The relationship of maternal factors and socioeconomic level to the reading achievement of 7-11 year-old black students

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE RELATIONSHIP OF
MATERNAL FACTORS AND SOCIOECONOMIC LEVEL
TO THE READING ACHIEVEMENT OF
7-11 YEAR-OLD BLACK STUDENTS
by
Joan M. Tassopulos
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Education)
December 1982
UMI Number: DP24948
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
U M I
Dissertation Publishing
UMI DP24948
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346
ProQuest
O k
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90 0 0 7
This dissertation, written by
Joan M. Tassopulos
under the direction of he.li.... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
D O C T O R O F P H I L O S O P H Y
Dean
Date.... Decembe r_ _ 1 982
DISSERTATION COMMITTEE
Chairman
Eddif e K. Williams, Ed.D.
Grayce M. Ransom, Ph.D.
^ - S ^ r —
Barbara Star, Ph. I).
IN APPRECIATION
I wish to express gratitude to the
Director, and other personnel, of
the Campus Reading Center at the
University of Southern California
for permitting their students and
facilities to be used for purposes
of this research study. Their con
sent to make the clinic available
for the collection of data, and
especially their support of personal
interview sessions for the mothers,
was greatly appreciated.
My gratitude is also extended to the
many mothers who were willing, without
exception, to come to the clinic for
special appointments in order to
communicate their attitudes and be
liefs about their children and give
information pertaining to the family.
Their support of this study was also
invaluable and appreciated. And, it
is because of their cooperation in
the project that extreme concern was
exercised to accurately reflect the
social realities of their subculture.
Joan M. Tassopulos
TABLE OF CONTENTS
Page
IN APPRECIATION 11
|LIST OF TABLES V I
Chapter
I. THE PROBLEM 1
Background of the Problem
I The Problem and Research Questions
Purpose of the Study
Conceptual Rationale
Substantive Hypotheses
Definition of Terms
Need for the Study
Setting of the Study
Scope and Delimitations
Limitations
I Organization of the Remainder of the Study
; II. REVIEW OF RELATED LITERATURE. .............. 13
Large-scale Surveys
Chicago and British Schools of Family
Environment Research
! Small-scale Educational Studies
Black Family Literature
Theoretical Framework of the Study
Concluding Remarks
| III. METHODOLOGY ......... .. .. .. .. 37
Research Design
Definition of the Variables
Instrumentation
The Mother-child Relationship
Evaluation (MCRE)
Peabody Picture Vocabulary Test (PPVT)
Wide Range Achievement Test (WRAT),
1976, Revised Edition
Gates-MacGinitie Reading Tests,
Second Edition, 1978
i i i
Chapter Page
III. METHODOLOGY (Continued)
| The Interview Schedule
I Improving Interview Method and
! Questionnaire Design
I Administration of Instruments
i The Sample
• Boundaries of Generalizability
Collection of Data
; Operational Hypotheses
Hypothesis I
Hypothesis II
Statistical Analysis
IV. PRESENTATION AND ANALYSIS OF DATA........... . 58
Preliminary Stage
Stage I
Stage II
Stage III
Findings from the Interview Schedule
Sources of Error
Summary of Findings
Preliminary Stage
Stage I
Stage II
Stage III
Interview Schedule
V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS . . . 91
Summary
The Problem
The Literature
The Methodology
Selected Findings
Conclusions
Recommendat ions
BIBLIOGRAPHY ................................ ...... 106
APPENDIXES........................... 114
iv
Page
APPENDIXES (Continued)
A. NORC-NYC Categories ............................ 115
B. Interview Schedule. ........................ 118
C. Measures Used for Variables......................122
D. Raw Data for Estimates, Expectancies
and Reading Test Scores.................... 124
E. Composites of Estimates, Expectancies
and Reading Test Scores........................127
F. Coding Scheme for Variables ................... 129
v
LIST OF TABLES
Table Page
1. Cumulative Frequencies and Cumulative
Percentage Frequencies for the
Distribution of Intelligence
Quotients .. ... .. .. ......... .... 59
2. Intercorrelations of Mother’s Estimates
for Word Recognition, Vocabulary and
Comprehension .................. 66
3. Intercorrelations of Mother's Expectancies
for Word Recognition, Vocabulary and
Comprehension ......................... 66
4. Intercorrelations of Reading Test Scores
for Word Recognition, Vocabulary and
Comprehension ..... 67
5. Means and Standard Deviations of Word
Recognition, Vocabulary and Comprehension
for Mother's Estimates, Mother's
Expectancies and Reading Test Scores......... 67
6. Crosstabulation of Maternal Nonacceptance
Scores and Socio-economic Index
(from lower to higher scale scores) ......... 69
7. Means and Range Values of Difference Scores
Calculated from Maternal Variables and
Reading Test Scores ....... 71
8. Partial Correlation Coefficients of all
Variables with Verbal Intelligence Used
as a Control Variable . . . . . . . . . . . . 74
9. Nonredunant Product-moment Correlations of
Maternal Variables to Determine
Multicollinearity . . . . . . . . . . . . . . 76
LIST OF TABLES (Continued)
Table Page
10. Summary Statistics for the Overall F
Test........... 77
11. Summary of Statistics for the Stepwise
Solution . . . . . . . . . . . 80
vii
CHAPTER I
THE PROBLEM j
| I
i |
Background of the Problem
In 1964 Stability and Change in Human Characteris- 1
j
j tics, by Benjamin Bloom, was published which was developed j
I i
around the theme that "much of what has been termed
individual variation may be explained in terms of environ
mental variation." (p. 9) He regards environments as j
i !
:having highly specific characteristics which have a bear-
iing on a child's growth and development. Since this
!
initial work by Bloom, a voluminous amount of material has |
i !
been written on the effects that environments produce on j
children's academic achievement.
j A similar position held by Wolf (1966) was that j
differences in test performance are primarily due to
environmental differences rather than individual differ
ences. On page 9 he cites Buros1 Mental Measurements
j Yearbook as containing thousands of instruments to measure ;
individual characteristics but few which measure environ
ments. An additional problem is that those environmental
measures that are available usually assess a general index
of socioeconomic status which summates a number of surface
environmental characteristics, obscuring the specific ways
lin which the environment affects behavioral characteristics.
Wolf (1966) postulates that the total environment is made j
!up of a number of sub-environments which directly relate to
'specific behavioral characteristics. In the case of the j
I
home environment, he is more concerned with ;
investigating what parents do in their interactions
! with their children rather than what parents are in
terms of status level of father's occupation, type j
of dwelling, source of income, and so forth. (p. 492)
' i
Evidence from family environment research has j
indicated that family measures have been successful in
accounting for variation in academic achievement scores. j
in a survey of reading comprehension in fifteen countries, I
|
Thorndike (1973) found that an appreciable amount of the j
I
prediction of reading achievement for individual students
was provided by information about their home and family j
i
i !
backgrounds. |
In final summary, then, the clear result is that good j
home and environmental backgrounds provide strong
differentiation between countries and within countries, ■
between students. (p. 179)
The Problem and Research Questions
The problem area in this study was one of discover
ing the modus operandi of family factors in the development
of reading differences in students. Identifying specific
family factors and differentially combining them to yield
varying results in reading achievement were the basic con
cerns of this study. Various lines of research have con
tributed to our understanding of "background" family
characteristics, but few investigations have been made into
the relationship between "specific" family characteristics
and reading achievement.
This study therefore addressed itself to three
major questions:
1. Which specific family factors are related to
children's reading achievement?
2. Do specific family factors have a stronger rela
tionship to reading achievement than background family
factors?
3. How great is the combined effect of background
family factors and specific family factors on reading
achievement?
Purpose of the Study
Exploration of the relationship between family
factors, specific and background, and children's reading
achievement was the primary intent of this study. To
investigate the direct effect of family factors on reading,
;it was deemed advisable to control for the presumed effects
of intelligence on reading achievement. The secondary
purpose of this study was to determine the differential
impact of specific family variables on children's reading
achievement. Due to the complex nature of the reading
process, three major components of reading achievement
(word recognition, vocabulary and comprehension) were used
in lieu of one generic concept.
Conceptual Rationale
Probably no single factor acting alone accounts for
differences in children's reading achievement; rather, such
achievement is the product of combinations of interacting
inputs. For example, specific family factors may interact
with background family factors to produce reading achieve™
ment in children. By themselves, some inputs are not .im
portant, but when combined with other inputs in the proper
proportions, they may become important. Bridge, Judd &
Moock (1979) have underscored this point.
Often it is not the presence of an input by itself
that is important. Rather it is the presence of two
(or more) inputs that makes a difference in the
production process. This is the phenomenon of
interacting inputs. (p. 16)
They also emphasize that the direction of effects
may not be one way.
In some cases, changes in an input will cause changes
in an output, which in turn will set off changes in
the original input variable. These reciprocal or
bidirectional effects can be tested empirically
with modern statistical models, (p. 30)
The present research adopted the foregoing concep
tual framework of Bridge, Judd & Moock (1979) to describe
the nature of the relationship between family characteris
tics and children's reading achievement. Specific family
factors were examined singly and jointly and in combination
with background family factors to determine when they become
important in reading achievement.
Substantive Hypotheses ;
i
The independent variables which were related to !
reading achievement included:
1. a set of specific family factors representing
maternal input (maternal nonacceptance of the child,
mother's prediction of child's reading ability, and mother's
lexpectancy of child's reading ability);
' i
2. one background family factor (socioeconomic level
I
iof the family); and
I . . . .
3. one personal characteristic of the child requiring
statistical control (verbal intelligence).
iThe three maternal variables were grouped into a logical j
cluster representing the effects on the child from one,
and perhaps the most significant, family member. This com
posite variable has been called the "maternal cluster." !
The dependent variables representing reading
achievement were three of the most commonly tested com
ponents of reading achievement: word recognition, voca
bulary, and comprehension.
The specific hypotheses tested were the following:
1. There is a positive relationship between family
characteristics--including a maternal cluster of factors
and socioeconomic level of the family--and children's
reading achievement, controlling for the effects of ver-
: bal intelligence.
2. There is a greater relationship between the
: maternal cluster and children's reading achievement than
between socioeconomic level and reading achievement, con
trolling for the effects of verbal intelligence.
Secondary consideration in this study was given
to:
1. the accuracy of the mother's knowledge of her
| child's reading ability,
2. the discrepancy between the mother's estimate of
her child's present grade level of reading and her expec
tancy of the grade level of reading performance that the
child should be able to reach (Harris & Hodges, 1981), and
3. the interaction between maternal nonacceptance and
the socioeconomic level of the family.
6
Definition of Terms
Some of the terms used in this study will be
defined in this section.
Family Environment: the social-psychological,
educational, and economic characteristics of the family
as a group which have a bearing on the child's develop-
| ment.
Maternal Nonacceptance of Child: the lack of an
adequate mother-child relationship in terms of sincerity
of affect expression, the fostering of independence,
interest in the child's pleasures and activities, and the
perception of the child as a good child (Roth, 1961).
Reading Achievement: the child's grade level of
reading performance obtained from standardized reading
tests in the areas of word recognition, vocabulary, and
comprehension.
! Socioeconomic Level: the location of the child's
| family in the social structure as determined by the occu
pation of the head of household, the income of the family,
and the education of the mother.
Need for the Study
The primary measure of family environment tradi
tionally used in research to account for variation in
7
student achievement has been socioeconomic status, a con
venient construct to measure. After reviewing the avail
able literature, Walberg & Marjoribanks (1976) conclude
that proximate, detailed assessment of factors in the
home provide for better predictions of children's ability
than socioeconomic status.
Comprehensive, proximal interview assessments are
preferable to socioeconomic status, family size, or
questionnaire measures of such things as the number
of books and magazines in the home. (pp. 548-549)
The importance of studying the processes occurring
in the home has been stressed by Bloom (1976)« He con
tends that
what adults do in their interactions with children in
the home is the major determinant of these characteris
tics rather than the economic level of the parents,
their level of education, or other status characteris
tics. (p. 2)
Although educators generally agree that family
environment influences academic achievement, they are only
beginning to understand the nature of the specific family
characteristics that affect the acquisition of this ability.
There appears to be a need for further research on this
level of analysis based on .insights already gleaned from
global measures. Attempt was made in this study to identify
with more clarity specific factors in the family environ
ment which relate to one academic area, that of reading.
Certain regularities of behavior can be understood
not only in terms of individual dynamics but also in terms
of group membership. After summarizing the theory,
research, and methods used in the study of whole families,
| I
Handel (1965) found a need to examine separate dyads within
the family. This study was designed to probe more deeply
into one family dyad, the mother-child relationship.
It is important to obtain more information regar
ding reading achievement for students from different
|ethnic backgrounds. Correlational and causal relationships
|
that have been established in the past for one ethnic group,
social class, or country may not hold for other groups.
The American black family has been studied historically
from diverse perspectives; new research findings are needed
to shed light on contemporary black family life.
Setting of the Study
! The Campus Reading Center at the University of
i
Southern California was chosen as the site for this study.
The community surrounding the University was classified as
a low-income area according to the figures in the 1970
census. This clinic is available to the public for diagno
sis and instruction in reading. Three sessions of instruc
tion are offered during the year: Fall, Spring, and
Summer sessions. Although no restrictions are imposed on
those children who enroll, the majority of students who
;attend are black and read below grade level.
9
; Scope and Delimitations
i
i
1. The topic of this empirical research was studied
within the framework of one ethnic minority.
2. The data collected for the study was obtained
solely from a university reading clinic. No attempt was
made to contact the student's full-day school for addi
tional information.
I
3. Only one subgroup of the family constellation was
intensively explored, that of the mother-child dyad.
Limitations
The limitations inherent in this investigation
!
were related to the data-gathering techniques, the instru
ments employed, and the basic assumptions of the theoretical
structure. The use of one interviewer to obtain informa
tion from the mothers may have introduced some personal
bias affecting the results. The study was perhaps also
limited by the reliability and validity of the assessment
J
instruments. The rating scale employed to measure attitudes
was subject to error resulting from the way in which respon-
I
dents characteristically distribute their ratings across a
scale, known as "response set" in the literature.
10
Concerning bias in item content of standardized
tests for minority groups, attention should be drawn to the
possibility that the tests utilized in this study may not
!accurately reflect the cultural experiences and attitudes
|of a black population. However, in the opinion of the
I
drafters of the Guidelines for Testing Minority Group
Children (Fishman, Deutsch, Kogan, North & Whiteman, 1964),
tests have not proved to be a panacea, but they are,
|nonetheless
among the most important evaluative and prognostic
; tools that educators have at their disposal, (p. 143)
Consideration must be given to the problem that
family theories generally have been devised with the white
family in mind and have been tested primarily on white,
middle-class samples (Heiss, 1975), It has been questioned
whether such theories can be applied equitably to all
groups in a pluralistic society such as the United States,
The present study may have been limited by this lacuna in
family theories, affecting the conceptual accuracy of
black family structure and function.
Organization of the Rema.inder of the Study
Chapter II consists of a review of the relevant
jliterature organized into large-scale surveys, family
i
environment research, small-scale educational studies,
11
black family literature, and the theoretical framework of I
i
|this study. The methodology adopted for the study is '
discussed in Chapter III; the research design, the defini- j
tion of the variables, the instrumentation, the sample, j
the procedures used to collect the data, and the techniques!
of statistical analysis are presented. Chapter IV contains
the findings of the study and the statistical analysis j
i
thereof following three sequential stages. A summary of the;
;
study appears in Chapter V along with the conclusions and
i
recommendations which were drawn from the findings. The i
|final section of the study includes a list of references
and the appendixes. I
CHAPTER II
REVIEW OF RELATED LITERATURE
This chapter will review some of the literature
which is relevant, either directly or indirectly, to the
purposes of the present study. The literature presented
below deals with the broad area of family environments
and children's academic achievement. The specific topics
covered are (1) large-scale national and international
surveys relating various inputs, including family variables,
to educational outcomes, (2) the "Chicago" and "British"
schools of family environment research, (3) small-scale
educational studies, (4) black family literature, and
(5) the theoretical framework adopted for this study.
Large-scale Surveys
In a survey of reading comprehension in fifteen
countries, Thorndike (1973) made cross-national comparisons,
concluding that schooling is relatively standardized
whereas extreme variations occur in home and family back
grounds. A major finding of the survey was that an appre
ciable part of the prediction for reading achievement of
individual students
is provided by information about their home and family
backgrounds. (p. 177)
After all, the home is the primary and continuing
source of influence on a child, and the school has
impact on him for a relatively small number of hours
a day and often for only half the days in the year
.... In final summary, then, the clear result
is that good home and environmental backgrounds pro
vide strong differentiation between countries and,
within countries, between students. (p. 179)
In the Plowden. survey (1967) , the environmental
icorrelates of cognitive performance were examined for
three age cohorts in a national sample of children from
173 primary schools in England. One of the criterion
variables was scores from the Watts-Vernon Reading Test.
A structured interview schedule was used to gather informa-
ition about the family environments of the children. Over
3000 interviews were achieved, and in most cases the mother
was chosen to be interviewed rather than the father. The
zero-order correlations computed from the Plowden data
showed that the amount of variance associated with the
social-status indicators for the achievement scores indi-
j
cated that social status has a pervasive influence on
children's outcomes.
Parents' attitudes were also surveyed in the
Plowden study--those attitudes regarding the education
their children were receiving, the kind of relationship
the parents have with the schools, and the role they
14
expect the schools to take in the upbringing of their
children. Comparing the variation in parental attitudes
with the variation in home background, of which socio
economic status was only one of the variables, the follow
ing conclusion was reached:
The variation in parental attitudes can account for
more of the variation in children's school achieve
ment than either the variation in home circumstances
j or the variation in schools. (p. 181)
i Further support for the argument that family
i
characteristics have a greater influence on academic
achievement than school characteristies is provided by the
landmark Equality of Educational Opportunity (EEO) survey,
commonly referred to as the "Coleman Report" (1966). This
survey, which was mandated by the Civil Rights Act of 1964,
!
involved 639,650 students in over 3,000 schools, represen
ting a five percent sample of all schools in the United
States. Taking into account all of the results, Coleman
and his co-workers drew the following implications from
the survey:
That schools bring little influence to bear on a
child's achievement that is independent of his back
ground and general social context; and this very
lack of an independent effect means that the
inequalities imposed on children by their home,
neighborhood, and peer environment are carried
along to become the inequalities with which they
confront adult life at the end of school. (p. 325)
; Relevant to the present study was the finding in
the Coleman Report that the average minority pupil scored
15
distinctly lower on the achievement tests at every level
than the average white pupil. For the black minority
I
group in particular, the students were unable to overcome
an initial deficiency in achievement, and this deficiency
in achievement became greater at progressively higher
grade levels.
A reanalysis of the EEO data set was performed by
Mayeske & Beaton (1975) from the U. S. Office of Education,
arriving at primarily the same results that the original
analysis had produced.
For virtually all students, the role of family back
ground factors in achievement exceeds that of school
factors. (p. 125)
They also noted that the unique portion of the total vari
ance explained by "family process" variables (e.g., expec
tations for excellence, educational plans, attitude toward
life) exceeded the unique portion for "home background"
variables (which included the student's SES). The common
portion of family process variables and home background
variables exceeded both unique portions.
Attitudinal and motivational factors tend to play a
greater explanatory role in achievement than do
social class factors, but there is also substantial
overlap among them. This overlap may represent the
effect of the interplay of social class and motiva
tional factors on one another over time. (p. 125)
16
I Chicago and British Schools
i of Family Environment Research
i
i
Bloom (1964) highlighted the fact that relatively
little effort had been devoted to the measurement of
environments in which individuals interact. Wolf's (1966)
subsequent work provided a rationale for the study of im-
jportant differences among environments. His conception of
the environment was that it was made up a number of
sub-environments, of which each has an effect on an indivi-
--------------------
dual's specific characteristics instead of measuring
surface manifestations of an environment. He endeavored
to concentrate on
what parents do in their interactions with their
children rather than what parents are in terms of
status level of father's occupation, type of dwelling,
source of income, and so forth. (p. 492)
The empirical findings from his study showed a correlation
!of 0.80 between the total environmental rating (defined by
i
-13 process characteristics) and the total achievement
battery score. Because this represents a high level of
relationship, Wolf suggested that it be considered suffi
cient for purposes of prediction. This study resulted in a
higher level of relationship than that found in previous
studies which used social status or some other general
index as a measure of the nature and quality of the
envi ronment.
17
Using the same 60 children that Wolf had examined,
Dave (1963) conducted a related investigation in which he
identified six press variables to define the family
environment: academic press, language models, academic
guidance, activeness of the family, intellectuality in the
home, and work habits in the family. The following list of
process characteristics, subsumed under the achievement
press variable, served to develop measurements of family
characteristics in the present study:
1. parental aspirations for the education of the child
I 2. parents' own aspirations
3. parents' interest in academic achievement
4. social press for academic achievement
5. standards of reward for educational attainment
6. knowledge of the educational progress of the child
7. preparation and planning for the attainment of
educational goals.
Some years later Bloom (1976) espoused a theory of
school learning, making the point that
the home is a powerful environment (for good as well
as harm) for the development of some of the basic
characteristics of the child that are fundamental
to further learning in the schools. Some homes do
j it well, while other homes do it rather poorly. (p. 2)
The "Chicago" school (represented in the works of
Bloom (1976), Wolf (1966) and Dave (1963) measured family
environments with greater precision than they had been
measured previously. Their studies generally were limited
by using small samples of children from only one ethnic
group, shortcomings which were overcome in part by studies
done in the 1970's in Australia and Canada.
Keeves (1972) carried out one of the few longitu
dinal inquiries in family environment research. His
paradigm for the study of educational environments was
'concerned, not with the level of academic performance, but
with the change in performance over time. He obtained
jmeasures from tests given one year apart to 242 twelve-
[year-old children in Australia. Of incidental interest
j
'regarding the reading test given to the subjects was the
correlation of 0.86 found between vocabulary development
and reading comprehension; it was decided therefore to
calculate a total reading score, or the sum of the
separate scores for vocabulary and comprehension. Keeves'
study was one of the first educational studies to use path
analysis to hypothesize a network of causal relations
i
between environmental factors and cognitive performance.
His analysis was restricted to the direct effects of
environmental measures on achievement, but it went beyond
most prior family environment research since it created a
model including a variety of environmental situations--
family structure, family attitudes, and family processes.
19
In another multivariate study of approximately
185 eleven-year-old boys in Southern Ontario, Canada,
Walberg & Marjoribanks (1973) correlated three socio
economic indicators, three family structure measures, and
eight home-environment processes with scores on tests of
verbal, number, spatial, and reasoning ability. An
environment schedule was constructed to assess the eight
press variables. A factor analysis of the environmental
forces produced two eigenvalues greater than unity, accoun
ting for 72 percent of the variance; the implication was
that a more parsimonious description of the home environ
ment might be made by using less than eight categories.
From computations of canonical correlations, high loadings
on the first canonical variate indicated that
environmental forces contribute more strongly to
the prediction of abilities than do the social
status indicators and family structure
characteristics. (p. 366)
Thus, several studies in recent years (Bloom, 1964;
Wolf, 1966; Plowden Report, 1967; Walberg & Marjoribanks,
1973; Mayeske & Beaton, 1975) have presented evidence which
supports the proposition that family process variables
account for more of the variance in children’s cognitive
scores than other environment measures such as social-
status characteristics and family structure variables.
20
Boocock (1972), argues to the contrary.
The family characteristic that is the most powerful
predictor of school performance is socio-economic
status (SES): the higher the SES of the student's
family, the higher his academic achievement. This
relationship has been documented in countless
studies and seems to hold no matter what measure
of status is used. (p. 36)
Small-scale Educational Studies
Some investigators have sought to identify those
parental attitudes and behaviors which are correlated with
achievement behaviors in children from differing socio
economic populations.. The two following research studies
were designed to use a middle-class or a lower-class sample
the findings in these studies indicate an inconsistency in
the relationship between parental variables and children's
academic outcomes. A study in the early 1960fs
(Crandall, Dewey, Katkovsky & Preston, 1964) conducted at
the Fels Institute, used families residing in villages in
southwestern Ohio, which generally had educational back
grounds and a vocational status higher than current
national norms. It revealed that mothers' evaluations of
their children's intellectual competence were positively
related to the children's actual academic performances. A
study (Solomon, Houlihan, Busse & Parelius, 1971) with a
sample of lower-class black students from the west side of
Chicago demonstrated that none of the parent behaviors
explored related to boys' academic achievement but two
parent factors were related to girls' academic achieve
ment— mothers' warmth and fathers' encouragement of inde
pendent achievement efforts.
One segment of the literature on parental atti
tudes and children's academic achievement has focused on
; the learning-disabled population. The research in this
area indicates a difference in attitudes between parents
;who have normally-achieving children; mothers in families
with a child who has learning problems tend to have low
expectations for that, child and high expectations for the
normal sibling. Several studies document these generali
zations. (Discussion is restricted to summary results of
the studies.)
1. Using only girls, Della-Piana & Martin (1966)
showed that mothers of underachieving girls exhibited
significantly more negative social-emotional reactions
than mothers of over-achieving girls.
2. The study of Wetter (1972), from a sample of out
patients from the U.C.L.A. Center for Health Sciences,
discovered that mothers of children with learning disorders
displayed greater overprotection, overindulgence, and re
jection toward their child than mothers whose children did
not present a learning disorder.
22
3. Humphries (1977) investigated consistency of
childrearing attitudes in mothers of boys with a learning
idisorder and in mothers of normally-achieving boys; a
’significant difference between the two groups was observed.
4. Comparing normally-achieving and learning disabled
children, Chapman & Boersma (1979) discovered that mothers
of LD children reported more negative and fewer positive
| reactions to their children, and they also held lower
academic performance expectations for their children.
' 5. In another study (Epstein, Berg-Cross & Berg-Cross,
1980) maternal expectations and birth order in families
with learning-disabled children were examined. Results
from the questionnaire data suggested that these mothers
became more overprotective toward the disabled child than
toward the normal sibling.
Some of the research effort has concentrated on
one area of achievement, namely, reading ability, as it is
influenced by parental factors. Both academic and intellec
tual home environments were reported by Koppenhaver (1972)
as being significantly related to reading achievement.
Another investigation conducted by Oidick (1977) studied
(mothers' behavior and sons* reading achievement and projec-
j
ited feelings. He hypothesized that pressure from mothers,
in the form of instigation and participation in reading
23
activities with their sons, interferred with the develop
ment of their sons' reading; his hypothesis was supported
by the findings. The study of Dix (1977), however, did not:
reveal a significant family influence on children's read
ing. He studied fifth grade students in Harlem and found
parent participation in school affairs and interaction
between mother and child in the home to have very low |
correlations with student's reading achievement. j
i j
These studies and others in the 1970's were spear-
t i
headed by a conference on Parents and Reading jointly
j sponsored by the International Reading Association and the j
National Congress of Parents and Teachers (Smith, 1971).
It was the first IRA Conference solely devoted to the role
of parents and the home in reading instruction. A high
degree of interest on the topic parents and reading was j
j
generated by this conference.
: i
I Presently, a study funded by the National Institute
of Education, is being conducted in Boston on a longitu
dinal basis with a sample of 30 elementary children and
their families, all of which are working class and low
income (Snow, 1982). An objective of the study is to
assess the relationship between children's out-of-school
environment and their growth in literacy skills and tested
achievement. One of the questions asked during the per
sonal interview was regarding parental knowledge of the
child's actual school achievement. The results of this
intensive study have not yet been published.
There has also been an analysis in the literature
of psychosocial factors related to the home environment
that affect reading achievement. It is Sartain's (1981)
|belief that
concept development experiences supplied in the home
provide an essential foundation for learning to read,
but it appears that the emotional environment pro
vided by the family may be equally important. (p. 7)
Harris (1970) has observed that remedial programs in read
ing are often planned with as much attention to the child's
emotional and social problems as to cognitive handicaps.
Taylor (1964) has summarized the literature over a
thirty-year period regarding the relationship between
personality traits and discrepant achievement. One con
clusion he reached was that the less conflict a student
i
has over independence-dependence relationships, the more
|effort the student places on achievement. In 1969 a report
of the thirty-year follow-up project of Athey & Holmes,
supported by the U.S. Office of Education, was published
which related reading success to personality characteris
tics. One of their findings was that the poor reader
;seemed to have a strong family orientation--he/she led a
more sheltered life and was treated more like a child by
his/her parents.
25
Because of the established interrelatedness of
reading achievement and intellectual ability, a number of
the aforementioned research studies incorporated a measure
of intelligence: Crandall, Dewey & Katkovsky (1964);
Della-Piana & Martin (1966); Solomon, Houlihand, Busse &
Parelius (.1971) ; Koppenhaver (1972) ; Oidick (1976) ; and
Humphries (1977). It should be noted, though, that in
recent years IQ testing has become a controversial issue,
especially with respect to minority groups*
Black Family Literature
Turning to the topic of black family structure and
function, one finds the repeated statement in the litera
ture that black family lifestyles are misrepresented and
often stereotyped* Instead of depicting the diversity
within the domestic units of blacks, the family structure
has been characteristically depicted as an undifferenti-
ated phenomenon. Peters (1981) discovered this unitary
perspective in the research on the behaviors and attitudes
of black parents, that is, a perspective that does not
recognize the variations in black family lifestyles. Her
historical review of the available studies on black family
life revealed many problems and methodological weaknesses,
some of which are cited herewith:
26
1. The black family is viewed as monolithic; few
studies include both lower-class and middle-class groups.
2. Culture-specific influences, such as delineation
of race of the participants, are ignored in many studies.
3. The primary focus of research on the black family
has been directed toward so-called problem populations—
single parents, nonachieving children, low-income families,
4. The research design frequently used to study a
black population has centered on intervention with the
intent of changing the target population.
In essence, much of the interdisciplinary work on
the black family has been undertaken with the implication
that a social problem is involved. This approach has been
labeled the "social pathological model" for it postulates
that the familial, social, and cultural milieu accounts for
behavior that deviates from the normative system; accep
tance of the model has likewise affected social planning,
educational policy, and research (Samuda, 1975). Baratz &
Baratz (1970), opponents of the model, found that the black
mother has been described by some researchers as inadequate
and responsible for producing "linguistically and cogni
tively impaired children who cannot learn," (p. 36) They
have denounced the "inadequate mother hypothesis" and its
corollary that we need parent training programs to eradi
cate the problem.
27
In the past decade social scientists have shifted
from this pathological model (a model emphasizing family
disorganization and dysfunction) to an ecological approach.;
I
The writings of Lewin (1935) had an important influence in j
providing a theoretical framework for ecologically-oriented:
j
research. This approach analyzes parental and child be
havior within the environment in which it occurs according
to the value system of the family's indigenous culture
(Peters, 1981).. The ecological approach was adopted for
purposes of this study. !
Descriptive information available for the black
I
population can perhaps give us some knowledge of the |
realities of the lives of black families. Data from the
U.S. Bureau of Census, 1975, indicates that the typical j
black family was found to be a two-parent household unit
I
with a stable marriage, controverting the stereotype of a ;
i
female-headed home; however, the proportion of husband-
wife families between the years 1970-75 declined from 68 ;
percent to 61 percent while the proportion of female heads
increased from 28 percent to 35 percent. Female heads
did comprise a majority among low-income black families.
In contrast to white women, black women were more likely
to be in the labor force.
Another reliable source of data is available from
reports of the National Urban League. In its publication
The State of Black America 1980, prominent characteristics
— — ----------------------
of the present-day black family are described. They
include strong kinship bonds, strong work orientation, j
flexibility of family roles, strong achievement orientation,
and strong religious orientation. From an overview of the
1970's, a revival of the extended black family was evident
in terms of support being provided for family members.
Contrary to the popular belief of black progress during the|
last decade, the facts revealed this theme to be an illu- i
i
sion. In actuality, there were more blacks who were poor j
at the end of the 1970's than at the beginning. The median
income in 1974 for black families was estimated at $7,800,
and for white families it was $13,400 (U.S. Census, 1975). ;
The U.S. Commission on Civ.il Rights (.1978) has also
made information available on the disparity between the j
position of black and white Americans, a matter of some j
controversy. While the educational, social, and economic
conditions of middle-class blacks have improved, the con- j
ditions of poor blacks have worsened. A variety of
economic indicators are presented in the Commission's re
port to support this view. For instance, in the period
1969-75 there was a decline in average earnings differen
tials between blacks and whites but an increase in poverty
rate differentials.
I
2 9!
A sociological study by Heiss (1975) investigated
a number of traits of the black family--size, interaction
patterns, stability, age at marriage, household composit ion,|
and kin relations. His data came from a secondary analysis'
of information originally gathered by Cain and the National
Opinion Research Center. The findings, in summary, which !
are of particular relevance to this study are: (
i
1. the structure of the black family is different from;
that of the white family in a number of ways; '
2. these differences are not merely SES differences j
in disguise; and
3. the differences, however, are not great. |
He concluded that the family dynamics for blacks and whites
i
are similar, not identical. |
I
This conclusion reached by Heiss differed from that
of Peters (1981),. Her analysis, it should be remembered, |
is in effect a "historical" one since it was based on a
survey of research studies spanning several decades, whereas!
that of Heiss issued from one large study using a national
probability sample. To quote Peters: :
Research on black families overwhelmingly shows that
the behaviors and lifestyles of black people are
different from those of whites. Their child-rear.ing
priorities, attitudes, and patterns of behavior have
developed out of the exigencies of the unique economic,
cultural, and racial circumstances in which they have
lived. (p. 221)
30
The literature was searched for information on one
specific aspect of the lives of black families--the educa
tional studies have recorded lower average scores on
academic achievement tests of black youths than white
youths. Recent trends however are changing, according to
an article in the Educational Researcher (Burton & Jones,
1982). From surveys conducted by the National Assessment
of Educational Progress, Burton & Jones reported that dif
ferences between black and white students along several
indicators of fundamental skills and knowledge are
decreasing. Their discovery of this recent improvement in
the relative position of black students was unmistakably
expressed in the following:
The results summarized here show that during the
1970's the discrepancy in average achievement level
between the nation's white and black youth has be
come smaller in five important learning areas at
ages 9 and 13. Typically, when achievement for
white students has declined, that for black students
has declined less; when whites have improved, the
blacks have improved more. The difference between
; the races has decreased at both ages in mathematics,
I science, reading, writing and social studies. (p. 14)
The related topic of achievement orientation has
been discussed in The State of Black America 1980. The
findings reported indicate that strong achievement orienta
tion is just as prevalent in one-parent black families as
in two-parent black families. In a review of the litera
ture on black educational achievement, Banks & McQuator
(1976) examined the evidence regarding the socialization
31
pf black children toward academic success. They did not j
I
find a lack of value placed on academic achievement. j
A search for family and early socialization contribu- j
tions in academic failure seems misdirected. The |
empirical evidence discounts the hypothesis that
black homes instill values contrary to academic !
success; also, initial value orientations alone are I
unable to sustain the achievement orientations |
necessary for goal attainment. (p. 6) !
; i
McAdoo (1977), in a study of 100 middle-income !
black parents, considered components of educational achieve-!
ment in terms of socio-economic factors. Aspirations for j
; I
higher educational achievement were viewed as a prerequisite)
to upward social mobility for the black individual. The !
subjects in the study declared that the desire for higher j
education was instilled early by their parents, who made j
qreat sacrifices to help them achieve middle-class status. j
To summarize the major points of the foregoing !
discussion of the American black family, the decade of the j
j
j
1970's for blacks was a period in which ;
|
1. the predominant family unit had two parents but [
this type of family structure was decreasing in frequency, :
2. social and economic factors improved for middle- ;
class blacks but worsened for poor blacks,
3. the extended family provided many forms of kin j
help, and
32
4. the differences in educational achievement for
black and white youth decreased in measures of outcome for
a variety of learning areas.
Theoretical Framework of the Study
The information gathered in this study was analyzed
within the context of an interactional model to understand
the multi-dimensional patterns of reactions for individuals
within the family. Recent formulations of interactional
psychology posit that
1. behavior is determined by a continuous process of
interaction between the individual and the situation he/she
encounter s,
2. the psychological meaning of the situation to the
individual is an essential determinant of behavior, and
3. cognitive factors are important in interaction.
Thus, not only do situations affect individuals; the person
also actively selects and intentionally influences the
situations in which he/'she interacts. Endler & Magnusson
(1976) have aptly stated this position:
According to interact ionism, actual behavior is the
result of an indispensable, continuous interaction
between the person and the situations he encounters
(i,e., B= f (P , S) ) . , This implies that the individual's
behavior is influenced by significant features of the
situations, but furthermore the individual chooses
the situations in which he performs and selects
significant situational aspects which then serve as
cues for his activities in these situations. He sub
sequently affects the character of these situations.
(P- 3)
33
A principal focus of conceptualization about the
family has been interaction patterns. Because of the
questionable utility of the cause-effect model of parent
influencing child, the interaction concept has come to the
fore as an alternative approach for analyzing familial be
havior. The adequacy of this concept, however, is also
problematic, regardless of whether the data are obtained
by direct observation of behavior or by subject reports.
One difficulty is the fact that an ahistoric framework for
studying the family in a laboratory-type setting is not
well-suited to this kind of social group because of its
enduring interpersonal relationships. The dilemma as to
whether interaction or the interpersonal relationship will
prove to be a fruitful concept in understanding families
has yet to be resolved.
Traditionally, most of the study of subfamily
pairs within the nuclear family has been on the mother-
child dyad. Handel (1965), after summarizing the theory,
research and methods for the study of whole families, saw
the need for studying the person within the context of
dyadic relationships in the family. Earlier support for
this position was given by Adler (.1923) who argued that
the individual could be understood only by studying the
social groups in which he/she lived. He influenced much
_3Al
of the conceptualization of family relationships by intro
ducing the concepts of maternal indulgence and rejection.
Such terms as rejection and overprotection seem to
have emerged out of the child guidance movement which had
its beginnings in the early 1920's, according to Symonds
(1949) , after performing an extensive review of the litera
ture. He developed a two-dimensional model for parent be-
! I
havior of acceptance-rejection and dominance-submission, ;
anticipating more recent studies. From the field of excep
tional children, Schaefer (1975) contends that there is ;
abundant evidence to demonstrate that academic achievement j
:is related to parental acceptance and family support for |
:academic achievement.
Another important part of the lit
relationships has dealt with the theme of
Concerned with the concept of positive me
Jahoda (1958) has advocated a multiple cr
to explain this phenomenon in lieu of a u
Three of her criteria (which have particu
this study) share an emphasis on reality
autonomy, perception of reality, and envi
Thus, there are various currents
which endorse an interactional model to u
patterns.
erature on family
adaptat ion.
ntal health,
iterion approach
nitary cause,
lar relevance to
orientation—
ronmental mastery
in the literature
nderstand family
Concluding Remarks
Research studies and theoretical considerations ;
i
i i
|have been presented which bear upon the relationship j
between family factors and academic achievement, with j
j
special attention given to the black family. The environ
ment provided by the family has a powerful impact, undis-
! !
putably, on a child's educational attainment. Black |
families, as one population subgroup, have a set of parental
|
attitudes and beliefs regarding academic achievement which j
developed out of the unique circumstances in which they J
I
' -i • !
live.
We can conclude from the foregoing literature
review that both maternal behavior and reading behavior
!are complex phenomena which necessitate a multiple factor
'theory for understanding their interrelatedness. This
study was therefore conceived as an investigation of a
network of constructs for these two categories of pheno- j
j ;
!mena within the framework of interactionism. I
CHAPTER III
METHODOLOGY
Research Design
Without attempting to obtain premature clarity
concerning the causal nature of the variables, it was
I
decided to do an exploratory--type investigation in which
a set of family variables was related to a set of reading
achievement variables. Intensive analysis of both sets of
variables was needed to more fully understand the nature of
these variables and to determine if they were independent
i
phenomena or if, in fact, they represented overlapping
characteristics of a more unitary phenomenon. Due to the
multiplicity of variables, a multivariate correlation j
j
technique was chosen to describe the joint and unique \
contributions of the several family variables. Thus, the ■
study sought to examine the "relationship" of family j
factors to reading achievement in lieu of offering a causal,
explanation of this relationship.
Definition of the Variables
Initially there were nine independent variables
and three dependent variables, all of which were attribute
in nature. They are listed below along with their opera
tional definitions.
1. Independent Variables:
a. maternal nonacceptance: the mother's nonac
ceptance of her child assessed by two attitude scales taken
from the Mother-child Relationship Evaluation (hereafter
referred to as the MCRE)
b. mother's estimate of word recognition: the
grade-level estimate of the child's word recognition
ability as stated during the personal interview
c. mother's estimate of vocabulary: the grade-
level estimate of the child's vocabulary ability as stated
during the personal interview
d. mother's estimate of comprehension: the grade-
level estimate of the child's comprehension ability as
stated during the personal interview
e. mother's expectancy for word recognition: the
grade-level expectancy for the child's word recognition
ability as stated during the personal interview
f. mother's expectancy for vocabulary: the grade-
level expectancy for the child's vocabulary ability as
stated during the personal interview
.38
g. mother’s expectancy for comprehension: the
grade-level expectancy for the child's comprehension
j
I ability as stated during the personal interview ■
j
1 h. socio-economic level of the family: a socio
economic index of the child's family primarily based upon
i
I
the occupation of the head of the household, using the
rating scheme of the National Opinion Research Center ;
Occupation Scale and the New York City Planning Commission ,
I
on Labor Force Participation (hereafter referred to as the
NORC-NYC scale) !
i. verbal intelligence of the child: the child's j
j
IQ score obtained from the Peabody Picture Vocabulary Test
(hereafter referred to as the PPVT).
2. Dependent Variables: !
I a. word recognition test score: the grade-level
equivalent score of the child's performance on the reading i
| i
subtest of the Wide Range Achievement Test (hereafter I
referred to as the WRAT)
b. vocabulary test score: the grade-level
equivalent score of the child's performance on the vocabu- :
lary subtest of the Gates-MacGinitie Reading Tests (here
after referred to as the Gates-MacGin.itie Test)
c, comprehension test score: the grade-level
equivalent score of the child's performance on the
comprehension subtest of the Gates-MacGinitie Reading
Test s.
The variable maternal nonacceptance was measured
by the overprotection and rejection scales of the MCRE.
Only these two scales from the instrument were used,
although the entire questionnaire was answered by the
mothers, because these scales are measures of the attitude
dimension being presently examined, nonacceptance, and
because of their stronger psychometric properties. In the
manual for the instrument (Roth, 1961) the two terms are
def ined.
1. Overprotection is an expression of prenatal
anxiety in terms of prolonged infantile care,
prevention of development of independent
behavior, and an excess of control. (p. 2)
2. Rejection is the denial of love and an
expression of hate towards a child in terms
of neglect, harshness, severity, brutality,
and strictness. (p. 1)
This instrument has been used for research purposes in a
school setting by Hilliard & Roth (1969) to examine the
mother-child relationship of two groups of students:
achievers and under-achievers.
The variable socio-economic level, according to
Miller (1970), has been found to have occupation as its
single best predictor. He claims that researchers will
achieve a high degree of predictive efficiency from an
occupational scale alone. For this reason, it was decided
that occupational prestige would be the major criterion for
measuring this variable. The classification scheme for
rating occupations selected for the study was the NORC-NYC
scale which incorporates seven categories: laborers,
service, operatives, craftsmen, clerical and sales,
managerial, and professional. However, because of practi
cal problems which customarily arise in research when using
only an occupational scale to determine socio-economic
level, more information was gathered in the personal inter
view. Educational attainment of both parents and yearly
family income were two additional measures obtained for
this variable to be used in cases with questionable data.
There was a need in the study for an easily
administered test to measure intelligence that would pro
vide a global score. In a comparison of several measures
of intelligence, Pikulski (1973) recommends the Peabody
Picture Vocabulary Test for the purpose of obtaining a
rough, global measure of intelligence. This test provides
an estimate of the subject's verbal intelligence through
measuring his/her hearing vocabulary, using pictorial
representations. The PPVT does not depend on reading
ability, which makes it particularly well suited for a
study using subjects, many of whom have reading problems.
41
Therefore, the PPVT was chosen for the present study
because of its practicality in a clinical setting and ;
because it does not depend on reading ability. i
i
For the collection of the dependent variables j
representing reading achievement, three comprehensive
reading skills were selected--word recognition, vocabulary,|
and comprehension— since these skills represent major areas;
of reading instruction for the elementary-school child.
Persistent problems have been related to measurement of j
the dependent variable in reading research; the validity
of measures of specified subskills in reading has been the ;
subject of a number of investigations (Farr & Roelke, j
1971). Because the issue of discriminant validity has not ;
yet been resolved, it was decided for this study to corre- j
late three test scores, representing performance in diversej
reading-skill areas, to determine if these skills were i
sufficiently independent of one another to be treated as
separate variables. The reading test scores were taken
from those tests that were routinely given to all students !
at the reading clinic during the beginning of their session.
The reading subtest of the WRAT is given as a word recogni
tion and placement test; both components of the Gates-
MacGinitie (the vocabulary subtest and the comprehension
subtest) are also given to the students. The type of
test score used for purposes of this study was a grade-
I
ilevel score because it is more readily understood by a
i
mother. This is not to say that such scores are more
meaningful than other kinds of scores but that they are
more familiar to a lay person.
Inst rument at ion
Information pertaining to the psychometric
characteristics of the scales and tests employed during
The data gathering process was found to be limited. The
technical manuals reported reliability coefficients based
on standardization groups with different characteristies
than that of this study; not all of them used black sub
jects in their norm group. It is suggested, therefore, j
i
that the following information be interpreted with this j
|
limitation in mind. j
i
The Mother-child Relationship Evaluation (MCRE) j
i
By Robert M. Roth (1961): 80 middle-class white j
women living in a suburb of Austin, Texas, comprised the ;
i
standardization group. For this group, split-half relia- j
bility coefficients for the subscales ranged from 0.41 to
i
0.57, and the intercorrelation of the overprotection and !
i
rejection scales, indicating construct validity, was 0.40. ;
In their review of the instrument for Personality Tests and
Reviews, Bell (1970) and Harris (1970) caution users on
I
43 !
the limitations of the standardization group and the lack
of precise definitions of the abstract attitude categories
being measured. However, in the study of Jillings,
Adamson & Russell (1976) the construct validity of the
instrument was supported; the scale intercorrelations
showed a pattern similar to that cited by Roth (1961)
despite the fact that the characteristics of the sample,
consisting of mothers of handicapped children, differed
from those of the standardization group.
The author of this self-report inventory states
that it is primarily an exploratory and experimental one
(Roth, 1961). The scale is global, evaluating the
acceptance-nonacceptance dimension of the mother-child
relationship as defined by Symonds (.1949). By indicating
the extent of agreement or disagreement on a five-point
Likert-type scale, the subject rates each opinion state
ment providing an estimate of the strength of her current
attitudes toward child-rearing practices. Direct scores
are provided for the at.titud.inal dimensions, two of which
are overprotection and rejection.
Peabody Picture Vocabulary Test (PPVT)
The PPVT was designed by Lloyd M. Dunn (1965) to
provide an estimate of the subject’s verbal intelligence
through measuring his hearing vocabulary. Form A of this
44
instrument was administered in the present study to arrive
at a quick estimate of the student's verbal intelligence.
Standardization of the PPVT was carried out in a geographi
cally circumscribed area, that of Nashville, Tennessee:
only white children were used. Studies of alternate-form
reliability, reported in the manual, were comparable to
those obtained for the standardization sample, ranging
from 0.67 to 0.84. Test-retest reliability studies have
produced mixed results. Thirty-three specific validity
studies using various subpopulations are listed in the
manual, also with mixed results. In those studies correla
ting the PPVT with measures of academic achievement, con
current validity coefficients ranged from 0.04 to 0.91.
In the validity study of P.ikulski (1973) previously cited,
in which he compared three brief measures of intelligence
with the WISC, he found the correlation between the full-
scale WISC and the PPVT to be high (0.74) and significant
beyond the .01 level. Thus, the PPVT was chosen for this
study because of its technical characteristics to measure
one aspect of intelligence with the awareness of its
limited generalizability.
This test does not depend upon verbalization or
reading ability; it uses pictorial representation to
measure recognition vocabulary. The subject sees a series
of plates graduated in difficulty on which there are four
45
line drawings. When the examiner says the corresponding
stimulus word, the subject points to the word which best
describes the stimulus word. Administration time ranges
from 10 to 20 minutes. The raw score obtained is then
converted to a derived score, an intelligence quotient.
Wide Range Achievement Test (WRAT), 1976/ Revised Edition
The WRAT, by J. F. Jastak and S. R. Jastak, is a
general achievement test incorporating a reading subtest
that was used in this study. It is a brief reading test
that provides a rough indication of one component of read
ing achievement, namely, word recognition. Because of the
elementary age level of the students in this study,
Level I (intended for the 5-11 age group) of the reading
subtest was used. The clinical reliability of the reading
subtest reported in the manual for the 5-1.1 age group was
from 0.90 to 0.95 with a SEM ranging from 1.05 to 1.3 9.
Validity of the instrument for the measurement of word
recognition ability was also strong in the standardization
group. Aside from the matter of measuring word recognition
ability was the problem in this study of determining an
appropriate level for the administration of the Gates-
MacGinitie survey test. To serve both purposes, the read
ing subtest of the WRAT was chosen.
Gates-MacGinitie Reading Tests, Second Edition, 1978
The second edition of the reading tests was com
piled by Walter H. MacGinitie, Joyce Kamons, Ruth Kowalski,
!
Ruth K. MacGinitie, and Timothy MacKay. For the second
edition of this survey test, restandardization was carried
out in the school year 1976-77. This edition was field-
tested on both minority and majority youngsters from low,
middle, and high socio-economic backgrounds to obtain new
national norms. Part of the national tryout of items was
to reexamine those items that were particularly difficult
for black students for the possibility of bias. Using the
Kuder-Richardson formula, reliability coefficients were
computed for each grade level of the test. The range for
the vocabulary section is from 0.90 to 0.95, and for the
comprehension section it is 0.88 to 0.93.
j The Gates-MacGinitie tests measure two components
of reaching achievement, vocabulary and comprehension,
which purportedly represent somewhat different abilities.
For the vocabulary test the individual is required to read
and know the meanings of single words— all of the vocabu
lary words are real words. In the comprehension test the
individual is given whole sentences and paragraphs to
read and understand--both literal and inferential questions
are included. Grade equivalent scores were derived from
the student's corresponding raw score for number of items
answered correctly. The use of grade equivalent scores
was based on the fact that it is common practice to report
them for indicating current achievement to parents.
The Interview Schedule
Recognizing the shortcomings of self-administered
forms such as the questionnaire to obtain data regarding
maternal attitudes, beliefs, and expectancies, it was
decided to construct an Interview Schedule for this inves- I
I
tigat ion. Data collect ion was needed for a descr iption of I
the demographic characteristics of the sample and for
assessment of the maternal variables and the socio-economicj
level of each family. Today parents have the right to |
know what the school knows about the abilities and problems;
of their child. Whether or not access to this information j
I
was provided by the school, or to rephrase this statement, j
whether or not the school communicated the information in j
I
an understandable fashion, was a tangential concern of the j
present study.
Improving Interview Method and Questionnaire
Design, by Bradburn and Sudman (1979), was utilized as a
reference source for designing the schedule. The problem
of question threat was of particular concern.
Empirical evidence shows that the impact of question
threat is mediated by several variables, particularly
question structure and question length. (p. 14)
48
In keeping with their recommendations, open-ended questions
were used for threatening topics. In addition, longer
questions with easy-to-understand words were used to
decrease response effects for threatening items. However,
there is no source for researchers which specifies which
kinds of questions are threatening. Horowitz, Shah &
Simmons (1967) suggest the use of the random response
model to reduce response distortion on threatening or per
sonal questions. This technique uses unrelated questions
to overcome the respondent's suspicions and increase
cooperation. Thus, the sequence of questions in the
Interview Schedule was modified so that unrelated and in
nocuous questions would be interspersed between more pro
bing questions. Attempt was made to avoid social class and
racial bias in question structure and in wording.
The format of the Interview Schedule included both
close-ended and open-ended questions. These questions were
designed to obtain data regarding the demographic charac
teristics of the sample and to assess the maternal and
socio-economic variables. The major areas explored were:
1. demographic information regarding the family,
2. preschool information indicative of problems,
3. pertinent educational facts regarding the child,
4. the mother's estimates and expectancies concerning
her child's reading level,
49
5. the mother's expectations for her child's educa
tional and occupational attainment, and
6. the mother's understanding of her child's reading
development.
This semi-structured format allowed for deviation
;from exact wording of the sentences but no addition of or
elimination of items. The mother was informed before the
interview in the instructions that she could decline to
answer any question if she so desired. The individual
questions of the Interview Schedule can be seen in
Appendix B. Copyright registration was obtained for this
study as of October 27, 1981, with the Copyright Office
under Registration No. TXu 81-433. Extensive pretesting o
the Interview Schedule was not possible because of the
difficulty in finding subjects who met the necessary
requirements. Nevertheless, during administration of the
schedule minor alterations were made by expanding some of
the demographic categories,
iAdministration of Instruments
All of the above instruments were administered at
the beginning of the summer and fall sessions to students
!enrolled at the reading clinic. It was necessary to
i
employ both individual and group testing procedures. The
students were given the reading tests by their regular
classroom teacher, which is the common practice of the
clinic. They were given the IQ Test on an individual
basis in a separate room. The mothers were contacted by !
telephone for an appointment before or after their child's
class. Although a large number of the mothers were work- j
ing, none of them refused to come for the interview |
session. During this session, the Interview Schedule was
administered first and then instructions were given to the j
mother for answering the items on the attitude question- 1
naire. Interviewing of the mother and administration of
the IQ Test were conducted by the investigator of the
i
study. i
The Sample |
Subjects were obtained from a university reading !
clinic. The university is located within the central part I
I
of the City of Los Angeles, surrounded by a low-income i
i
population of primarily ethnic groups. Although students i
who attend the clinic come from many parts of the city,
I
the majority of them live near the university. There are
no restrictions placed upon who may enroll. This reading
clinic is available to the public for diagnosis and for
instruction in small classes of six or less students. As a
result, youngsters from both public and private schools
enroll in classes for more individualized instruction than
jthey can obtain from their regular schools.
For purposes of the present study, certain guide
lines were established for the selection of subjects. Both
boys and girls of late childhood, those from 7 to 11 years
of age who had not entered a secondary school, were
included in the sample. Only black children and their
natural mothers were chosen during the selection process.
A further restriction imposed was that the student must
have entered the clinic for the first time— otherwise the
student's mother would have had access to the child's
reading scores from the clinic. Following these guide
lines, it was necessary to draw students from two sessions
to arrive at a sufficient number of subjects, 62 in all or
31 mother-child dyads.
Boundaries of Generalizability
It can be assumed that the population of students
from a reading clinic have different characteristics than
the population of students from a regular day school.
Those students who attend a reading clinic are presumably
there to improve their reading skills. Parents of students
have made commitments to provide transportation for their
child and to pay an enrollment fee. Differences in child
achievement motivation and parent involvement may exist
between the population of a clinic and that of a regular
day school. Therefore, extrapolation of the findings from
this study to other student populations should be made with
extreme circumspection.
The criterion of racial status for subject selec
tion also limits the generalizability of the findings of
this study. Only the correlates of black families with
respect to reading achievement were investigated; those
of nonblack families were not empirically studied. There
fore, generalizations made from the results of this study
are limited to the black population.
i Collection of Data
All of the data of this investigation were derived
from primary sources. The design called for the data to
be collected from students and their mothers. A subgroup
iof students at the reading clinic was identified according
to guidelines heretofore outlined. Mothers of these
students were called by telephone to make a thirty-minute
appointment for a personal interview. Only natural mothers
j were contacted which meant that stepmothers and grand
mothers were excluded. Every mother contacted agreed to
ibe a subject in this research study. Voluntary participa-
1tion on the part of all the mothers meant in effect that
no bias was created due to missing subjects. A summer and
fall session were needed to obtain subjects meeting the
requirements.
The data were collected at the beginning of the
sessions so as not to be influenced by the student's
experiences at the clinic. During the interview the
mother was apprised of the nature of the study to increase
cooperation and reduce suspicions; permission was requested
to administer an IQ test to her child. Uniform instruc
tions were read to the mother pertaining to both of the two
instruments used to measure family characteristics, the
Interview Schedule and the Mother-child Relationship Evalu
ation (a self-administered scale). This approach was
|intended to minimize overreporting of socially desirable
acts and underreporting of socially undesirable acts.
Children of these mothers were individually administered
;the PPVT--in their classrooms they were given two reading
tests, the WRAT and the Gates-MacGinitie.
Each mother-child pair was given a code number so
the data could be subsequently analyzed by family unit.
The instruments were hand scored. On the attitude scale
only half of the items were needed to arrive at a nonac
ceptance score for each mother, using the overprotection
and rejection items. The reporting of occupation on the
54
Interview Schedule was compared with the NORC scale to j
j
|classify each family as either lower class or middle class.j
The ambiguous cases were decided by use of the additional
information regarding family income and mother's educa- |
tional attainment.
i
; j
There were relatively few problems encountered I
during the process of data collection, primarily because ’
of the cooperation of the mothers and the students.
Selective bias due to some subjects refusing to participate!
was not a problem. The mothers generally responded freely j
i
during the interview (one mother needed help in reading
,the attitude scale). The students seemed to like the pic- j
torial nature of the intelligence test, and they displayed |
jample motivation throughout the testing period. j
i Operational Hypotheses
j Two hypotheses concerned with the multivariate j
nature of the relationship between family factors and
reading achievement were empirically tested. They were ‘
formulated as follows: ;
Hypothesis I
j There is a significant multiple cor relat ion
between a maternal cluster of variables and socio-economic
level with reading test scores, controlling for the effects
of verbal intelligence.
i
Hypothesis II
There is a greater correlation between the maternal
cluster with reading test scores than between socio-eco
nomic level with reading test scores, controlling for the
effects of verbal intelligence.
Statistical Analysis
To test the above research hypotheses, various
statistical treatments were administered. An incremental
method of data analysis examining the measured variables,
independent and dependent likewise, was chosen; data re-
duction by formation of variable composites was initiated
when advisable. To implement this method, single variables
were analyzed, then pairs of variables, and finally sets
of variables. A multiple regression model was ultimately
employed, following the transformation of variables to
composites, with a .01 level of significance established.
i
Partial correlation coefficients were necessary to obtain
prior to regressing the dependent variable composite on
the set of independent variables to control for the pre
sumed effects of verbal IQ. More than one multiple
regression approach was used to examine the collective and
56
separate contribution of the independent variables: an ;
overall goodness of fit test and decomposition of explained;
variance by blockwise and stepwise methods of entry. ;
Other statistical treatments were utilized in i
processing the data from the instruments that measured
family characteristics and in analyzing certain variables
and bivariate relationships of special interest. Analysis ;
i
of the instrumentation involved deriving a reliability |
coefficient for the MORE and reporting the responses for j
!
each item on the Interview Schedule. Descriptive statis
tics were employed to present the IQ scores of the students
and the socio-economic indexes of the families. A corre- j
lational method was selected to depict the relationship
between maternal nonacceptance and each of two variables, j
j
verbal IQ and socio-economic level. The variables invol- ;
ving reading, both maternal-related ones and test scores, j
were extensively examined by each subskil.1 and compara- i
tively reported; they were also grouped into composites andj
related to one another, using the technique of scattergrams.
CHAPTER IV
PRESENTATION AND ANALYSIS OF DATA
Three broad strategies of analysis carried out
sequentially were used in this study:
1. data reduction through construction of composites
from sets of variables,
2. examination of bivariate relationships between
variables or variable sets, and
3. various analyses of multivariate relationships
controlling for the influence of one variable.
This procedure, as advocated by Hartwig & Dearing (1979),
followed a course in which single variables were analyzed,
then pairs of variables as relationships were analyzed, andj
|lastly groups of variables were built into a multivariate j
model by this incremental method. Other types of analysis !
i
that were not part of this specific sequence were also j
I
i
used. One such form of analysis was the use of descriptive1
i
statistics which shall be reported as a preliminary to the
j
major analyses. !
58
Preliminary Stage
To describe the location and spread of the verbal
IQ scores obtained from the PPVT, the median and inter- ;
quartile range were chosen in lieu of the mean and standard;
j i
deviation because of outlying values. The median and j
interquartile range, based on the rank order of the values,
; are not as sensitive to extreme values which are present j
in this particular distribution. The median value found !
!
was 96.38, and the interquartile range was from 86.0 to !
100.0. This spread of 14 points is considerably less than ;
i I
! i
i that of a. normal population. A more detailed account of |
! the distribution of this variable is rendered in Table 1
which represents the cumulative frequencies and the cumu- |
i
I
lative percentage frequencies by class interval for the :
; l
j intelligence quotients. j
Table 1
Cumulative Frequencies
and CumulatTve Percentage Frequencies
for the Distribution of Intelligence Quotients
Class
Interval Frequency
Cumulative
Frequency
Cumulative
Percentage
Frequency
140 1
3
1
3
31
30
27
26
100.00
96.77
8 7.10
83.87
110-114
105-109
100-104
59
Table 1 (Continued)
Cumulative Frequencies
and Cumulative Percentage Frequencies
for the Distribution of Intelligence Quotients
Class
Interval Frequency
Cumulative
Frequency
Cumulat ive
Percentage
Frequency
95-99 12 23 74.19
90-94 2 11 35.48
85-89 4 9 29.03
80-84 3 5 16.13
75-79 1 2 6.34
70-74 1 1 3. 23
Evaluation of the two scales (overprotection and
rejection) selected from the MCRE, the measuring instrument
i
; I
; for maternal nonacceptance, was achieved through computa- j
tion of a coefficient of reliability. The alpha coeffici- j
!
ent obtained from the questionnaires was 0.77. In compari-j
1 i
■ I
son the split-half reliability coefficients found in the
I original standardization sample ranged from 0.41 to 0.57. j
i i
i
i
The reliability of this instrument in the specific context |
of the present study was thus supported. Jillings, j
Adamson & Russell (1976) used the MCRE with mothers of
handicapped children (autistic, mongoloid, and learning
jdisabled), and they arrived at alpha coefficients of 0.73
I
: and 0.44 for the overprotection and rejection scales, re
spectively. Therefore, the alpha coefficient estimated
i
for the scales used in this study is greater than that
found in other reported research investigations.
The socio-economic index for each family was ob
tained from information on the Interview Schedule. Occupa-
j tional status of the head of household (father or mother)
was used as the principal criterion for defining the
socio-economic level of the families. In cases where an
joccupation was not on the NORC-NYC list, it was necessary
I
\ for the investigator to infer the prestige standing of the
occupation. Tabulation of the assignment of ratings to
t occupational categories resulted in the following sums of
families in each category.
Code Number Category Sum
(0) Welfare Recipients 4
(1)
Laborer s 1
(2) Service 1
(3) Operatives 6
(4) Craftsmen 5
(5) Clerical and Sales 12
(6) Manager ial 0
(’ 7) Professional 2
The category of Clerical and Sales occupations
clearly represents the modal occupational group for the !
jsubjects in this study. It should be noted that in four
family constellations where the family was intact and where j
both the father and mother were employed, the father's
I
loccupation was used even though it fell in a lower category
than that of the mother. Because of this problem and
!
because of the possibility of unreliability of response on
I
|a single measure, redundancy of information for the socio-
i
economic index was obtained through the use of two
additional measures— educational attainment of parents and
family income. The breakdown for educational attainment by
Bachman's categories (1970) and for family income is as
jf ollows:
i
Educational
Attainment
| Less than
| High School
Some High School
Completed
High School
' Some College or
Trade School
Completed College
Unknown
Mother Father
1 3
3 3
13 8
11 10
3 2
6
Family Income in Dollars
Income Number
Less than $5,000 8
$5,000 - $10,000 8
$10,000 - $15,000 7
$15,000 - $20,000 2
$20,000 - $25,000 2
More than $25,000 4
The use of education and income along with occupa
tion as the principal criterion to establish a socio
economic level for each family group resulted in a total
of 17 lower-class families and 14 middle-class families.
Although reported income levels fell primarily in the
lower three categories representing family incomes of less
than $15,000, the educational attainment of parents indica
ted that many of these parents had received some post
secondary instruction which had a bearing on the family's
rating for socio-economic level.
Stage I
The first broad analysis strategy could be de
scribed as a process of data reduction. Clusters of highly
intercorrelated variables of a kindred nature were sought.
Composite variables were calculated by finding the
arithmetic mean of a subject's responses or test scores
63
which measured a common characteristic. This technique
of building composite variables was utilized if concomitant
variation was revealed through bivariate correlation
|analysis.
Measures of the
i
behavior— word recogniti
were targeted for analys
I
bles were chosen:
1. mother's estimat
and comprehension, and
2. mother's expecta
lary, and comprehension.
The set of reading test
i vocabulary, and comprehe
:variables was also earma
word recognition, vocabulary, and comprehension components
of the three sets of variables are displayed in Appendix C.
JPearson product-moment correlation coefficients were com
puted for pairs of variables in each set, resulting in
very high zero-order correlations. Matrices of the non-
redundant coefficients are presented in Table 2 for
smother's estimates, in Table 3 for mother's expectancies,
and .in Table 4 for reading test scores. Mean values and
standard deviations for each variable are listed in
Table 5.
64
three components of reading
on, vocabulary and comprehension—
is. Two sets of independent varia-
e of word recognition, vocabulary,
ncy for word recognition, vocabu-
scores for word recognition,
nsion representing the dependent
rked. The raw data representing
Based on these results, it was decided to construct
a composite for each set of variables. Transformation of
the variables was effected by computation of the arithmetic
! mean of the three reading component values for each subject
with each set. See Appendix D for a listing of the
composite values attributed to each mother-child dyad for
the three sets of variables. Labelling and coding of the
newly created variables, which purportedly represent a
common underlying dimension, was established as follows:
Composite of Mother's Estimates EST110
Composite of Mother's Expectancies EXP120
Composite of Reading Test Scores SCR710.
The problem of differentiation of the skills and
'abilities that comprise reading achievement has been con-
i sidered by MacGinitie (1973), one of the authors of the
Gates-MacGinitie Reading Tests used in this study. In a
!
'discussion of vocabulary and comprehension subtests on
standardized reading achievement tests, he states that he
has found a high correlation between subskill scores,
especially the more advanced skills.
65
Table 2
Intercorrelations of Mother's Estimates
for Word Recognition, Vocabulary
and Comprehension
WRlll V0CAB112 COMP113
WRlll
VOCAB112
COMP113
1.00 0.83 0.82
1.00 0.80
1.00
Table 3
Intercorrelations of Mother's Expectancies
for Word Recognition, Vocabulary
and Comprehension
WR12.1 VOCAB122 COMP123
WR121
VOCAB122
COMP123
1.00 0. 94
1.00
0.93
0.95
1.00
Table 4
Intercorrelations of
Reading Test Scores for
Word Recognition, Vocabulary and Comprehension
WR711 V0CAB712 COMP713
WR711 1.00 0.91 0.77
VOCAB712 1.0 0 0.8 4
COMP713 1.00
Table 5
Means and Standard Deviations of
Word Recognition, Vocabulary and Comprehension
for Mother's Estimates, Mother's Expectancies and
Reading Test Scores
Var iable
Mother's Estimates
WR111
V0CAB112
C0MP113
Mother's Expectancies
WR121
VOCAB12 2
COMP123
Reading Test Scores
WR711
VOCAB712
COMP713
Standard
Mean Deviation
3.44 1.46
3.98 1.41
3.71 1.61
5.47 1.66
5.50 1.73
5.31 1.76
3.65 1.41
2.81 1.58
2.54 1.23
67
Corroboration of his finding was given by this
study which produced an r_ = 0.84 for vocabulary and compre
hension test scores. MacGinitie expressed the following
view:
Correlations between the vocabulary subtest and the
comprehension subtest generally approach the relia
bility of the individual subtest. There is still
room for the two subtests to be measuring somewhat
different achievements, but for individual pupils
the difference between the vocabulary score and
the comprehension score must generally be very large
before we can put much faith in this difference
actually reflecting a true difference in achieve
ment in the two areas. (p. 40)
As reported in the manual, the reliability
coefficients for the Gates-MacGinitie Reading Tests range
from 0.90 to 0.95 for the vocabulary section and 0.88 to
0.94 for the comprehension section. The product-moment
correlation for vocabulary and comprehension scores arrived
at in this study comes near these ranges.
Stage II
The second strategy consisted of analysis of sub
sets of variables to discover important relationships
among the variables. (As used hereafter "variable" can
refer to a single variable or a composite variable.) To
this end, several statistical techniques of association
were employed such as crosstabulation and correlational,
procedures. i
68
Maternal nonacceptance was correlated with child's
verbal IQ, resulting in a negative correlation, of -0.27.
This maternal variable was also analyzed with respect to
the socio-economic index. A contingency table represen
ting the quartile intervals of the MCRE and the two socio
economic levels can be seen in Table 6 along with row and
column totals and percentages. A dramatic difference
between lower-class and middle-class mother's acceptance
attitudes is evident. If one were to look at just half of
the table, columns 3 and 4, the combined categories would
give a percentage of 70.6 lower-class mothers high in the
nonacceptance attitude whereas only 14.3 percent of the
middle-class mothers had scores which fell in the upper
half of the percentile rankings.
Table 6
Crosstabulation of Maternal
Nonacceptance Scores and Socio-economic Index
(from lower to higher scale scores)
Row
1 2 3 4 Total
Lower Class 2 3 7 5 17
54 . 8
Middle Class 4 8 2 0 14
45.2
6 1 1 9
19.4 35.5 29.0
5
16.1
31
100.0
69
Other subsets of variables receiving intensive
analysis were those relating to reading— mother's estimates,
mother’s expectancies, and reading test scores. Difference
scores for each subskill were computed between mother's
estimates of reading grade level and child's test scores
and between mother's expectancies regarding child's read
ing grade level and actual test scores. The means and j
I
ranges for these difference scores are shown in Table 7. ;
A new block of coded variables was created to represent
these difference scores. Reference is made to Appendix F !
for the complete coding scheme of all variables and
i
variable composites. j
Special note should be taken of the high degree of j
accuracy with which mothers could estimate their child's
grade level for word recognition skills (mean value of
-0.22). However, there was almost a three-year discre
pancy (mean value of 2.77) between the child's test score
in comprehension and the grade level at which the mother
believed her child could be reading in this area. The
range column of difference scores displays considerable
variation in each subskill for both estimates and
expectancies.
70
Table 7
Mother's
Estimates
Mother's
Expectancies
Means and Range Values of Difference Scores
Calculated from Maternal Variables and Reading Test Scores
Initial Difference
Variables Variable Mean Range
WR111 - WR711 WR811 -0.22 4.00
V0CAB112 - VOCAB712 VOCAB812 1.17 6.20
C0MP113 - COMP713 C0MP813 1.17 4.10
WR121 WR711 WR821 1.82 6.40
VOCAB122 - VOCAB712 V0CAB822 2.69 6.30
COMP123 - COMP713 C0MP823 2.77 5.50
, To more fully understand the bivariate relationships!
i
between the reading composite variables, scattergrams were ;
plotted. One scattergram related mother's estimates to
reading achievement; another scattergram related mother's
expectancies to reading achievement. The two scattergrams j
thus showed the pairings of two maternal composite variables!
with the dependent variable. Of interest was the general
pattern represented by the data points and the graphic lo
cation of the bulk of the data. As expected, a positive I
i
;linear relationship was evident in both graphs. More j
I
informative was the clustering of points, for it was dis
covered that in the scattergram employing the variable i
i
mother’s estimates 18 cases (over half of the points) fell j
within the third quadrant, or lower-left quadrant. In the j
other scattergram using the variable mother's expectancies
only 9 cases fell within the same quadrant. The means for
the two maternal variable composites were considerably :
!apart from one another, 3.71 for mother’s estimate and 5.42;
for mother's expectancy. In contrast, the standard devia
tions showed about the same degree of dispersion, 1.40 and
1.68, respectively.
Stage III j
The third of the broad analysis strategies was
directed to the task of examining the multiple relation-
!
ships of the set of independent variables with the dependentj
variable, which at this stage was a composite variable. !
The design of the study called for a partialling out of !
jverbal IQ before the major analyses were performed; this j
necessitated that partial correlations be computed. j
i
Another preliminary step to the regression analyses, |
which seemed to be indicated by the nature of the maternal
variables, was to determine if a high degree of multicol- j
i
linearity existed among these variables. The regression |
solutions were accomplished through use of blockwise and j
i
stepwise procedures to evaluate the separate and combined
I
contributions of the independent variables within a
tnul t ivar iate context.
To statistically control the influence of verbal '
IQ, partial-correlation coefficients were calculated. As j
i i
ja result, new independent and dependent variables were ■
constructed which reflect the adjustment for the effects of
the controlling variable. The complete correlation matrix :
can be seen in Table 8; this matrix was used as an input in '
all subsequent regression problems. As shown in the Table,
73
Table 8
Partial Correlation Coefficients of all
Variables with Verbal Intelligence Used as a Control Variable
SCR710 N0NAC100 EST110 EXP120 SES200
SCR710 1.00
0
O
1
0.80 0.63 0.05
N0NAC100
i
o
i —*
o
1.00 -0.06 -0.07 -0.53
EST110 0.80 -0.06 1.00 0.80 0.14
EXP120 0.63 -0.07 0.80 1.00 0.13
SES200 0.05 -0.53 0.14 0.13 1.00
ithe dependent variable (coded as SCR710) and the four
: j
independent variables are represented with the zero-order j
partials indicated for each pair. The column for maternal [
i
; j
jnonacceptance has values, all of which are in a negative j
! |
^direction; there is a substantial negative correlation |
1 i
!(-0.53) between this variable and the socio-economic index.I
!
The other consideration regarding the regression j
I
analyses was the possible problem of multicollinearity of !
! i
|the independent variables. Some of the independent j
j
jvariables in this study, namely the three maternal variables
I
I were suspected of being moderately correlated. Therefore,
i
j the intercorrelations of these three variables— maternal
j
j nonacceptance, mother’s estimates, and mother’s expectan- j
cies— were computed. The nonredundant correlation I
; |
coefficients can be seen in Table 9. The results indicate |
I
that the only substantial correlation, that of a Pearson j
j i
' £ of 0.80, is between one pair of variables— mother's j
I j
jestimates and mother's expectancies. Since a bivariate j
icorrelation of this magnitude was rather high but since
! it did not violate the assumption of the multiple regres- ;
! sion model requiring the absence of perfect multicol-
'linearity (Lewis-Beck, 1980), it was decided to put both
I ;
| independent variables into the regression equations. !
75]
{ To analyze the collective and separate contribu-
tions of the independent variables, which in effect tested
the research hypotheses of the investigation, three
!multiple regression procedures were employed:
I
| 1. an overall F test for goodness of fit of the re-
I
gression equation,
2. a blockwise technique with a predetermined order
of entry for the variables, and
3. a stepwise technique in which the variables were
jentered one-by-one according to their contribution to
explained variance.
i Table 9
|
j Nonredunant Product-moment Correlations of
Maternal Variables to Determine Multicollinearity
EXP12Q
-0.07
0.80
1.00
NONAC10 0 EST110
NONAC10 0 1.0 0 -0.0 6
EST110 1.00
EXP120
76
The overall test for the null hypothesis that the
multiple correlation is zero for the set of independent
variables together with the dependent variable, controlling
for verbal intelligence, produced an F ratio of 12.30. The
summary statistics are presented in Table 10. Comparing
this F value to the critical F value for 4 and 26 degrees
of freedom, we find that it is significant at the .01
level. Thus Hypothesis I, relating the set of family
variables in combination to reading achievement, was
supported.
Table 10
Summary Statistics for the
Overall F Test
Multiple R 0.81
R Square 0.65
Adjusted R Square 0.60
Standard Error 0.84
Analysis of Sum of Mean
Var i ance DF Squares Square F
Regression 4 34.72 8.68 12.30
Residual 26 18.35 0.71
77
The standard error of estimate found in this test
was 0.84, which is considerably smaller than the standard
i
deviation of the dependent variable, or 1.33. Kerlinger
Pedhazur (1973) state that if the standard error of esti
mate is relatively large when compared to the standard
j
deviation of Y, the estimate of Y on the basis of the X's
is poor. Thus, the prediction of the reading achievement
composite based on this set of independent variables is
expected to be relatively successful.
! Also shown in Table 10 is a statistic called
adjusted R square. This statistic, in conjunction with
the conventional R square, is part of the output from the
SPSS multiple regression subprogram. In a footnote,
j Kim & Kohout (1975) provide the following explanation of
adjusted R square.
Adjusted R2 is an r2 statistic adjusted for the
number of independent variables in the equation
and the number of cases. It is a more conserva
tive estimate of the percent of variance explained,
especially when the sample size is small. (p. 358)
Because of the number of subjects in this study, it was
deemed advisable to use adjusted R square in .lieu of
R square as a more appropriate statistic to report.
To decompose the explained variance, the three
maternal variables were entered as a block after socio
economic index so that the total influence of the index
would be reflected. The rationale for blocking the
maternal variables was one of conceptual similarity. The
first variable, socio-economic index, to enter the predic
tion equation uniquely accounted for .2 percent of varia
tion; 60 percent of explanatory power was provided by the
maternal block. Clearly, the contribution of the socio
economic index was practically nil.
Entering the same four variables one at a time for
a stepwise solution to find their incremental contribu
tions to explained variance resulted in the following
order of entry: mother's estimates, socio-economic index,
and mother's expectancies, was not chosen for entry
because it did not meet the statistical criteria. The
regression coefficient (-0.03) for this variable was
unexpectedly small relative to the coefficient for
mother's estimates, and it was statistically insignificant
(F = 0.02). The body of statistics for this stepwise
inclusion approach is located in Table 11, which displays
the added amount of explained variation made by each suc
cessive independent variable. The variable that was
entered first, explaining the greatest amount of variance
in reading achievement, was mother's estimates; socio
economic index was entered second.
Table 11
Summary of Statistics for the
Stepwise Solution
R Square
Variable Multiple R R Square Change Simple R* b** F ratio
EST110 0.80 0.64 0.640 0.80 0.770 51.56
SES200 0.80 0.64 0.004 0.05 -0.329 25.32
NONAC100 0.81 0.65 0.010 -0.10 -0.175 17.01
*Zero-order correlation between the dependent variable and each of the independent variables
**Unstandardized regression coefficients
co
o
The findings from both of these methods used to
decompose the total explained variance supported
Hypothesis II. The contribution of the maternal block of
variables to explained variance was significantly greater
than the amount of variance explained by the socio-economic
index.
Findings from the Interview Schedule
The bulk of the information obtained from inter
viewing the mothers has heretofore been reported as repre
sentations of the variables socio-economic index, mother's
estimates, and mother's expectancies. Four questions on
the Interview Schedule pertained to developing a socio
economic index for each family unit. Three questions
ascertained the mother's estimates and expectancies of
her child's grade level of reading. Information from the
other pertinent questions shall be presented. Data from
these questions placed as fillers between the more probing
questions shall not be reported unless special significance
was noted. See Appendix B for a copy of the Interview
Schedule.
Demographic statistics obtained from Questions 1
and 2--not as subject to change over a period of time as
maternal attitudes, beliefs, and expectancies— are the
following:
81
Number of Children
in Family Frequency
1 8
2 12
3 6
4 1
5 1
more than 5 3
Birth Order of
______Child
Firstborn 12
Middle born 4
Last born 7
Only child 8
Marital Status of
Mother _
Single 6
Married 14
Separated 3
Divorced 8
82
A group of items (Questions 7, 8, 9, 10, and 11)
dealt with important aspects of the child’s preschool years.,
There were 8 mothers who indicated that there was a problem!
! !
jduring the preschool years in their child's eating, sleep-
iing, or toilet training habits; 7 of these mothers reported;
j i
j :
problems in toilet training. More of the mothers described!
their preschooler as very active than as moderately active; j
the numerical breakdown for the trichotonious response
categories follows:
very active 17
moderately active 14
inactive 0.
| was
I
I
! adj
; mar
1
wer
the
add
chi
a r
: chi
had
Almost none of the mothers said that their child
difficult to raise during the preschool years. The
ectives they selected to describe the child were pri-
ily of a positive nature.
Several questions (Questions 12, 13, 14, and 15)
e designed to obtain information as to whether or not
child had received special educational services in
ition to instruction in a regular classroom. Eighteen
Idren had received special help in reading at school in
eading lab and/or from a reading specialist. Only one
Id was in a special education program; three children
had speech therapy; one had been in a special motor
83
skills program. Nine children had previously attended
another reading clinic where they obtained both diagnosis
land instruction. Another finding from this group of
questions was that nine children had been retained at some
grade level and had not been normally promoted to the next
Igrade level.
Two additional questions (Questions 21 and 22)
were asked regarding long-range expectations about occupa-
|
tional choice and educational attainment for the child.
jMost of the respondents had difficulty in specifying an
occupational category. Ten mothers believed that their
child would complete high school, while 21 said their
child would go to college. These findings can be compared
to those of Coleman (1966) who points out that there is
"lesser concreteness in blacks' aspirations" (p.279),
greater hopes but lesser plans.
The last two open-ended questions (Questions 24 and
s25) were directed toward ferreting out the mother's belief
regarding why her child is not reading at a higher grade
i
level. This subjective report is at once a consequence of
the child's earlier educational experiences and other fac
tors and an influence in later educational performance.
The answers fell into four broad categories with the
following frequencies:
Family Factors:
Maternal: insufficient amount
of time spent with child 6
Family Problems 2
Student Factors:
Physiological Problems 2
Short Attention Span 4
Lack of Motivation 6
Educational Factors:
Lack of Basics 2
Poor Teaching 4
School System 2
No Idea, Not Pertinent: 3
! Sources of Error
i
Efforts were made to keep the errors at a low level
at each phase of data collection and data analysis. Errors
resulting from the Interview Schedule included possible
jmisrepresentation of facts by respondents; therefore, each
schedule was reviewed for inconsistencies on the basis of
'other information on the form. The MCRE was checked for
omissions immediately after being completed by the mother.
i
Scoring of the three tests given to the child---the PPVT,
the WRAT, and the Gates-MacGinitie— was done by hand with
oareful attention to accuracy which, nevertheless, could
have entailed some clerical errors. Although the mathe
matical computations were performed by electronic computers
the raw data and matrix inputs were verified. Thus, check
■measures were utilized throughout the data-handling
operation.
Summary of Findings
! In general three broad analysis strategies were
employed to arrive at an understanding of the multivariate
irelationships. In addition, descriptive analyses were
lemployed to depict how the respondents were distributed
ialong single variables and how they answered questions on
the Interview Schedule. The major findings are summarized
as follows:
Preliminary Stage
1. Almost 75 percent of the students had a verbal IQ
!
of 99 or less, indicating a positively skewed distribution.
2. The reliability coefficient found for the MCRE
instrument was 0.77, a higher coefficient than that
obtained in other known research studies.
3. For the socio-economic index, the families were
almost equally distributed. There were 17 lower-class
families and 14 middle-class families.
86
Stage I
1. For mother's estimates, the Pearson r's were 0.83
(word recognition and vocabulary), 0.82 (word recognition
and comprehension), and 0.80 (vocabulary and comprehension)
2. For mother's expectancies, the Pearson r_'s were
0.94 (word recognition and vocabulary), 0.93 (word recogni
tion and comprehension), and 0.95 (vocabulary and compre
hension) .
' 3. For reading achievement, the Pearson £ 1s were 0.91
(word recognition and vocabulary), 0.77 (word recognition
and comprehension), and 0.84 (vocabulary and comprehension)
Stage II
1. A small negative correlation of -0.27 was found
between maternal nonacceptance and child’s verbal IQ.
2. It was found that 70.6 percent of lower-class
mothers fell within the two upper intervals in maternal
nonacceptance in comparison with the standardized popula
tion, whereas 14.3 percent of middle-class mothers fell
within these same intervals.
3. The mean of difference scores between mother's
estimates of grade level of word recognition and
children's test scores in word recognition was -0.22, very
close to a perfect prediction.
4. The mean of difference scores between mother's
expectancies of grade level of comprehension and children's
test scores in comprehension was 2.11, almost a three-year
spread.
5. The variables mother's estimates and mother's
iexpectancies were both linearly related to the dependent
variable, reading achievement,
6. The mean for the mother's estimate composite was
, 3.71, and the mean for the mother's expectancy composite
I was 5.42. Generally the mothers expected their children to
be reading at a level almost two years higher than they
were actually reading.
Stage III
j 1. All of the partial zero-order correlations between
maternal nonacceptance and the other variables, adjusting
for verbal IQ, were in a negative direction. The largest
Pearson r_ or -0.53 was for the association between maternal
; nonacceptance and socio-economic index.
2. Testing for multicollinearity in the maternal
block of variables, one substantial correlation was found.
A Pearson product-moment correlation of 0,80 was the calcu
lated value of the relationship between mother's estimates
and mother's expectancies. '
3. The overall test of multiple correlation produced
an F ratio of 12.30, significant at the .01 level of
881
probability. The standard error of estimate found was 0.8
considerably smaller than the standard deviation of 1.33
found for the dependent variable. The linear regression
of this variable on the indenendent variables together,
controlling for verbal IQ, resulted in a multiple R of
0.81 and an adjusted of 0.60.
4. Decomposition of the total explained variance
resulted in 60 percent accounted for by the maternal block
when it was entered into the prediction equation after
socio-economic index, the contribution of which was
negligible.
5. The findings of the forward stepwise approach
were as follows:
mother's estimates accounted for 0.64 percent,
socio-economic index added 0.004 percent, and
maternal nonacceptance added another 0.01 percent
to explained variance.
The variable mother's expectancies was never
entered into the equation.
Interview Schedule
1. The modal number of children per family was two.
2. The modal birth order position of the child was
that of first born.
3. Seventeen of the 31 mothers were either single,
separated or divorced.
4. The most prevalent problem reported during the
child’s preschool years was that of toilet training.
5. Over half of the children were categorized as very
active as preschoolers.
6. Attending a reading lab and/or seeing a reading
specialist in school was reported for 18 of the children.
7. Almost one third of the students had previously
attended another reading clinic.
8. Grade level retention had been experienced by 9
of the students.
9. Occupational plans for the child, as a rule, had
not been crystallized by the family.
10. The majority of mothers, 21 in actual number,
expected that their child would go to college in the future.
11. Although a variety of reasons were given to
explain the child's present reading level, the two most
common responses were: the mother had not spent enough
time with the child, and the child did not have sufficient
motivation in learning to read.
9 0
CHAPTER V
SUMMARY, CONCLUSIONS
AND RECOMMENDATIONS
Summary
The study in the broadest sense was an exploration
of the relationship between the family environment and
children's academic achievement. From a host of possible
j
family environment variables, a small number, presumed to
be associated with reading achievement, were selected for
intensive analysis. Special emphasis was placed on the
role of maternal factors. Even though the investigation
was undertaken on a modest scale, it attempted to address
some of the crucial problems in educational research per
taining to family effects. A recapitulation of the main
i
ideas of each of the preceding chapters will be given
before discussing conclusions derived from the study and
possible recommendations.
The Problem
The basic problem of the study was that of identi
fication of specific family factors which are significantly
91
jrelated to child outcomes. It was the purpose of the study !
I
i
■to examine certain maternal factors and the family socio- j
economic level as they relate to children's reading j
achievement. The three major questions posed were:
1. Which family factors can be named as having a |
I
t
dominant effect on reading achievement? i
I
I
2. How strong is the relationship of both maternal '
factors and socio-economic level on reading achievement?
j
3. Is there a stronger relationship between maternal j
factors and reading achievement than between socio- I
|
economic level and reading achievement? j
Consideration was also given in the study to minor |
questions of particular interest. How accurate is the j
|
mother's knowledge of her child's reading ability? Is
there a marked discrepancy between the mother’s estimate
and her expectancy of the child's grade level of reading? |
I
Does the variable maternal nonacceptance interact with i
I |
ithe variable socio-economic level in a significant way? j
Because ability has been regarded traditionally as <
a major predictor of educational outcomes, the possible
confounding effects of this variable were deemed important
to control. The delimitation of the population to one
minority group was another concern. The generalizability
of the study was limited since the relationships examined
may not hold for other segments of the population.
The Literature
The main aspects of the interactional model were
joutlined which hinge on the premise that there is an
I
jinterdependency between personal and situational factors
jin eliciting behavior. Not only is the person's behavior
influenced by the situation; the individual chooses the
situation in which to perform, cognitively ascribes his/
;her own meaning to the situation, and subsequently changes
I
the character of the situation.
Multiple factor designs were found to be advocated
’in much of the literature for research on educational
achievement, beginning with the Chicago School which
pioneered the measurement of family environments and con
tinuing up to the present time with the input-output
^approach of analyzing educational outcomes in terms of a
set of interacting inputs.
! Large-scale and small-scale research studies,
i
1
| investigating the family's impact on children's educational
achievement, were reviewed. Of special interest in these
studies was the selection of varying types of family fac
tors (structural, attitudinal and process variables). The
importance of family characteristics as determinants of
academic achievement was supported throughout the studies,
those specifically designed to investigate reading
behavior as well.
Study of the family was found to have been con
centrated on the mother-child dyad; historically, terms
such as maternal rejection and overprotection were
developed to conceptualize parental patterns of behavior.
The black family, in particular, was revealed as having a
strong achievement orientation and a continuing family
support system. The social pathological model, used in
the past in research and compensatory programs for black
youth, was described and contrasted with the ecological
approach, a culture-specific perspective.
The Methodology
Percent of explained variance was chosen as the
principal measure of association in this study. Since
the student was the unit of analysis, the variance to be
explained was the total variance among students. To
analyze the collective and separate contributions of the
independent variables, the technique of multiple regression
was employed.
The set of independent variables was represented
by family characteristics; the set of dependent variables,
basic reading skills. The data collection procedures for
these variables, which were carried out at a university
reading clinic, included a personal interview and an
attitude questionnaire involving the mother and a battery
of reading achievement and IQ tests given to the child.
94
Regarding the psychometric characteristics of these scales
land tests, it was found that validity information was
i
limited and that primarily whites had been chosen for the
standardization populations. In the sample for this study,
only black students attending the clinic and their mothers
were used as subjects. The students ranged in age from
7 to 11 years; all of the mothers of these children con
sented to be subjects.
Two operational hypotheses were formulated to
explain variance in the dependent variable as follows:
1* Hypothesis I stated that there would be a signifi
cant relationship between the set of family variables
together and reading achievement, controlling for verbal
IQ.
2. Hypothesis II stated that the proportion of
variance explained by the maternal block of variables
would be greater than that contributed by the sole variable
'socio-economic level, controlling for verbal IQ.
Testing of these hypotheses was conducted through
a sequence of stages in which the statistical analysis
followed an incremental method starting with univariate
examination and proceeding to bivariate and multivariate
relationships. The techniques used included a variety of
descriptive statistics, Pearson product-moment correlations,
95
a partial correlation to control for verbal IQ, and
i
multiple regression analyses employing blockwise and
stepwise entry methods. The more complex statistical
; t
j
procedures were calculated by electronic computer using
SPSS subprograms.
; Selected Findings
Analysis of the research data proceeded through
| ' !
; three broad stages in addition to intensive examination
I
: I
of certain variables and instruments. The following
: selected findings were produced from this analysis. j
1. Preliminary Stage. The reliability coefficient j
; computed for the MCRE was 0.77. Seventy-five percent of
the students scored at 99 or below on the PPVT. Near j
!
equal distribution on the socio-economic index for lower I
: I
and middle-class categories was found.
Stage I. High intercorrelations between word j
nition, vocabulary and comprehension were found for
r's estimates, mother's expectancies, and reading j
scores. As a result, composite scores were computed
ach of the variable sets, using the mean value to
sent the three reading subskills.
Stage II. A Pearson product-moment correlation
.27 was found between maternal nonacceptance and the
scores; lower-class mothers were high in nonacceptancei
middle-class mothers were low. Estimates of word !
2
recog
, mothe
test
for e
repre
3
: of -0
PPVT
! while
recognition ability were very accurate. Expectancies for
comprehension were almost three years above actual reading
scores.
4. Stage III. Partial correlations controlling for
verbal IQ revealed negative values between maternal non-
acceptance and all of the other variables. In the test
for multicollinearity of the maternal variables, a Pearson,
i
_r of 0.80 was found for the relationship between mother's j
j
estimates and mother's expectancies. The overall F test >
was significant at a probability level of .01 for the
linear regression of the dependent variable composite on j
all of the independent variables taken together. Adjusted
2
R was 0.60, viz 60 percent of total variance was accoun
ted for by the set of independent variables. Decomposi
tion of the explained variance showed that the maternal
block accounted for most of the explained variance; the j
I
contribution of socio-economic level was negligible. In
the stepwise analysis, the variable mother's estimates was
, entered first, accounting for most of the explained vari
ance.
5. Interview Schedule. Some important findings
i
: derived from interviewing the mothers were: (a) over
half of the mothers were either single, separated, or
divorced, (b) over half of the students had received
special help in reading at their regular school,
(c) nine of the students had been retained at some grade
level, and (d) the ratio of mothers that expected their
child to go only to the high school level and those that
expected their child to go on to college was 1.2.
| Conclusions
The general pattern of findings from the empirical
study and the implications that can be derived thereof
.offer some justification for the contention that mother's
beliefs and attitudes are significantly related to
i
children's reading performance, and that such maternal
factors contribute more to explained variance than the
oft-used socio-economic index. That mothers are important
[determiners of their children's ability to perform academi
cally is a premise accepted by many educators. But pre
cisely how mothers affect specific areas of learning, for
example reading, has been surmised but not fully under
stood. To help fill this lacuna, the argument was pre
sented that certain maternal factors (which were carefully
chosen to represent dominant attitudes, beliefs, and ex
pectancies regarding the child) are directly related to
the reading level attained by the child, and evidence was
gathered to test this contention. Moreover, the
98
differential contributions of maternal factors and socio
economic level were compared with the latter proving to
be a poor predictor when combined with maternal factors.
Identification of meaningful and alterable family
variables, other than static characteristics of parents
I (Bloom, 1980) , was a major task of the study. Although
the literature was scanned in both theoretical and empiri-
: cal domains to aid in this identification, it was dis-
!
; covered that a well-articulated theoretical structure
incorporating educational constructs had not yet been
developed. Thus interaction theory, supplemented by con
structs from the Chicago School of family environment
research and the child guidance movement, served as a
working model for the questions of this research study.
Shortening of the initial variable list by con
struction of composite variables was undertaken because
strong interdependence was found among the variables
(independent and dependent) involving reading. This
pruning operation indicated in effect that reading sub
skills assessed on standardized reading tests are so
highly intercorrelated that for research purposes the dis
tinction may not have great utility.
Within the maternal block of variables was one,
maternal nonacceptance, which was not a strong contributor
99
to the variance in reading achievement when combined with
the particular set of maternal variables in this research i
i
!
design. The fact, of course, that it was not a main !
source of variation is, in itself, important information
for future research. The high reliability coefficient j
i
obtained for the instrument measuring this variable was j
an indication that the MCRE seems to be a suitable instru- j
ment for this type of population--black mothers who, for I
the most part, have graduated from high school or have had i
some postsecondary education. If this instrument has j
i
equal reliability for a less educated group of black i
mothers is a moot point that was not addressed within the '
boundaries of this research design.
When maternal nonacceptance was related to other j
independent variables, some noteworthy findings were
revealed. In harmony with the literature, maternal non- j
i
i
acceptance was negatively correlated with verbal IQ,
though the correlation was small. The unintended conse
quence of this maternal attitude is that it may prove to
be harmful for the educational attainment of less able
students. The other relationship that merits attention
is that between maternal nonacceptance and socio-economic ;
level. Lower-class mothers were high in maternal nonac
ceptance and middle-class mothers, low. Does this repre
sent the fact that one group, the more educated and |
esteemed group, responds in a more "socially desirable" !
i
manner? The question remains unanswered. A final comment j
regarding this variable is that maternal attitudes affec- !
ting the child may shift as the age of the child changes. |
The results from this cross-sectional design, therefore, j
may not be the same as those obtained from a longitudinal
study. j
i
The other two maternal variables, mother's esti- j
mates and mother’s expectancies, were closely related |
phenomena. Designating these measures as independent j
variables is debatable since they are in some sense an !
1
r
i
outcome of the educational history of the child— these j
variables may also be affected by the age of the child. ;
!
Nevertheless, employment of mother's estimates and mother'sj
i
expectancies as independent variables has value for educa- j
tional research because they are, in part, subject to !
j
school manipulation over time whereas background family j
variables are not. Owing to the possibility that verbal j
IQ had biased the estimates and expectancies, an !
I
adjustment process was instituted in the form of a partial '
correlation to correct for this factor. High maternal
expectancies were found likewise for student reading
grade level and for ultimate educational attainment of
I
the child. The wide discrepancy between mother's j
101 ,
estimate of reading comprehension and the child's test >
score in comprehension is cause for concern since it can
, be asked if these estimates represent a denial of the ;
: child's learning problems. Parents have the right to know |
about the abilities, the performance level, and the
! I
|
1 problems of their child at school. This study indicated
that maternal understanding of the child's academic abili
ties and problems was lacking in some degree. j
The three maternal variables discussed above were
selected because they were presumed to have an impact on ;
reading achievement for this population subgroup. However,
there were idiosyncracies in the sample which might limit |
wide generalization of the results. (Characteristics of '
the sample were reported in Chapter IV from answers to |
the questions appearing on the Interview Schedule.) There j
are also situational differences that invariably exist j
from one clinic to another and cultural differences that
appear from community-to-community. For these reasons,
caution should be exercised in extrapolation from the
data of this sample to differing subpopulations until there;
is corroborating evidence from other research studies.
Improper attribution of cause should not be made
regarding the block of maternal variables. Survey data
such as that obtained in this nonexperimental study
inherently are not amenable to a cause-and-effect
102 i
inference. The conclusions reached come from a multiple
regression model, and therefore they pertain only to the
set of independent variables inserted into the regression
equation; a different collection of maternal variables
predictably would account for more or less explained vari
ance in reading achievement. It must also be remembered
that the unique contribution found for each of the inde
pendent variables should be interpreted as a function of
the particular set of variables that is included in this
regression analysis.
From the results of the study it can be concluded
within the framework of interactional psychology, that
the student's reading achievement is related to the per
vasive and continuing effects of his/her interaction with
his/her mother. This interaction, and the cognitive
factors which emerge from it, shape the individual's
capacities and performance in reading and probably in
other academic areas as well.
An awareness of the dominant maternal factors
affecting reading performance will give us a much clearer
perspective from which to develop educational policy in
volving the parent. This study has relevance for other
researchers who are also attempting to identify those
few maternal variables among the network of candidate
variables that are important sources of reading variability.
i
Because these variables operate in an interrelated, over
lapping fashion, it seems likely that only a select few
are needed to predict reading achievement. Perhaps, there
fore, our search will not be long or arduous.
Recommendat ions
The findings of the study and the conclusions
drawn therefrom led to the following recommendations for
educational research and educational practice.
1. Since there is still a need for theories and/or
models employing family constructs to account for
children’s academic achievement, it is recommended that
contributions be made to this area of study.
2. To identify and examine the dominant maternal
variables affecting student reading achievement, it is
irecommended that there be controlled inquiry investigating
!
■this domain of variables.
3. To further improve the measurement of family
environments, it is recommended that the personal inter
view format be utilized as a viable tool for aggregating
data from small samples in lieu of the questionnaire.
4. Inasmuch as maternal knowledge, attitudes and
expectancies in regard to the child's academic achievement
104
are susceptible to change over time, it is recommended
that longitudinal studies be designed to document the kind
and degree of change that occurs.
5. To compare and contrast the nature of maternal
variables among unlike subpopulations, it is recommended
that research studies in academic achievement be conducted
on different ethnic and racial groupings and in various
regions of the country.
6. Because many mothers seem to lack knowledge
regarding their child's reading level, it is recommended
that procedures be initiated by educators to supply parents
with this important information.
105
BIBLIOGRAPHY
BIBLIOGRAPHY
Adelman, II., Taylor, L. , Fuller, W. , & Nelson, P.
Discrepancies among student, parent, and teacher ratings;
of the severity of a student's problems. American j
Educational Research Journal, 19 79, 1_6 (1) , 38.41. i
Adler, A. Practice and theory of individual psychology.
New York: Harcourt. Brace, 1923. ;
Athey, I. J., & Holmes, J.A. Reading success and per
sonality characteristics in junior high school students.;
University of California Publications in Education j
(Vol. 18), Berkeley, California: University of
California Press, 1969. '
j
Bachman, J.G. Youth in transition, Volume II: The impact '
of family background and intelligence on tenth-grade
boys. Ann Arbor, Michigan: Survey Research Center, j
Institute for Social Research, 1970.
Banks, W.C., & McQuater, G*V*. Achievement.motivation and ,
black children. IRCD Bulletin, 1976, _11(4). Columbia j
University, New York, N.Y. Institute for Urban and ;
Minority Education. i
I
i
Baratz, S.S., & Baratz, J.C. Early childhood intervention: ^
The social science base of institutional racism.
Harvard Educational Review, 1970 , 4_0, 29-50.
Bell, J.E. The Mother-child Relationship Evaluation. In
O.K. Buros (Ed.), Personality tests and reviews.
Highland Park, New Jersey: Gryphon, 1970, 1124-1125.
Bloom, B.S. Stability and change in human characteristics.
New York: John Wiley & Sons, Inc., 1964.
Bloom, B.S. Human characteristics and school learning.
New York: McGraw-Hill Book Company, 1976.
Bloom, B.S. The new direction in educational research:
Alterable variables. Phi Delta Kappan, 1980, 61
382-385.
Boocock, S.S. An introduction to the sociology of learning.
New York: Houghton MTFflin Company^ T97 2.
107
Bradburn, N.M., & Sudman, S. Improving interview method
and questionnaire design. San Francisco: Jossey-Bass
Publishers, 1979.
Bridge, R.G., Judd, C.M., & Moock, P.R. The determinants I
of educational outcomes. Cambridge, Massachusetts:
Ballinger Publishing Company, 1979. j
Bryant, E.C., Glaser, E., Hansen, M.H., & Kirsch, A. |
Associations between educational outcomes and background
variables: A review of selected literature. Monograph i
of the National Assessment of Educational Progress, a
project of the Education Commission of the States,
Denver, Colorado, 1974.
Burton, N.W., & Jones, L.V. Recent trends in achievement
levels of black and white youth. Educational Researcher
1982, 11 (4) , 10-14 .
Chapman, J.W., & Boersma, F.J. Learning disabilities,
locus of control, and mother attitudes. Journal of
Educational Psychology, .197 9, 7JL (2), 250-258.
!
Coleman, J.W., Campbell, E.Q., Hobson, C.F., McPartland, J . , |
Mood, A.M., Weinfeld, F.D., & York, R.L. Equality of 1
educational opportunity. Washington, D.C.: U.S. ;
Department of Health, Education and Welfare, Office of
Education, U.S. Government Printing Office, 1966. ;
Crandall, V. , Dewey, R., Katkovsky, W., & Preston, A.
Parents attitudes and behaviors and grade-school
children's academic achievements. The Journal of
Genetic Psychology, 1964 , .104, 53-66.
Cronbach, L.J. Coefficient alpha and the internal structure
of tests. P sychomet r i ka , 1951, 1_6, 297-334 .
Dave, R.H. The identification and measurement of
environmental process variables that are related to
educational achievement. Unpublished doctoral
dissertation, University of Chicago, 1963. :
Della-Piana, G., & Martin, H. Reading achievement and ;
maternal behavior. The Reading Teacher, 1966, 20,
225-230. ;
Dix, W.E. Relationship between selected family variables
and student achievement in school. (Doctoral
dissertation, Teachers College, Columbia University,
1977). Dissertation Abstracts International, 1977,
38(01), 40-A. (University Microfilms No. 77-14, 715).
108
Endler, N.S., & Magnusson, D. Personality and person by
situation interactions. In N.S. Endler & D. Magnusson
(Eds.), Interactional psychology and personality.
New York: John Wiley & Sons, Inc., 1976.
Epstein, J., Berg-Cross, G., & Berg-Cross, L. Maternal i
expectations and birth order in families with learning j
disabled and normal children. Journal of Learning |
Disabilities . 198 0, .13 ( 5) , 273-280 . :
Farr, R.C., & Roelke, P.L. Measuring subskills of reading:
Intercorrelations between standardized reading tests, :
teachers' ratings, and reading specialists' ratings. J
Journal of Educat .ional Measurement, .1971, 8 * (1) , 27-32.
Earr, R.C., & Tuinman, J.J. The dependent variable:
measurement issues in reading research. Reading
Research Quarterly, Spring 1972, _7(3), 413-423.
Fishman, J.A., Deutsch, M., Kogan, L., North, R.,
& Whiteman, M. Guidelines for testing minority group
children. Journal of Social Issues Supplement, 1964,
20 , 129-145. |
Handel, G. Psychological study of whole families. i
Psychological Bulletin, 1965 , 6J3, 19-41. ;
Harris, A.J. (Ed.) Casebook on reading disability. ;
New York: David McKay Company, Inc., 1970. |
Harris, D.B. The Mother-child relationship evaluation.
In O.K. Buros (Ed.), Personality tests and reviews.
Highland Park, New Jersey: Gryphon, 1970, 1125.
Harris, T.L., & Hodges, R.E. (Eds.) A dictionary of
reading and related terms. Newark, Delaware:
International Reading Association, 1981.
Hartwig, F., & Dearing, B.E. Exploratory data analysis.
Quantitative applications in the social sciences,
series no. 07-016. Beverly Hills, California: Sage
Publications, 1979.
Heiss, J. The case of the black family: A sociological
inguiry. New York: Columbia University Press, 197 5. !
Hilliard, T., & Roth, R.M. Maternal attitudes and the
nonachievement syndrome. Personnel and Guidance
Journal, January 1969, 424-428.
109
Horowitz, D.G., Shah, B.V., & Simmons, W.R. The unrelated :
question randomized response model. Proceedings of |
the American Statistical Association. Washington, D.C.:j
American Statistical Association, 1967. |
Humphries, T.W. An investigation of consistencies and !
inconsistencies in mothers' attitudes toward rearing j
their learning disabled boys having specific types of 1
learning disorders (Doctoral dissertation, Northwester^
University, 1976). Dissertation Abstracts International
1977 , 3_7(7-8), 4281-A. (University Microfilms i
No. 77-1271) !
Jahoda, Marie Current concepts of positive mental health. !
! New York: Basic Books, Inc., 1958. i
Jillings, C.R., Adamson, C.A., & Russell, T. An applica- |
tion of Roth's Mother-child Evaluation to some mothers !
of handicapped children. Psychological Reports, 1976,
38, 807-810. :
Reeves, J.P. Educational environment and student achieve- j
merit. Stockholm: Almquist & Wiksell, 1972. !
Kerlinger, F.N., & Pedhazur, E.J. Multiple regression in I
behavioral research. New York: Holt, Rinehart &
Winston, 1973. j
t
Kim, J., & Kohout, E.J. Multiple regression analysis:
Subprogram regression. In N.H. Nie, C.H. Hull, j
J.G. Jenkins, K. Steinbrenner, & D.H. Bent (Eds.), ;
SPSS: Statistical package for the social sciences, J
second edition. New York: McGraw-Hill, Inc., 1975. J
!
Koppenhaver, A.H. Reading achievement as influenced by"""
certain home factors (Doctoral dissertation, University ;
of Southern California, 1971). Dissertation Abstracts J
International, 1972, _3_2(7-8), 4243-A. (University
Microfilms No. 72-6078)
Lemon, N. Attitudes and their measurement. London:
B.T. Batsford, Ltd., 1973. ;
Lewin, K. A dynamic theory of personality. New York:
McGraw-Hill, 1935.
Lewis-Beck, M.S. Applied regression: An introduction. :
Quantitative applications in the social sciences,
series no. 07-022. Beverly Hills, California: Sage
Publications, 1980.
110
MacGinitie, W.H. An introduction to some measurement
problems in reading. In W.H. MacGinitie (Ed.),
Assessment problems in reading. Newark, Delaware:
International Reading Association, 1973.
Marjoribanks, K. Families and their learning environments.
London: Routledge & Kegan Paul, Ltd., 1979.
Mayeske, G.W., & Beaton, Jr., A.E. Special studies of our
nation's students. Washington, D.C.: U.S. Government
Printing Office, 197 5.
McAdoo, H.P. Components of educational achievement and
mobility in black families. Paper presented at the
Annual Meeting of the American Educational Research
Association. (New York, N.Y., April 3-8, 1977)
(ERIC Document Reproduction Service No. ED-141-208)
Miller, D.C. Handbook of research design and social
measurement (2nd ed.). New York: David McKay Company,
Inc., 1970.
New York City Planning Commission. Community planning
district profiles: Socioeconomic characteristics.
New York: City Planning Commission, 1974.
Oidick, M.S. Mothers' behaviors and sons' reading achieve
ment and projected feelings (Doctoral dissertation,
Wayne State University, 1976). Dissertation Abstracts
International, 1977 , 3_7(5-6), 2750-A. (University
Microfilms No. 76-26163)
Peters, M.F. Parenting in black families with young
children: A historical perspective. In H.P. McAdoo
(Ed.), Black families. Beverly Hills, California:
Sage Publications, 1981.
Pikulski, J.J. The validity of three brief measures of
intelligence for disabled readers. The Journal of
Educational Research, 1973 , 6_7 (2) , 67-68 , 80.
Plowden Report. Children and their primary schools.
A report of the Central Advisory Council for Education
(England). London: Her Majesty’s Stationery Office,
1967.
Reiss, A.J., Jr., Duncan, L.D., Hatt, P.K., & North, C.N.
A socioeconomic index for all occupations, Chapter VI
in Occupations and social status. New York: The Free
Press of Glencoe, Inc., 1961.
Ill
Revicki, D.A. The relationship among socioeconomic status,
home environment, parent involvement, child self
concept and child achievement. Paper presented at the
Annual Meeting of the American Educational Research ,
Association (Los Angeles, California, April 13-17, 1981);
Roth, R.M. The Mother-child Relationship Evaluation Manuals
Los Angeles: Western Psychological Services, 1961. I
I
Samuda, Ronald J. Psychological testing of American
minorities: Issues and consequences. New York: j
Dodd, Mead & Company, 1975. ■
Sartain, H.W. Research summary: Family contributions to
reading attainment. In H.W. Sartain (Ed.), j
Mobilizing family forces for worldwide reading success.
Newark, Delaware: International Reading Association,
1981.
Schaefer, E.S. Family relationships. In J.J. Gallagher
(Ed.), The application of child development research
to exceptional children. Reston, Virginia: The
Council for Exceptional Children, 1975. j
i
Schaefer, E.S., & Bayley, N. Maternal behavior, child j
behavior and their intercorrelations from infancy |
through adolescence. Monograph of the Society for ;
Research in Child Development, 19 6 3 , _28 , No. 3 . \
Smith, C.B. (Ed.) Parents and reading. Perspectives in
Reading Series: No. 14. Newark, Delaware: !
International Reading Association, 1971.
Snow, C.E. The objectives and research design of the
NIE-sponsored study on families and literacy. Paper
presented at the Annual Meeting of the American
Educational Research Association (New York City,
March 19-23, 1982).
Solomon, D., Houlihan, K., Busse, T., & Parelius, R.
Parent behavior and child academic achievement,
academic striving, and related personality characteris
tics. Genetic Psychology Monographs, 19 71, 8_3 , 173-273i
Symonds, P.M. Dynamics of parent-child relationships.
New York: Bureau of Publications, Teachers College,
Columbia University, 1949.
Taylor, R.G. Personality traits and discrepant achieve
ment. Journal of Counseling Psychology, 1964, 11(1),
76-82. '
112
Thorndike, R.L. Reading comprehension education in fifteen j
i countries: International studies in education (Vol. j
III). New York: John Wiley and Sons, 1973. |
j
U.S. Bureau of the Census. The social and economic status |
of the Black population in the United States 1974. j
Current Population Reports. Special Studies Series P - 2 3 j
! No. 54. Washington, D.C.: U.S. Department of Commerce,:
July, 1975.
U.S. Commission on Civil Rights. Social indicators of j
equality for minorities and women. Washington, D.C.:
U.S. Commission on Civil Rights, 1978. |
I
Walberg, H.J., & Marjoribanks, K. Differential mental
abilities and home environment: A canonical analysis.
Developmental Psychology, 1973, 9_, 363-368.
Walberg, H.J., & Marjoribanks, K. Family environment and j
cognitive development. Twelve analytic models. J
Review of Educational Research, 1976, _46, 527-551. j
i
Wetter, J. Parent attitudes toward learning disability. |
Exceptional Children, 1972, 490-491. j
Williams, J.D. (Ed.) The state of Black America. i
New York: National Urban League, Inc., 1980.
Wolf, R. The measurement of environments.
In A. Anastasi (Ed.) , Testing problems in perspective. j
Washington, D.C.: American Council on Education, 1966. j
! I
APPENDIXES
APPENDIX A
NORC-NYC CATEGORIES
NORC-NYC Categories
Occupations were classified into the following
seven categories of the National Opinion Research Center
pccupation Scale and New York City Planning Commission on
Labor Force Participation:
1. Laborers
a. freight
b. stock and material handlers
c. farm laborers
d. unpaid family workers
t 2. Service
| a. janitors
b. security watchmen
! c, waiters, waitresses
| d. servants
i e« domestics
i 3. Operatives
I a. workers in factories and industry
| b. truck drivers
! c. other transport workers
4. Craftsmen
a. auto mechanics
b. body repairmen
c. other mechanics and repairmen
d. machinist
j e. other metal craftsmen
f. carpenters
g. construction craftsmen
5. Clerical and Sales
a, bookkeepers
b. secretaries
c» stenographers
st s
r clerical workers
s personnel in retail
factur ing
r industries
116
d. typi
e. othe
f . sale
g. manu
h. othe
6. Managerial
a. managers
b. retailers
c. manufacturers
d. farmers
e. farm managers
7. Professional
a. technicians
b. engineers
c. physicians
d. dentists
e. medical and other health workers
f. public school teachers
g. administrators
117
APPENDIX B
INTERVIEW SCHEDULE
COPY
Name of Mother
Name of Child
Subject No.
Date
INTERVIEW SCHEDULE
Instructions: I am going to ask you a few short questions
for purposes of this research study. I hope you will
answer as truthfully as possible so that the results of
this study will be accurate. However, if you would rather
not respond to any of the questions, you're under no
obligation to do so. You can be assured that this
informat ion will be held in the strictest of confidence
and that it will be used in this study only.
1. How many children are there in your family?
What are their ages?
2. Are you single, married, separated, or divorced?
What is your husband's occupation?
3. Do you work?
What is your occupation?
4. What was your approximate family income last year?
less than $5,000__________ $15 , 000-$20 , 000_______
$5,000-$10,000_
$10,000-$15,00 0
$ 2 0, 0 0 0-$ 2 5,0 0 0__
more than $25,000
5. How far did you go in school?
less than high school________
some high school______________
completed high school________
some college______
completed college
6. How far did your husband go in school?
less than high school
some high school_____
completed high school
some college______
completed college
119
COPY Interview Schedule (Continued)
7. What kinds of things does your family do together?
8. When your child was a preschooler, did he/she have
any problem in eating, sleeping, or becoming toilet
trained?
child's general activity
school?
active inactive
10. During these preschool years, was your child difficult;
to raise? In other words, did he/she have colic,
did he/she get into trouble frequently, would you ;
say he/she was cranky or irritable? j
i
11. How did your child get along with his/her family j
when he/she was a preschooler? j
12. What words would you use to describe your child dur- j
ing his/her preschool years? (e.g., quiet, happy,
aggressive, disagreeable, friendly) j
13. Has your child ever received special help in reading j
at school? If yes, in what way?
14. Was he/she ever put in a special educat ion class?
Did he/she
(e.g., spee
receive any
ch therapy)
special services at school?
15. Was he/she ever retained or held back at a certain
grade level?
16. Have you ever taken your child to another clinic
for diagnosis or instruction in reading?
17. What would you estimate as the grade level of your
child's reading ability if he/she were given a list
of isolated words to read? (Show the WRAT)
18. What would you estimate to be the grade level of his/
her reading ability as far as his/her understanding
of the mean ing of isolated words is concerned? In
other words, what do you think is his/her vocabulary
level? (Show the Gates-MacGinitie)
9. How would you describe your
level before he/she went to
very active_____ moderately
120
COPY Interview Schedule (Continued)
19. At what grade level would you estimate his/her
reading comprehension to be in terms of understanding
the meaning of a short paragraph? (Show the Gates-
MacGinitie)
20. About how many hours a week does your child normally ;
watch TV?
21. Does he/she have any special hobbies or interests?
22. What do you think he/she might do for a living
some day?
23. How far do you expect him/her to go in school?
24. Where do you think he/she could be reading at now,
as far as recognizing a list of words is concerned?
Where do you think he/she could be reading at now,
as far as his/her vocabulary and comprehension in
reading are concerned? I mean grade level, of course.
WR_____ V_____ C______
25. What facts or events in your child's life would you
imagine were important in his/her reading development?
26. Do you have any ideas why he/she is not reading at a
higher level than he/she presently is?
Interviewer's subjective appraisal of subject's
socio-economic level.
C Copyrighted 1981
APPENDIX C
MEASURES USED FOR VARIABLES
MEASURES USED FOR VARIABLES
Measure
i Maternal
| Nonacceptance
Mother's Estimates
j of Reading
; Achievement
Mother's Expectancies
of Reading Achievement
Socio-economic level
Verbal Intelligence
Reading Achievement
De scr iption
Ordinal measure based on the
overprotection and rejection
subscales of Roth's Mother-
child Relationship Evaluation !
Reported statement from j
Interview Schedule of child's
estimated grade level in word
recognition, vocabulary and
comprehen sion
Reported statement from
Interview Schedule of child's
expected grade level in word
recognition, vocabulary and
comprehension
Index of socio-economic level
of family based on 4 items
from Interview Schedule,
using 7 occupational categor
ies from the NORC/NYC j
Occupational Scale j
Individually-admini stered
Peabody Picture Vocabulary
Test |
Individually-administered j
reading subtest of the WRAT |
and group-administered !
Gates-MacGinitie Reading Tests!
APPENDIX D
DATA FOR ESTIMATES, EXPECTANCIES
AND READING TEST SCORES
125
RAW DATA FOR ESTIMATES, EXPECTANCIES AND READING TEST SCORES
MATERNAL VARIABLES
s
Child's
Grade
Level
Mother
Child's
's Estimate of
Reading Ability
Mother
Child's
s Expectancy of
Reading Ability
Child's Reading
Test Scores
WR Voc Comp WR Voc Comp WR Voc
1 5 3.5 4.5 3.5 7.5 7.5 7.5 3.1 2.1
2 6 6.5 6.5 5.5 8.5 7.5 7.5 5.7 4.5
3 3* 2.5 2.5 1.5 4.5 4.5 4.5 2.2 1.6
4 5* 3.5 5.5 3.5 8.5 8.5 8.5 2.7 1.8
5 6 4.5 4.5 4.5 7.5 7.5 7.5 6.2 4.5
6 4 2.5 4.5 2.5 4.5 4.5 2.5 2.8 1.8
7 3 3.5 3.5 3.5 4.5 4.5 4.5 3.8 3.1
8
2*
1.5 2.5 2.5 2.5 3.5 3.5 2.5 1.5
9 5 3.5 3.5 4.5 6.5 6.5 6.5 3.1 2.7
10 5* 3.5 3.5 4.5 6.5 6.5 6.5 4.1 5.3
11 5 2.5 3.5 2.5 6.5 6.5 6.5 3.3 2.4
12 5* 5.5 5.5 5.5 6.5 6.5 6.5 3.6 2.5
13 4 2.5 2.5 2.5 4.5 4.5 4.5 2.5 1.6
14 5 4.5 4.5 4.5 6.5 6.5 5.5 4.4 2.1
15 6 5.5 7.5 6.5 6.5 8.5 7.5 5.5 4.9
16 4* 2.5 3.5 3.5 4.5 4.5 4.5 2.8 1.7
17 4 4.5 4.5 6.5 6.5 7.5 7.5 6.6 7.0
18 3* 1.5 3.5 2.5 3.5 4.5 3.5 2.9 2.0
19 5 3.5 4.5 3.5 5.5 5.5 4.5 4.1 3.6
20 5 1.5 4.5 4.5 5.5 5.5 5.5 1.7 1.3
21 7 5.5 5.5 6.5 7.5 7.5 7.5 5.7 4.2
22 6 5.5 5.5 4.5 6.5 6.5 5.5 6.5 6.0
CRITERION VARIABLES
Comp
1.6
3.2
1.5
2.5
3.2
2.0
3.6
1.7
2.8
4.3
2.5
1.9
1.4
1.7
5.6
2.3
4.7
2.3
2.8
1.5
5.1
3.2
RAW DATA FOR ESTIMATES, EXPECTANCIES AND READING TEST SCORES (Continued)
MATERNAL VARIABLES
WR
Child * s
Grade Mother's Estimate of Mother s Expectancy of
S
Level Child's Reading Ability Child's Reading Ability
23
WR Voc Comp WR Voc Comp
5 5.5 5.5 5.5 6.5 6.5 6.5
24 4* 3.5 3.5 3.5 5.5 5.5 5.5
25 3 2.5 2.5 1.5 4.5 3.5 3.5
26 2 1.5 2.5 0.5 2.5 2.5 2.5
27 5 4.5 4.5 5.5 5.5 4.5 5.5
28 3 1.5 1.5 2.5 3.5 3.5 3.5
29 2 2.5 2.5 2.5 3.5 2.5 2.5
30
2*
1.5 1.5 1.5 3.5 3.5 3.5
31 3 3.5 3.5 3.5 3.5 3.5 3.5
Word Recognition
Voc: Vocabulary
Comp: Comprehension
^Retained in one grade level
CRITERION VARIABLES
Child's Reading
Test Scores
WR Voc Comp
4.5 3.8 4.1
2.4 1.7 1.6
1.8 1.4 1.3
2.3 0.5 0.5
4.2 4.1 2.9
2.3 1.2 1.3
3.2 1.7 1.6
2.6 1.7 2.0
4.1 2.9 1.9
i—1
to
APPENDIX E
COMPOSITES OF ESTIMATES,
EXPECTANCIES AND READING TEST SCORES
COMPOSITES OF
ESTIMATES, EXPECTANCIES AND READING TEST SCORES
Child 1s
Child's Mother's Reading
§. Grade Level Estimate Expectancy Test Score
1 5 3.8 7.5 2.3
2 6 6.2 7.8 4.5
3 3* 2.2 4.5 1.8
4 5* 4.2 8.5 2.3
5 6 4.5 7.5 4.6
6 4 3.2 3.8 2.2
7 3 3 . 5 4.5 3 . 5
8 2* 2.2 3.2 1.9
9 5 3.8 6.5 2.9
10 5* 3.8 6.5 4.6
11 5 2.8 6.5 2.7
12 5* 5.5 6.5 2.7
13 4 2.5 4.5 1.8
14 5 4.5 6.2 2. 7
15 6 6.5 7.5 6.0
16 4* 3.2 4.5 2.3
17 4 5.2 7.2 6.1
18 3* 2.5 3.8 2.4
19 5 3.8 5.2 3.5
20 5 3.5 5.5 1.5
21 7 5.8 7.5 5.0
22 6 5.2 6.2 5.2
23 5 5.5 6.5 4.1
24 4* 3.3 5.5 1.9
25 3 2.2 3.8 1.5
26 2 1.5 2.5 1.1
27 5 4.8 5.2 3.7
28 3 1.8 3.5 1.6
29 2 2.5 2.8 2.2
30 2* 1.5 3.5 2.1
31 3 3 . 5 3.5 3.0
*Retained! in one grade level
128
APPENDIX F
CODING SCHEME FOR VARIABLES
130
CODING SCHEME FOR VARIABLES
Variable or Variable Composi t e
Maternal Block (100 series)
Nonacceptance, Estimates, Expectancies
Maternal Nonacceptance:
Mother-child Relationship Evaluation. .... ............
Composite of Mother's Estimates
of Reading Grade Levels .....................................
Word Recognition. . . . . . . . . . . . . . . WR111
Vocabulary. .... ........ V0CAB112
Comprehension . . . . . . . . . . . . . . . . C0MP113
Composite of Mother's Expectancies
of Reading Grade Levels . . .................................
Word Recognition. ........... . WR121
Vocabulary. ................ ......... VOCAB122
Comprehension . .......... COMP123
Socio-economic Status (200 series)
Index of Socio-economic Level of Family ..........
Verbal Intelligence (300 series)
Verbal Intelligence of Child:
Peabody Picture Vocabulary Test . . . . . . . . . . . . . .
Code
NONAC100
EST110
EXP 12 0
SES200
PPVT300
1 3 1
CODING SCHEME FOR VARIABLES (Continued)
Variable or Variable Composite Code
Reading Scores (700 series)
Composite of Child’s Reading Test Scores. .... ...........
Word Recognition: Wide Range Achievement Test. . WR711
Vocabulary: Gates-MacGinitie Reading Tests . . .VOCAB712
Comprehension: Ga t e s-*MacGini t i e Reading Tests. . COMP713
Transformed Variables (800 series)
Temporary Transformation of Reading Variables
to Difference Scores
Mother's Estimates Less Reading Scores
WR111 - WR711..................................... WR811
VOCAB 112 - VOCAB 7 1 2 ...................................VOCAB812
COMP 113 - COMP 713 COMP 813
Mother's Expectancies Less Reading Scores
WR121 - WR711..................................... WR821
VOCAB 12 2 - VOC AB 71 2 ...................................VOCAB822
COMP 123 - COMP 7 1 3 ................. COMP823
SCR710

Asset Metadata

Creator Tassopulos, Joan M. (author)

Core Title The relationship of maternal factors and socioeconomic level to the reading achievement of 7-11 year-old black students

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Education

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

Tag Black studies,education, reading,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c26-518814

Unique identifier UC11245600

Identifier usctheses-c26-518814 (legacy record id)

Legacy Identifier DP24948.pdf

Dmrecord 518814

Document Type Dissertation

Rights Tassopulos, Joan M.

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