, English/American language and
cultural exposure potentially contribute to the development o f complex reasoning
skills.
The NeSBHis-R Language Domain is comprised of two tests, the PS-BNT,
and three Verbal Fluency subtests (FAS, Animals, and Fruits). The current
sample’s performance on these subtests suggests that while age and gender do not
contribute to improved neuropsychological language skills, other demographic
variables (e.g., education; culture— only Spanish/Latino Language and Behavior
variables) and linguistic skills (e.g., distinct levels of bilingual and monolingual
language skills) do contribute to significantly improved NeSBHis-R subtest scores.
Specifically, on the Verbal Fluency subtests, clinicians should consider that while
the “FAS” and “Animals” subtests are most sensitive to bilingual and monolingual
language skills, the “Fruits” subtest may be more influenced by cultural and
educational experiences. Both the “PS-BNT” and the “Verbal Fluency” subtests
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
were strongly influenced by Spanish/Latino language and cultural exposure, which
may explain, in part, the discrepant abilities between bilingual groups on these
subtests. While all bilingual groups were assessed as having at least an average
level o f language skills in both Spanish and English, the English Dominant
Bilingual group was still outperformed by the Strong Spanish Dominant and
Strong Balanced Bilingual groups. This finding implies that the level of language
proficiency appears to influence word retrieval skills. It is also important to
consider the potential contribution o f cultural exposure in the interpretation of
bilingual discrepancies. Finally, the consistently lower scores o f the Weak
Spanish Monolingual group suggests that education and language development
may be more influential than culture in the interpretation of language knowledge
and retrieval.
The NeSBHis-R EF-AC Domain is comprised of four subtests: Stroop,
Digit Span, Digit Symbol, and Color Trails. The current study showed that while
gender had a minimal impact on EF-AC test performance, clinicians should be
most aware o f the education and age related confounds. That said, there is an
additional consideration for bilingual groups. Strong English Dominant Bilinguals
still show a disadvantage as compared to Strong Dominant Bilinguals and Strong
Balanced Bilinguals, particularly on the Stroop and Digit Span subtests. On the
other hand, Strong English Dominant Bilinguals are at an advantage on the
nonverbal EF-AC subtests (e.g., Digit Symbol and Color Trails) when compared to
other bilingual groups. Nonverbal EF-AC subtests also showed preference to the
.j-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190
bilingual groups over the monolingual groups. This finding is consistent with
reports from studies which assess metalinguistic skills in child samples. It has
been shown what bilingual children exhibit stronger executive functioning skill
(fluid ability and cognitive complexity) than monolingual children (Bialystok,
1999). Clinicians should consider the possible culture confound o f the EF-AC
subtests as the current study showed that Spanish/Latino culture variables were
associated with higher Stroop and Digit Span Scores, and that English/American
culture variables were influenced by higher Digit Symbol and Color Trails scores.
The NeSBHis-R Memory Domain is comprised of two subtests: AVLT
and Rey-O Delay. Of all the NeSBHis-R Domains, this one showed the least
influence of culture and language on improved test performance. Education
showed a small contribution, gender was insignificant, and age showed a weak to
moderate correlation. Clinicians should note that there were no differences
between the bilingual and monolingual groups nor the strong and weak language
groups. However, Strong Spanish Dominant Bilinguals did show a slight
advantage over Strong Balanced Bilinguals and Strong English Dominant
Bilinguals on the AVLT Long-Term Recall subtests. This finding was further
supported by the small influence of culture on the AVLT Long-Term Recall
subtest when Spanish Proficiency was high. Suffice it to say that the AVLT and
Rey-O appear to be very strong measures of learning and memory because
language and culture are generally not associated with improved performance on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191
these subtests. However, Spanish language proficiency/skill may potentially
interact with the retrieval o f long-term verbal information.
The NeSBHis-R Visual-Spatial Domain is comprised o f two subtests:
Block Design and Rey-O Copy. While both subtests showed a moderate
correlation with education, only the Block Design subtest was significant as
compared to age (negative and moderate correlation). Also, the Rey-O Copy
subtest showed a one point advantage to women. Relative to bilingual and
monolingual groups, there is some support for bilingual groups over monolingual
groups on the Block Design subtests; however, it is more likely that education and
age are the strongest confounds o f successful performance on this subtest.
Similarly, while the English/American culture variables showed the strongest
influence on visual-spatial tasks (particularly the constructional, abstract reasoning
skills of the Block Design subtest), other demographic variables showed stronger
and more consistent support for improved scores on the Visual-Spatial NeSBHis-R
Domain subtests.
The NeSBHis-R Motor Domain is comprised of one subtest: Grooved
Pegboard (Dominant and Non-Dominant Hands). Compared to the current
sample, it was notable that this subtest showed weak to moderate correlations with
education and age, and there were no gender differences. As with the Visual-
Spatial Domain, these demographic influences showed greater and more consistent
support than did language and culture findings. Nevertheless, clinicians should
note that a small trend showed that strong bilingual groups outperformed the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
Strong Spanish Monolingual group on this subtest. Also, although both
Spanish/Latino and English/American culture variables showed influence on
improved Grooved Pegboard scores, the latter were most influential.
In sum, the current findings provide extensive support for the use o f at least
three moderator variables in both clinical and research settings which attempt
neuropsychological assessment of Hispanics. These are: (a) Objective Language
Measures which identify bilingual and monolingual levels o f skill, (b) Hispanic
Stress Inventory, and (c) Culture Measures which consider the bicultural
experience.
Limitations o f Study
It is important to note a number o f limitations of this study. First,
generalizability is limited by several factors. Due to logistical and financial
constraints, it was not possible for the current study to achieve a large sample size,
therefore limiting attempts to include equal numbers among diverse Latino/a
ethnic groups, generation levels, bilingual category levels, and social economic
groups. The small sample size also reduced normative cell sizes (e.g. education,
age, gender: see Appendix D). These limitations constrain the generalizability of
the current findings to other groups.
For example, the current sample is most representative o f the Mexican
ethnic group, which limits the possibility o f generalizing to the Hispanic group as
a whole. While one could argue that the current sample is a strong representation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
o f the Mexican ethnic group, there is a risk of making broad and inaccurate
generalizations about this subgroup. The majority o f the sample was also first
generation, which limits the generalizability to other generation groups.
A third demographic variable that merits attention is social economic status
(SES). The current sample most represented the middle to middle-lower social
economic group, limiting our understanding o f how both the lower and higher ends
of SES groups would present with this research model. The literature makes
frequent mention of the potential confound o f SES in clinical and research
interpretation (Dana, 1993), yet it is noted that this variable receives little attention
in neuropsychological assessment research. So while the current study reported
the SES data, it is limited in that this information was not included in the statistical
analyses. It is possible that SES groups could differ on several levels, including
culture and language issues.
Moreover, while there was a diverse age group represented in this study,
the smaller sample size as a whole limits the possibility of generalizing about any
one age group. In fact, because the current sample was relatively small, the
normative data of this sample cannot be applied toward statistical interpretation of
individual neuropsychological test scores in clinical settings. It is also likely that
the current findings could be altered if this model were applied to a larger sample.
It is equally important to consider the limitation of the bilingual and
monolingual groups in this study. While it was clinically and empirically useful to
consider how “Strong” levels o f language proficiency (e.g., at least average level
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
of proficiency) present across language categories, it limits the possibility of
generalizing to all bilingual and monolingual groups (e.g., “Strong” and “Weak”
levels). The current study does not consider “Weak” levels of bilingualism (e.g.,
below average and well below average); therefore, generalizing beyond the current
sample should be exercised with caution due to the sampling bias.
Internal and external validity was compromised with the identified
measures in that the CIS is an instrument that was developed with an adolescent
sample. Hence, the current study offered a first attempt to apply this measure to an
adult sample. Finally, given that the current study serves as a normative data
collection, it was important to consider the influence of stress, Hispanic stress, to
the findings. While there is a large benefit in stating that the current study
controlled for Hispanic stress (e.g., NeSBHis-R scores are less influenced by
external factors), the insignificant levels of stress could also limit the degree to
which current subjects endorsed culture-related stress items on the CIS and Marin.
Future Directions
Future research on the NeSBHis-R should continue to look at the influence
of Objective Language Categories as well as the Bicultural experience on subtest
performance. The current study suggests that both Language and Culture offer a
significant contribution to performance on neuropsychological subtests when
controlling for Hispanic stress factors.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
Future research should focus on obtaining larger sample sizes and a
balance of ethnic groups. Either one Hispanic ethnic group should be of focus
(e.g., Puerto Rican) or the number o f groups considered should be limited, to
ensure equal numbers between groups (e.g., Mexican, Cuban, and Puerto Rican).
Similarly, efforts should be focused on obtaining a balance among other
demographic characteristics such as SES, generation levels, and education
location, to name a few. These demographic variables should be considered in the
statistical analyses.
Greater diversity among bilingual groups is also suggested. Future
research should include both “Strong” and “Weak” levels o f bilingual proficiency
across all categories (Balanced Bilinguals, English Dominant Bilinguals, and
Spanish Dominant Bilinguals). While the current study offered some support for
the Threshold Theory, a larger number of language categories would provide a
more comprehensive assessment o f this theory and would also allow for greater
generalizability o f the findings. In an effort to understand which language
(Spanish versus English) serves as a benefit or deficit to cognitive functioning
among bilingual groups, it would also be useful to add an English language
neuropsychological test battery which would be administered to the bilingual
groups.
Given the complexity of the bilingual experience (e.g., levels and
categories) future research should continue to focus on (a) the operalization of
bilingualism and, (b) methods o f statistically analyzing the relationship between
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196
bilingual skills and neuropsychological abilities. The current study conceptualized
bilingualism as a categorical variable (Balanced Bilingual; Spanish Dominant
Bilingual; English Dominant Bilingual) and then considered the amount of
variance that each category explained relative to neuropsychological domains
(ANOVA and ANCOVA statistical analyses). While it will be important continue
testing this model empirically, other models should also be considered.
For instance, level or “degree” o f bilingualism could be operationalized as
an absolute value (e.g., English Proficiency Sum of Standard Scores subtracted
from Spanish Proficiency Sum o f Standard Scores). Higher values would
represent a larger discrepancy between English and Spanish language skills (e.g.,
low degree of bilingualism), and lower values would represent a small discrepancy
between English and Spanish language skills (e.g., High degree of bilingualism).
Comparisons between “degree of bilingualism” and performance on
neuropsychological domains/subtests could be analyzed using correlational
analyses (e.g., See Appendix H). A summary o f Appendix H shows that
performance on four of the six NeSBHis-R Domains (Nonverbal Reasoning; EF-
AC; Visual-Spatial; Motor) improved as degree o f bilingualism increased (“High
degree o f bilingualism”).
Overall, bilingualism shows a strong influence on neuropsychological
performance; therefore, it will be important to further investigate both the
operalization and statistical models used to analyze these variables. Also, future
research with adult Latino/a samples should continue to test well developed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197
theories from child/developmental literature, such as the study of metalinguistic
skills (e.g., crystallized versus fluid ability).
The current study also offered a contribution to the understanding of
culture and neuropsychological skills. However, culture is a complex experience
which may not be fully captured by a Likert scale questionnaire. Therefore, a
qualitative approach to understanding culture and cognitive functioning is
recommended. At minimum, however, future research and clinical interventions
with Hispanics should include three moderator variables: (a) Objective Language
Proficiency, (b) Bicultural Identity, and (c) Hispanic Stress.
In addition, future research should include a larger sample in which both
culture and language could be analyzed conjointly with the NeSBHis-R. Although
the current study suggests that objective and subjective language proficiency and
exposure have the strongest influence on neuropsychological test performance, it
would still be useful to determine how the two variables (language and culture)
interact with the NeSBHis-R.
While there is a great need for the development of normative data in the
Latino/a communities, there is a greater need to apply the existing NeSBHis data
to clinical populations. These populations are not limited to brain injured groups,
but also learning disabled populations, the medically ill (e.g., HIV and cancer), and
psychiatric patients. Finally, there is a need to identify a sample which endorses
clinically significant levels of stress so that the neurocognitive contribution of
stress among Hispanics could be further understood.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198
This comprehensive investigation of both language an culture within a
neuropsychological framework offers a clarification of the culture literature and
presents new methods o f linguistic analysis. Therefore, the current study serves as
a model for future research and clinical recommendations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
199
American Psychological Association. (1992). Ethical principles o f psychologists
and code o f conduct. American Psychologist 12. 1597-1611.
Arbona, C., Flores, C.L., & Novy, D.M. (1995). Cultural awareness and ethnic
loyalty: Dimensions o f cultural variability among Mexican american
college students. Journal o f Counseling and Development. 73. 610 - 614.
Ardila, A. (1990) Neuropsychology in Latin America. Clinical
Neuropsvchologist. 4C20. 121-132.
Ardila, A. (1995). Directions o f research in cross-cultural neuropsychology.
Journal of Clinical & Experimental Neuropsychology. 17(TL 143-150.
Bahrick, H. P., Hall, L. K., Goggin, J.P., Bahrick, L.E., & Berger, S. A .. (1994).
Fifty years o f language maintenance and language dominance in bilingual
Hispanic immigrants. Journal of Experimental Psychology: General.
123(3), 264-83.
Beminger, V. W., & Abbott, R. D. (1992). The unit o f analysis and the
constructive processes o f the learner: Key concepts for educational
neuropsychology. Psychologist. 21(2 \ 223-242.
Betancourt, H., & Lopez, S.R. (1993). The study o f culture, ethnicity, and race in
American psychology. American Psychologist. 48. 629-637.
Bialystok, E. (1986). Factors in the growth of linguistic awareness. Child
Development. 57. 498-510.
Bialystok, E. (1988). Levels of bilingualism and levels of linguistic awareness.
Developmental Psychology. 24(4). 560-567.
Bialystok, E. (1999). Cognitive complexity and attentional control in the bilingual
mind. Cognitive Development. 70f3T 636-644.
Bialystok, E., & Majumder, S. (1998). The relationship between bilingualism and
the development o f cognitive processes in problem solving. Applied
Psycholinguistics. 19fD- 69-85.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
Buriel, R. (1993). Acculturation, respect for cultural differences, and biculturalism
among three generations o f Mexican American and Euro American school
children. The Journal of Genetic Psychology. 154I4V 531-543.
Bumam, M. A., Telles, C. A., Hough, R. L., & Escobar, J. I. (1987). Measurement
of acculturation in a community population o f Mexican Americans.
Hispanic Journal of Behavioral Sciences. 9(21. 105-130.
Cervantes, R. C., & Acosta, F. X. (1992). Psychological testing for Hispanic
Americans. Applied & Preventive Psychology. 1. 209-219.
Cervantes, R. C., Padilla, A. M. & Salgado de Snyder, N. (1991). The Hispanic
Stress Inventory: A Culturally Relevant Approach to Psychosocial
Assessment. Psychological Assessment: A Journal of Consulting and
Clinical Psychology. 3131. 438-447.
Cherry, S. A. (1993). Neuropsychological considerations in the evaluation and
placement o f adult learning-disabled students. Dissertation Abstracts
International. 53I8-A1 2741.
Chung, R.G. (1994, August). Measurement o f Acculturation and Cultural Identity:
A Comparative Study. Paper Presented at The American Psychological
Association, Los Angles, California.
Clarren, S. B., Martin, D. C., & Townes, B. D., (1993). Academic achievement
over a decade: A neuropsychological prediction study. Developmental
Neuropsychology. 9f3-4\ 161-176.
Comas-Diaz, L. (1988). Feminist therapy with Hispanic/Latina women: Myth or
reality? Women & Therapy. 6(4). 39-61.
Corey, G. (1994). Theory and practice o f group counseling. Pacific Grove,
California: Brooks/Cole Publishing Company.
Costantino. G. (1992). Overcoming bias in educational assessment of Hispanic
students. In K. F. Geisinger (Ed.), Psychological testing of Hispanics (pp.
89-97). Washington, DC: American Psychological Association.
Cuellar, I., Harris, L.C., Jasso, R (1980). An acculturation scale for Mexican
American normal and clinical populations. Hispanic Journal o f Behavioral
Sciences. 2(3), 199-217.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
201
Cuellar, I., Arnold, B., Maldonado, R. (1995). Acculturation rating scale for
Mexican Americans-II: A revision o f the original ARSMA scale. Hispanic
Journal of Behavioral Sciences. 17f3 \ 275-304.
D ’Amato, R. C., & Rothlisberg, B. A. (1996). How education should respond to
students with traumatic brain injury. Journal o f Learning Disabilities.
29(6), 670-683.
Dana, R. H. (1993). Multicultural assessment perspectives for professional
psychology. Massachusetts: Allyn & Bacon.
Dana, R. H. (1994). Testing and assessment ethics for all persons: Beginning and
agenda. Professional Psychology: Research and Practice. 25. 249-354.
Dana, R. H. (1995). Culturally competent MMPI assessment of Hispanic
populations. Hispanic Journal o f Behavioral Sciences. 17f3 \ 305-319.
Diaz, R. M. (1985). Bilingual cognitive development: Addressing three gaps in
current research. Child Development. 56. 1376-1388.
Duran, R. P. (1992). Clinical assessment o f instructional performance in
cooperative learning. In K. F. Geisinger (Ed.), Psychological testing of
Hispanics (pp. 137-156). Washington, DC: American Psychological
Association.
Felix-Ortiz, M. & Newcomb, M. D. (1995). Cultural identity and drug use among
Latino and Latina adolescents. In G. J. Botvin, S. Schinke, & M. A.
Orlandi (Eds.), Drug abuse prevention with multiethnic youth (pp. 147-
165). Thousand Oaks: Sage.
Felix-Ortiz, M., Newcomb, M. D., Myers, H. (1994). A multidimensional
measure of cultural identity for Latino and Latina adolescents. Hispanic
Journal of Behavioral Sciences. 16(2). 99-115.
Franklin, M. E., James, J. R., & Watson, A. L. (1996). Using a cultural identity
development model to plan culturally responsive reading and writing
instruction. Reading & Writing Quarterly: Overcoming Learning
Difficulties. 12(TL 41-58.
Franzen, M. D., & Lovell, M. R. (1987). Neuropsychological assessment. In R. E.
Hales & S. C. Yudofsky (Eds.), Textbook of neuropsychiatry (pp. 41-45).
Washington, DC: American Psychiatric Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
Galindo, R., & Escamilla, K. (1995). A biographical perspective on Chicano
educational success. Urban Review. 27f P . 1-29
Garcia, E. E. (1986). Bilingual development and the education o f bilingual
children during early childhood. Special Issue: The education of Hispanic
Americans: A challenge for the future. 95(1). 96-121.
Geisinger, F. (Ed.). (1992). Psychological testing of Hispanics. Washington, DC:
American Psychological Association.
Groth-Mamat, G. (1990). Handbook of psychological assessment. New York:
John Wiley and Sons, Inc.
Guzman, M. E. (1986). Acculturation o f Mexican American adolescents.
Dissertation Abstracts International. 47151. 2166-2167.
Herrera, L., Ponton, M. O., Corona, M., Gonzales, J., & Higareda, I. (November
1997). Poster Presentation, National Academy of Neuropsychology,
Acculturation impact on neuropsychological test performance in a Hispanic
population. Las Vegas, NV.
Huang, G. G. (1995). Self-reported biliteracy and self-esteem: A study of
Mexican-American 8th graders. Applied Psycholinguists. 16(3). 271-291.
Kaplan, M. S., & Marks, G. (1990). Adverse effects of acculturation:
Psychological distress among Mexican American young adults. Social
Science in Medicine. 311121. 1313-1319.
Keefe, S. E. (1992). Ethnic identity: The domain of perceptions o f and
attachment to ethnic groups and cultures. Human Organization. 51(1). 35-
44.
Kemp, A. D. (1992). Counseling center psychologist in neuropsychology:
Counseling neuropsychology. Counseling Psychologist. 20f4). 571-604.
LaCalle, J. J. (1987). Forensic psychological evaluations through an interpreter:
Legal and ethical issues. American Journal of Forensic Psychology. 5(4).
29-43.
Lezak, M.D. (1983). Neuropsychological assessment. New York, NY: Oxford
University Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
Lopez, S. R., & Romero, A. (1988). Assessing the intellectual functioning o f
Spanish-speaking adults: Comparison o f the EIWA and the WAIS.
Professional Psychology: Research and Practice. 19(31. 263-270.
Lopez, S. R., & Taussig, I. M. (1991). Cognitive-intellectual functioning of
spanish-speaking impaired and nonimpaired elderly: Implications for
culturally sensitive assessment. Psychological Assessment: A Journal of
Consulting and Clinical Psychology. 3(31. 448-454.
Mahurin, R. K., Espino, D.V., & Holifield, E. B. (1992). Mental status testing in
elderly Hispanic populations: Special concerns. Psvchopharmacologv
Bulletin. 28(41. 391-399.
Manaster, G. J., Chan, J.C., & Safady, R. (1992). Mexican-American migrant
students' academic success: Sociological and psychological acculturation.
Adolescence. 27(105), 123-136.
Marin, G., Sabogal, F., Marin, B. V., Otero-Sabogal, R., & Perez-Stable, E. J.,
(1987). Development of a short acculturation scale for Hispanics. Hispanic
Journal of Behavioral Sciences. 9(2). 183-205.
Matsumoto, D. R. (1996). Culture and Psychology. Pacific Grove: Brooks/Cole
Publishing Company.
Matute, B. M. E. (1986). Ethnic identities and patterns of school success and
failure among mexican-descent and japanese-american students in a
California high school: An ethnographic analysis. Special Issue: The
education of Hispanic Americans: A challenge for the future. 95(11. 233-
255.
Miranda, A. O., & White, P. E. (1993). The relationship between acculturation
level and social interest among Hispanic adults. Individual Psychology.
49(1), 76-85.
Monjica, Y. R. (1991). Acculturation o f limited English-speaking and bilingual
Mexican-American high school students: A validation of the dual
acculturation scale-Spanish version. Dissertation Abstracts International.
53(3), 759-A.
Montgomery, G.T., & Orozco, S. (1984). Validation of a measure of acculturation
for Mexican Americans. Hispanic Journal of Behavioral Sciences. 6(1). 53-
63.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 04
Montgomery, G.T. (1992). Comfort with acculturation status among students
from south Texas. Hispanic Journal o f Behavioral Sciences. 14(2). 201-
223.
Negy, C. (1993). Anglo and Hispanic-Americans1 performance on the family
attitude scale and its implications for improving measurements o f
acculturation. Psychological Reports. 73. 1211-1217.
Negy, C. & Woods, D. J. (1992a). A note on the relationship between
acculturation and socioeconomic status. Hispanic Journal o f Behavioral
Sciences. 14(21 248-251.
Negy, C., & Woods, D. J. (1992b). The importance of acculturation in
understanding research with Hispanic-Americans. Hispanic Journal of
Behavioral Sciences. 14C2T 224-247.
Oetting, E. R., & Beauvais, F. (1990-91). Orthogonal cultural identification
theory: The cultural identification of minority adolescents. The
International Journal o f Addictions. 25T5A & 6A), 655-685.
Orozco, S., Thompson, B., Kapes, J., & Montgomery, G.T. (1993). Measuring the
acculturation of Mexican Americans: A covariance structure analysis.
Measurement and Evaluation in Counseling and Development. 25. 149-
155.
Padilla, A. M. (1995). Hispanic Psychology. Critical Issues in Theory and
Research. Thousand Oaks, CA.: Sage Publications.
Paniagua, F. A. (1994). Assessing and Treating Culturally Diverse Clients.
Thousand Oaks, CA.: Sage Publications.
Pennock-Roman, M. (1992). Interpreting test performance in selective admissions
for Hispanic students. In K.F. Geisinger (Ed.). Psychological testing o f
Hispanics (pp. 99-135). Washington, DC: American Psychological
Association.
Perez, A. J. (1990). The relationship between Chicano ethnic self-esteem, level of
acculturation, and cultural identity. Dissertation Abstracts International.
52(2), 1042.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
205
Pewitt, K. (1999). A profile o f America’s diversity— the view from the Census
Bureau, 1999. In R. Famighetti, W. A. McGeveran, Jr., L. P. Wiesenfeld,
B. R. Ellis, & K. Seabrooke (Eds.), The world almanac and book of facts
2000 (pp. 381-382). Mahwah, NJ: Primedia Reference, Inc.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for
use with diverse groups. Journal of Adolescent Research. 7(21. 156-176.
Phinney, J. S. (1996). When we talk about American ethnic groups, what do we
mean? American Psychologist. 51(91. 918-927.
Pialorsi, F. (1981). A pilot test to measure basic English pattern proficiency
among bilingual children. International Review of applied Second
Language. 19(TT 45-64.
Ponton, M. O., Satz, P., Herrera, L., Ortiz, F., Urrutia, C. P., Young, R., D’Elia,
L.F., Furst, C. J., & Namerow, N. (1996). Normative data stratified by age
and education for the neuropsychological screening battery for Hispanics
(NeSBHis): Initial report. Journal of the International Neuropsychological
Society. 2. 96-104.
Prediger, D. J. (1994). Multicultural assessment standards: A compilation for
counselors. Measurement and Evaluation in Counseling and Development.
27, 68-73.
Raven, J., Raven, J. C., & Court, J. H. (1993). Manual for the Raven’s progressive
matrices and vocabulary scales. Oxford: Oxford Psychologist Press.
Reynolds, C. R., & Kamphaus, R.W. (Eds.), (1990). Handbook of psychological
and educational assessment of children: Intelligence & achievement. New
York: The Guilford Press.
Ricciardelli, L. A. (1992). Bilingualism and cognitive development in relation to
threshold theory. Journal of Psvcholinguistic Research. 21(4). 301-316.
Riccio, C. A., Hynd, G. W., & Cohen, M. J. (1993). Neuropsychology in the
schools: Does it belong? School Psychology International. 14(4). 291-
315.
Rodriguez, O. (1992). Introduction to technical and societal issues in the
psychological testing o f Hispanics. In K.F. Geisinger (Ed.), Psychological
testing of Hispanics (pp. 11-15T Washington, DC: American
Psychological Association.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 0 6
Rogler, L. H., Cortes, D. E., & Malgady, R .G. (1991). Acculturation and mental
health status among Hispanics: Convergence and new directions for
research. American Psychologist. 46(6). 585-597.
Sabnani, H. B., & Ponterotto, J. G. (1992). Racial/ethnic minority-specific
instrumentation in counseling research: A review, critique, and
recommendations. Measurement and Evaluation in Counseling and
Development. 24. 161-185.
Sabogal, F., Marin, G., Otero-Sabogal, R., Marin, B. V., & Perez-Stable, E.
(1987). Hispanic familism and acculturation: What changes and what
doesn't?. Hispanic Journal o f Behavioral Sciences. 9(4). 397-412.
Saldana, D. H. (1995). Acculturative stress: Minority status and distress. In A.
M. Padilla (Ed.), Hispanic psychology: Critical issues in theory and
research (pp. 43-54). Thousand Oaks: Sage.
Salgado De Snyder, V. N. (1987). The role of ethnic loyalty among Mexican
immigrant woman. Hispanic Journal o f Behavioral Sciences. 9C 31. 287-
298.
Sandoval, J., & Duran, R. (1998). Language. In J. Sandoval, C. L. Frisby, K. F.
Geisinger, J. D. Scheuneman, J. R. Grenier (Eds.), Test interpretation and
diversity (pp. 181-211). Washington, DC: American Psychological
Association.
Standards for educational and psychological testing. (1985). Washington, DC:
American Psychological Association.
Stevenson, H. C. Jr. (1994). Validation of the scale of racial socialization for
African American adolescents: Steps toward multidimensionality. Journal
of Black Psychology. 20(4), 445-468.
Suinn, R. M., Rickard-Figueroa, K., Lew, A., & Vigil, P. (1987). The Suinn-Lew
Asian self-identity acculturation scale: An initial report. Educational and
Psychological Measurement. 47. 401-407.
Taussig, I. M., Henderson, V., & Mack, W. (1992). Spanish translation and
validation o f a neuropsychological battery: Performance of Spanish and
English speaking Alzheimer’s disease patients and normal comparison
subjects. Clinical Gerontologist. 11(3-4). 95-108.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
207
Wedding, D., Horton, A. M., & Webster, J. (Eds.)- (1986). The
neuropsychological handbook. New York: Springer.
Woodcock, R. W. (1991). Woodcock Language Proficiency Battery — Revised.
Chicago: Riverside Publishing.
Woodcock, R. W. & Munoz-Sandoval, A. F. (1995a). Woodcock — Language
Proficiency Battery — Revised. Spanish Form. Supplemental Manual.
Chicago: Riverside Publishing.
Woodcock, R. W. & Munoz-Sandoval, A. F. (1995b). Woodcock — Language
Proficiency Battery — Revised. Spanish Form. Chicago: Riverside
Publishing.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
APPENDIX A
SAMPLE FLIER
208
PLEASE CONSIDER BEING A PARTICPANT
IN AN IMPORTANT STUDY
At this time, normative data is not available to help doctors diagnose
neuropsychological problems o f Latino/a patients. We need your help to establish
normative data from individuals like yourself, who do not have serious or chronic
medical conditions. The test results will be used to improve the diagnostic process
and treatment intervention process of Latino/a individuals who experience
neurological deficits and psychological disorders.
The study consists o f a series o f tests which require approximately two hours to
administer. For example, you will be asked to answer questions, write, draw, and
manipulate small objects. During the test administration, you will be able to take
breaks, as needed. The majority o f participants find that the testing experience is
pleasant and interesting . The tests are administered at a local psychology clinic by
a licensed psychologist or graduate psychology student. All examiners are bilingual
and bicultural. All test results remain confidential and your name will not appear on
any of the data. Only raw scores and demographic data will be used.
To obtain further information, please call Marta Corona at (213) 975-9258.
Thank you for your time and collaboration with this project.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
20 9
APPENDIX B
Language Proficiency - Examiner's Form
Subject #:
Examiner:
Examiner's rating o f subject's Spanish and English language proficiency prior to administration o f
the Woodcock Language Proficiency Battery - Revised and the NeSBHis - Revised:
Level o f SPANISH language proficiency:
none a litde sufficient a lot
1 2 3 4
Comments:
Level o f ENG LISH language proficiency:
none a little sufficient a lot complete
1 2 3 4 5
Comments:
Which language was selected for questionnaires?
English Indifferent, but English
Spanish Indifferent but Spanish
complete
5
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
210
A PPE N D IX C
Exam iner’s Scoring Sheet
NAME:____________________ COUNTRY OF ORIGIN (Subject): _________________
AGE:_____________________________ Father:_____________Mother:____________
DOMINANT HAND: LEFT RIGHT FORMAL EDUCATION (Total Years):____________
SEX: M F Questionnaire Language Choice: SP ENG
Dem ographic D ata
C ultural Identity Scale
Language
Behavior/Familiarity
Value/Attitude
TOTAL
Hispanic Stress Inventory
I. Occupational/
Economic Stress
II. Parental Stress
III. Family/Cultural
Conflict
IV. Marital Stress
V. Immigration Stress
TOTAL
M arin A cculturation Scale
Language Use/
Ethnic Loyalty
Media
Ethnic Social Relations
TOTAL
#13
Exam iner’s Rating of
Language Proficiency
Test Results
English Spanish
Spanish Proficiency __________ ________
Sp. Language Preference_________ __________ ________
English Proficiency __________ ________
Familiarity with American __________ ________
Culture
Familiarity with Latino Culture __________ ________
Latino Activism __________ ________
Preferred Latino Affiliation _____ _ _
Percieved Discrimination
"Respeto”
Feminism
Immigrant Non-Immigrant
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
211
Woodcock-Johnson Language Proficiency Battery-Revised English Spanish
Oral Vocabulary _________ _______
Listening Comprehension __________ ________
Letter-Word Identification _________ ________
Dictation
END VII
+20
BEGIN__
DELAY
STROOP
COLOR
WORD
INTERFERENCE
FAS
F
S
TOTAL
WHO-UCLA AVLT
I
II
III
IV
TOTAL
Animals
Fruits
VI (B)
v n (ST)
vni (LT)
Repetitions
Intrusions
D C (RECOG).
Jseg.)
Jseg.)
Jseg.)
Jseg.)
_(errores)
_(errores)
_(errores)
Jerrores)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
212
END
BEGIN__
DELAY
GROOOVED PEGBOARD
RIGHT _________(sec.) LEFT (sec.)
________ DROPS ________
DROPS
P-S BOSTON NAMING TEST
WITHOUT HELP ________
SEMANTIC QUE ________
PHONEMIC QUE ________
D IG IT SPAN
FORWARD
BACKWARD
TOTAL
REY-OSTERREITH
COPY
+10” MEMORY
D IG IT SYM BOL
C O LO R TRAILS
DISENO DE CUBOS
RAVENS SPM
TOTAL
ERRORS
OLD TOTAL
A
B
C
. (sec.) 2
ERRORS
NEW
TOTAL
D
E
TOTAL
. (sec.)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
A P P E N D IX D
Demographics Questionnaire
English Version
Name: Subject #:
Date: Dominant Hand: R
Age: Date o f Birth: Sex: M
Personal Medical History
Do you drink alcohol? Y E S N O How many times per week?
Type o f alcohol?
Have you had alcohol within the last 24 hours? Y E S N O How much?
Are you currently using any type of medications? Y E S NO
Medicine: Dose:
Do you wear glasses or contact lenses? YES ___ NO
Do you or have you ever experienced hearing difficulties? Y E S NO
Have you ever had:
Epilepsy
Encephalitis
Brain Tumor
Convulsions
Tuberculosis
Diabetes
Cancer
Glaucoma
Meningitis
Arthritis
Embolism
Vascular Problems
Automobile accident which resulted in a head injuiy
Loss o f consciousness due to a head injury (duration:.
Psychological/Psychiatric treatment Hospitalization (duration:
)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 1 4
Generation
Where were you bom? Country:____ C ity:___Region:____ Rural Town Metropolitan City
How long have you lived in the United States? Y ears:______ Months:_______
Fathers place o f birth:_______ Region: Rural Town Metropolitan City
Mothers place o f birth:_______ Region: Rural Town Metropolitan City
Paternal grandfather's place of birth:
Paternal grandmother's place o f birth:
Maternal grandfather's place of birth:
Maternal grandmother's place of birth:
Paternal great grandfather's place o f birth:
Paternal great grandmother's place o f birth:
Maternal great grandfather's place o f birth:
Maternal great grandmother's place o f birth:
Main reason for migrating to the United States:
For you :_______ Economic
Political
Social
Other (Please specify:
)
Main reason for your family (for example: parents, grandparents, great grandparents):
Economic
Political
Social
Other (Please specify:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Marital Status
Are you:
Married
Separated
Widowed
Divorced (number o f marriages:_________________ )
Never married
Currently living with someone as though you are married
Education
What is the highest grade or year o f college that you completed?
None
Elementary and High School
Where?
When?
Circle the highest year completed:
K 1 2 3 4 5 6 7 8 9 10 11 12
College
Where?
________ When?
Circle the highest year completed:
13 14 15 16
Graduate or Professional School
Where?
________ When?
Circle the highest year completed:
17 18 19 20
Did you earn a high school diploma or pass a high school equivalency test?
NO
Diploma
Equivalency test (GED)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 1 6
What is the highest school degree that you have?
None
Vocational (Adult) school certificate o f completion
Associate o f Arts (A. A., junior college degree)
Bachelor’s (B.A., A.B.. o B.S.)
Master’s (M.A., M.S., M.S.W., M.P.H., etc.)
Doctorate or professional degree (Ph.D.. M.D., J.D.)
Where?
When?
Are you currently enrolled in school or college?
NO
YES, full-time
YES. part-time
Other Studies:
Your father's education:
None
Elementary and High School
Where?
________ When?
Circle the highest year completed:
K 1 2 3 4 5 6 7 8 9 10 11 12
College
Where?
________ When?
Circle the highest year completed:
13 14 15 16
Graduate or Professional School
Where?
________ When?
Circle the highest year completed:
17 18 19 20
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
217
Your mother's education:
None
Elementary and High School
Where?
________ When?
Circle the highest year completed:
K 1 2 3 4 5 6 7 8 9 10 11 12
College
Where?
________ When?
Circle the highest year completed:
13 14 15 16
Graduate or Professional School
Where?
________ When?
Circle the highest year completed:
17 18 19 20
Occupation
Which o f the following describes your occupation?
Executive I professional,
proprietor (owner) o f a large business
Manager or director o f a large business
proprietor (owner) of a medium size business,
middle management
Personnel Administrator,
owner o f a small business
lower management, farm owner
Clerk, office worker, or vendor,
technician, owner o f a small business,
farmer (manager)
Qualified worker, farm worker
Machine operator, worker
Less than full qualification
Student, housekeeper,
not employed, non-qualified worker
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
218
In what kind o f business or industry are you working? (record type of business — e.g.,
manufacturing, TV and radio, retail shoe store, etc.)
Specifically, what kind o f tasks are you doing in this field? (record a brief job description
emphasizing job title and tasks — e.g., typist, sales clerk electrical engineer, manager, etc.)
Job T itle:______________________________________________________________________
Job Tasks:_____________________________________________________________________
Which o f these best describes your life now?
Employed, 40 hours or more a week
Employed, less than 40 hours a week
Unemployed, looking for work
Unemployed, not looking for work
Student, Not Employed
Axe you self-employed? ______ NO YES
During the past 12 months, for how many months did you have a job?
Record the number o f months:____
Salary
Think about your total household income for the past year. By total household income I mean
salaries, wages, tips, commissions, social security, welfare, pensions, interest, dividends, property
sale or rental, alimony or child support received by you or anyone else in the family living with
you.
Less than $14,999
$ 15,000-$29,999
$30,000 - $44,999
$45,000 - $59,999
$60,000 - $74,999
$75,000 - $99,999
More than $ 100.000
How many persons were supported last year by the household's total income?
Record number o f persons:________
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 1 9
If you are married, please list vour spouse’s.
Education:
None
Elementary and High School______Where?______
Circle the highest year completed:
K1234 5 6 7 8 9 10 11 12
College
Where?
Circle the highest year completed:
13 14 15 16
Graduate or Professional School
Where?
Circle the highest year completed:
17 18 19 20
Occupation:
In what kind of business or industry is your husband/wife working? (record type of
business — e.g., manufacturing, TV and radio, retail shoe store, etc.)
Specifically, what kind o f tasks is your husband/wife doing in this field? (record a brief
job description emphasizing job title and tasks — e.g., typist, sales clerk electrical
engineer, manager, etc.)
Job Title:________________________________________________________________
Job Tasks:_________________________________________________________________
Religion
What is your religious preference?
Protestant
Catholic
Jewish
Buddha
Other (Explain:___ ______________________________ )
No religion
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
220
Level o f religious involvement?
None
A little
Some
A lot
A great deal
Language
Level o f SPANISH language proficiency:
complete
5
complete
5
Which language do you think i n ? ____________________________________________
Which language do you dream i n ? _________________________________ ___________
TH AN K YOU
none a little sufficient a lot
1 2 3 4
Level o f ENGLISH language proficiency:
none a little sufficient a lot
1 2 3 4
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
221
A P P E N D IX E
Reasons for Migration to the United States
Category “Other”
Main reason for migrating to the United States:
Husband came to study
To play soccer professionally and to get to know another country
Study
To learn English
To buy musical instruments and form a band in Guadalajara
Missionary
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
222
A P P E N D IX F
NeSBHis-R Means and Standard Deviations for Women by Age and Years of Education
(N=35) - Page One
A ge: 16-25 16-25 16-25 26-39 26-39 26-39 40-65 40-65 40-65
Years o f E ducation: 0-11 12-14 15-20 0-11 12-14 15-20 0 -1 1 12-14 15-20
Strooo Color
Mean: 73.00 72.50
-----
68.00 63.33 68.86 73.78 76.67 65.50
SD: 4.58 16.26
-----
12.11 6.43 10.01 9.48 11.50 14.55
StrooD Word
Mean: 47.00 53.50
------
43.75 44.67 41.00 44.56 41.33 43.75
SD: 7.94 16.26
------
6.8 4.04 6.03 5.41 4.93 8.81
StrooD Interference
Mean: 119.67 133.00 109.25 115.00 108.57 127.89 91.67 103.50
SD: 10.79 25.46
------
11.32 12.77 22.37 24.10 53.46 11.70
FAS Total
Mean: 34.00 26.00 25.25 36.33 39.43 38.11 27.00 33.50
SD: 10.58 14.14
------
8.81 1.53 11.30 10.55 5.00 5.92
FAS Animals
Mean: 16.00 16.00 17.50 19.00 17.14 18.56 15.33 15.75
SD: 7.21 11.31
------
3.79 3.00 5.73 5.32 3.06 2.23
FAS Fruits
Mean: 12.67 11.50 11.50 13.33 12.86 14.33 12.67 14.25
SD: 5.69 4.95
------
1.91 .58 3.85 2.55 2.52 2.99
AVLT Total
Mean: 50.67 54.50
------
56.25 53.67 51.29 50.67 56.00 54.50
SD: 8.02 3.54 6.24 2.52 3.86 6.32 2.65 8.85
AVLT Short Term Recall
Mean: 11.33 12.50
------
13.00 11.33 11.57 11.89 12.00 12.50
SD: 2.31 2.12
------
1.41 .58 1.13 2.57 2.00 2.38
AVLT Lone Term Recall
Mean: 12.00 14.00
------
13.50 11.67 11.43 11.89 12.33 13.00
SD: 1.00 .00
------
1.00 .58 1.62 1.90 1.53 2.45
AVLT Recoenition
Mean: 15.00 15.00
------
14.00 14.33 14.43 14.11 14.00 14.50
SD: .00 .00
------
1.15 .58 .79 1.27 1.73 1.00
Grooved Peeboard (Dom)
Mean: 70.33 70.00 66.50 72.33 64.00 83.11 64.67 70.50
SD: 9.45 4.24
------
5.26 6.03 7.59 23.44 2.89 6.66
Grooved Peeboard fN-Domf
Mean: 76.33 75.50 73.75 75.67 77.00 93.22 70.00 79.25
SD: 13.58 7.78 13.79 5.51 11.89 30.72 3.46 14.55
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
223
NeSBHis-R Means and Standard Deviations for Women by Age and Years of Education
(N=35) - Page Two
Age; 16-25 16-25 16-25
Years of Education: 0-11 12-14 15-20
Ponton-Satz BNT Total
Mean: 20.33 19.00
SD: 5.51 8.49
D ieit SDan Forward
Mean: 4.67 6.00
SD: .58 .00
Digit SDan Backward
Mean: 5.33 5.00
SD: .58 .00
Digit SDan Total
Mean: 10.00 11.00
SD: 1.00 .00
Rev-O Codv
Mean: 32.17 33.50
SD: 1.76 3.54
Rev-O Delav
Mean: 18.67 21.50
SD: 7.08 4.95
Color Trails 1
Mean: 29.33 36.50
SD: 4.04 12.02
Color Trails 2
Mean: 70.33 74.00
SD: 18.77 42.43
Digit Svmbol Total
Mean: 70.67 72.00
SD: 11.93 11.31
Block Design Total
Mean: 38.33 40.00
SD: 5.03 4.24
Ravens Total
Mean: 35.33 43.50
SD: 9.07 12.02
26-39
0-11
26-39
12-14
26-39
15-20
40-65
0-11
40-65
12-14
40-65
15-20
20.25
2.87
23.67
1.15
21.71
5.50
23.22
2.54
24.00
2.00
24.25
4.92
5.00
.82
4.67
.58
5.43
.98
4.78
.67
5.33
.58
5.25
.50
3.75
.50
4.67
.58
5.00
.58
3.89
.60
5.00
.00
4.00
.82
8.75
.96
9.33
1.15
10.43
1.27
8.56
.88
10.33
.57
9.25
.96
32.00
2.16
33.67
1.53
34.00
1.00
31.89
2.76
33.33
2.52
35.50
1.00
17.63
2.78
21.50
1.80
21.71
7.64
17.83
7.81
16.67
4.25
19.75
5.12
42.00
22.32
46.33
12.06
33.57
13.75
51.11
17.35
39.00
3.00
43.50
17.82
83.50
4.65
94.00
21.93
68.43
14.34
119.56
33.98
94.67
21.08
99.50
45.83
52.00
9.06
62.00
10.82
73.57
9.43
43.22
9.46
64.67
10.26
61.75
9.67
32.75
3.86
32.33
10.69
39.14
2.91
29.77
7.76
33.00
1.00
34.25
2.06
23.50
13.13
41.33
9.87
50.00
3.32
33.22
11.45
39.67
11.59
50.00
2.45
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2 2 4
A P P E N D IX G
NeSBHis-R Means and Standard Deviations for Men by Age and Years of Education
(N=43) - Page One
Age: 16-25 16-25 16-25 26-39 26-39 26-39 40-65 40-65 40-65
Years of Education: 0-11 12-14 15-20 0-11 12-14 15-20 0-11 12-14 15-20
Strooo Color
Mean: 70.00 65.00 70.33 72.00 62.60 67.73 95.00 57.00 63.50
SD: 10.60 12.79 7.51 5.29 11.80 15.02 24.04 14.88
StrooD Word
Mean: 55.00 41.44 48.67 56.00 47.80 42.65 51.00 37.00 38.83
SD: 13.75 7.09 8.62 9.54 8.38 6.14 7.07 5.00
Strooo Interference
Mean: 113.66 115.33 96.67 152.00 114.20 112.91 148.00 156.00 I01.0C
SD: 10.97 23.70 9.02 42.58 37.56 18.78 22.63 10.95
FAS Total
Mean: 27.67 36.78 26.00 26.00 32.80 38.55 27.00 58.00 39.17
SD: 11.93 8.27 13.23 .00 8.07 7.19 12.73 7.78
FAS Animals
Mean: 14.67 17.56 17.00 16.00 20.40 20.18 14.50 20.00 20.17
SD: .58 2.51 4.58 2.65 2.70 5.38 4.95
------
5.95
FAS Fruits
Mean: 12.33 12.78 11.33 10.00 13.80 15.55 13.50 11.00 14.50
SD: 1.15 4.12 3.06 2.65 2.68 4.74 .71 2.88
AVLT Total
Mean: 52.33 53.44 46.67 49.67 57.20 57.64 51.00 45.00 56.17
SD: 2.31 7.47 4.04 8.39 3.83 8.32 .00
------
6.82
AVLT Short Term Recall
Mean: 11.33 12.22 11.00 12.00 12.20 12.73 12.00 12.00 12.17
SD: 1.15 1.48 3.00 3.46 1.64 2.41 .00
------
3.60
AVLT Lone Term Recall
Mean: 11.33 11.67 10.67 12.33 12.40 12.82 12.00 11.00 12.83
SD: 2.08 3.32 2.08 3.79 1.67 2.99 .00
------
2.64
AVLT Recognition
Mean: 14.00 14.11 14.67 15.00 14.80 14.82 14.00 15.00 14.17
SD: .00 1.69 .58 .00 .45 .40 .00
------
1.17
Grooved Pegboard (Doml
Mean: 84.67 68.22 62.00 77.67 63.20 65.55 86.50 87.00 67.67
SD: 23.18 9.24 8.72 16.77 8.96 11.94 3.54 11.08
Grooved Pegboard (N-Dom)
Mean: 96.67 67.89 68.67 89.67 66.00 73.09 95.50 102.00 80.83
SD: 32.59 9.36 6.35 8.33 3.39 11.01 2.12
------
20.12
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
225
NeSBHis-R M eans and Standard Deviations for Men by Age and Years of Education (N=43) -
Page Two
Age: 16-25 16-25 16-25 26-39 26-39 26-39 40-65 40-65 40-65
Years of Education: 0-11 12-14 15-20 0-11 12-14 15-20 0-11 12-14 15-20
Ponton-Satz BNT Total
Mean: 15.00 21.67 23.00 21.33 27.20 25.09 20.50 25.00 25.67
SD: 7.00 4.47 3.46 2.08 .45 6.01 3.54
------
2.34
Digit Scan Forward
Mean: 6.33 5.11 5.00 4.67 5.20 6.00 5.00 5.00 6.33
SD: .58 1.05 .00 .58 .84 1.34 .00
------
1.37
Digit Soan Backward
Mean: 4.67 4.44 5.00 3.67 4.40 4.73 3.00 3.00 4.83
SD: 1.15 1.13 1.00 .58 1.34 1.56 1.41 ----- 1.72
Digit Soan Total
Mean: 11.00 9.56 10.00 8.33 9.60 10.73 7.50 8.00 11.17
SD: 1.73 1.81 1.00 .58 1.95 2.65 .71
-----
3.06
Rev-O C o d v
Mean: 33.33 33.72 36.00 34.00 34.60 35.18 30.50 34.00 33.83
SD: 2.31 3.23 .00 1.00 1.34 .40 2.12
-----
.75
Rev-O Delav
Mean: 21.33 22.17 24.67 20.67 23.10 25.18 16.50
—
16.67
SD: 4.04 5.54 1.15 8.08 2.82 6.26 2.12
—
4.97
Color Trails 1
Mean: 48.00 32.67 29.33 53.67 35.00 35.27 58.00 79.00 41.67
SD: 12.29 7.04 6.66 5.51 10.17 16.84 8.49
-----
6.12
Color Trails 2
Mean: 99.00 73.33 62.67 140.67 74.40 78.91 133.00 144.00 83.83
SD: 26.15 18.63 12.66 26.54 9.74 21.76 15.57
-----
15.66
Digit Svmbol Total
Mean: 46.67 66.44 54.67 43.33 59.20 63.36 35.00 38.00 66.17
SD: 20.23 4.45 13.05 11.93 4.97 14.66 9.90
__ __ _
8.16
Block Design Total
Mean: 30.67 37.44 40.67 34.67 37.80 40.64 24.00 20.00 38.16
SD: 5.77 6.35 3.06 7.64 3.90 5.14 5.66
-----
4.17
Ravens Total
Mean: 34.67 49.00 49.33 33.33 45.40 50.55 22.50 21.00 46.33
SD: 11.24 3.87 7.64 3.79 6.35 4.48 12.02
-----
13.23
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
A P P E N D IX H
226
NeSBHis-R Domains
Nonverbal Reasoning
Degree of Bilingualism
Ravens Total
- .48 **
r (78), * £ < .05; **£<.01.
Language
Degree of Bilingualism
Ponton-Satz Boston Naming Test (Total) - . 1 2
Verbal Fluency (FAS Total) - . 1 2
Verbal Fluency (Animals) - . 0 2
Verbal Fluency (Fruits) - .05
I (78), *£<.05; **£<.01.
Executive Functioning —Attention Concentration
Degree o f Bilingualism
Stroop Color . 1 2
Stroop Word . 35 **
Stroop Interference
3 4 **
Digit Span Total - .1 1
Digit Span Forward . 05
Digit Span Backward - . 2 2
Digit Symbol - .58 **
Color Trails 1
41 **
Color Trails 2
42 **
I (78), *£<.05; ** £ < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2 7
Memory
Degree o f Bilingualism
AVLT (Total: Lists 1-5) - .04
AVLT (Short Term Recall) .04
AVLT (Long Term Recall) - .01
AVLT (Recognition) .06
Rey-Osterrieth (Delay) - .05
1(78), * £ < .05; **£<.01.
Visual Spatial
Degree o f Bilingualism
Block Design (Total) -.23 *
Rey-Osterrieth (Copy) - .24 *
r (78), * £ < .05; ** £ < .01.
M otor
Degree o f Bilingualism
Grooved Pegboard (Dominant Hand) .48 **
Grooved Pegboard (Non-Dominant Hand) .40 **
I (78), *£<.05; **£<.01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Considering the impact of culture on the therapeutic relationship: A look at Latino school-age children in a special education setting

PDF

Development and validation of the Cooper Quality of Imagery Scale: A measure of vividness of sporting mental imagery

PDF

Predictors of dysphoria and substance abuse among Latino and African American youth

PDF

Effects of test interpretation style and need for cognition on elaboration, favorability, and recall

PDF

Mental and physical health consequences of intimate partner violence in a multi-ethnic sample of women

PDF

Gender differences in motivation for sexual intercourse: Implications for risky sexual behavior and substance use in a university and community sample

PDF

Hispanic culture, acculturation, and distress among caregivers of dementia patients

PDF

A cross-cultural comparison of marital power and dyadic adjustment among American, Indo-American, and East Indian dual-career and single-career couples

PDF

Assessment of racial identity and self -esteem in an Armenian American population

PDF

Intergenerational conflict, family functioning, and acculturation experienced by Asian American community college students

PDF

Integration of science and practice: A collective case study of scientist -practitioner programs in counseling psychology

PDF

An adolescent -parent conflict resolution skills training program for ethnically diverse families: A program evaluation study

PDF

A daily diary approach to compare the accuracy of depressed and nondepressed participants' estimation of positive and negative mood: A test of the depressive realism hypothesis

PDF

Effects of parenting style and ethnic identity on European American and Asian Indian adolescents' academic competence and self esteem

PDF

Child sexual abuse in a sample of male and female Hispanic and White nonclinical adolescents: Extending the reliability and validity of the Trauma Symptom Inventory (TSI)

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

PDF

A Positive Psychology goal intervention using strengths and values to enhance goals and increase well -being

PDF

Examining the differences in family functioning and meaning in life between Reform, Conservative, and Orthodox Jewish families

PDF

Psychological well-being, religious affiliation, and cultural identification in adult children of Jewish-Gentile interfaith marriages

PDF

Change beliefs and academic outcomes: A construct validation

Asset Metadata

Creator Corona-LoMonaco, Marta Elena (author)

Core Title Impact of language and culture on a neuropsychological screening battery for Hispanics

School Graduate School

Degree Doctor of Philosophy

Degree Program Education (Counseling Psychology)

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, mental health,OAI-PMH Harvest,psychology, clinical,psychology, psychometrics,sociology, ethnic and racial studies

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Newcomb, Michael (committee chair), Abreu, Jose (committee member), Jellison, Jerald (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-63221

Unique identifier UC11338703

Identifier 3017996.pdf (filename),usctheses-c16-63221 (legacy record id)

Legacy Identifier 3017996.pdf

Dmrecord 63221

Document Type Dissertation

Rights Corona-LoMonaco, Marta Elena

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA