University of Southern California Dissertations and Theses

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

(USC Thesis Other)

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content AN INTEGRATIVE ANALYSIS OF RACISM AND QUALITY OF LIFE:
A COMPARISON OF MULTIDIMENSIONAL MODERATORS IN AN
ETHNICALLY DIVERSE SAMPLE
Copyright 2005
by
Jennifer Lauren Best
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
(PSYCHOLOGY)
August 2005
Jennifer Lauren Best
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UMI Number: 3196777
Copyright 2005 by
Best, Jennifer Lauren
All rights reserved.
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 3196777
Copyright 2006 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ii
ACKNOWLEDGEMENTS
I would like to take this opportunity to express my warm thanks and
gratitude to several individuals without whom this ambitious endeavor would not
have gotten off the ground. I first and foremost thank my adviser, Dr. Gerald C.
Davison for his never-wavering confidence in my ability to actualize my
research passions within the area of ethnic minority health. I will never forget
his invaluable guidance and support throughout my graduate career and beyond.
I would also like to thank the other members of my dissertation committee,
which include Dr. Beth Meyerowitz, Dr. Shannon Daley, Dr. Richard John,
and Dr. Rodney Goodyear for their time and challenging inquiries, which
enhanced my ability to think more critically about how my research fits into the
context of the larger empirical literature. Thank you also to Dr. Barry Reynolds
and his staff at the USC Human Relations Center for allowing me to use this
non-traditional setting for collecting this research data. To my several diligent
and hardworking undergraduate research assistants over the last three years, I
applaud and thank you all deeply for your time and energy on the front lines of
the research process. Finally, I would like to thank my fellow colleagues,
friends, and family for your continued support of and faith in me realizing my
professional goals over the years.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
iii
TABLE OF CONTENTS
Section Page
Acknowledgements................................................................ ii
List of Tables........................................................................ iv
List of Figures....................................................................... v
Abstract.................................................................................. ix
Chapter I: Introduction.......................................................... 1
Chapter II: Method................................................................ 45
Chapter III: Results................................................................ 63
Chapter IV: Discussion........................................................... 140
Chapter V : Summary............................................................... 182
Chapter VI: Limitations and Future Areas of Research 183
Chapter VII: Conclusions and Implications............................. 188
References................................................................................... 191
Appendix...................................................................................... 213
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
iv
LIST OF TABLES
Table Page
Table 1: Descriptive Statistics for the Primary Study Variables 64
for the Full Sample
Table 2: Correlations Among Primary Study Variables 65
for the Full Sample
Table 3: Ethnic Group Differences on Demographic Variables 68
Table 4: Current Annual Household Percentages by Ethnic Group 69
Table 5: Ethnic Group Differences on Independent and Dependent 72
Variables
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
V
LIST OF FIGURES
Figure Page
Figure 1: Perceived quality of life/general health as 79
a function of hostility and the frequency of lifetime
racist events (in standard deviations from the mean) for
the Asian American participant group.
Figure 2: Perceived quality of life/general health as a 86
function of anger rumination and the frequency of lifetime
racist events (in standard deviations from the mean) for the
Asian American participant group.
Figure 3: Family member closeness as a function of anger 89
rumination and the frequency of lifetime racist events (in
standard deviations from the mean) for the African American
participant group.
Figure 4: Family member closeness as a function of ethnic 90
identity and anger rumination (in standard deviations from
the mean) for the African American participant group.
Figure 5: Family member closeness as a function of 92
ethnic identity, anger rumination and the frequency of
lifetime racist events (in standard deviations from the mean)
for the European American participant group.
Figure 6: Family member closeness as a function of ethnic 93
identity and anger rumination (in standard deviations from
the mean) for the Latino American participant sample.
Figure 7: Friendship closeness as a function of ethnic identity 95
and anger rumination (in standard deviations from the mean) for
the African American participant sample.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
vi
Figure Page
Figure 8: Friendship closeness as a function of anger rumination, 96
ethnic identity and the frequency of lifetime racist events (in
standard deviations from the mean) for the African American
participant group.
Figure 9: The frequency of positive health behaviors as a 99
function of ethnic identity and the frequency of lifetime racist
events (in standard deviations from the mean) for the African
American participant sample.
Figure 10: The frequency of positive health behaviors as 100
a function of health values, ethnic identity and the frequency
of lifetime racist events (in standard deviations from the mean)
for the African American participant group.
Figure 11: Perceived quality of life/general health as a 105
function of ethnic identity and the frequency of positive health
behaviors (in standard deviations from the mean) for the Latino
American participant sample.
Figure 12: Somatic distress as a function of the frequency 106
of positive health behaviors and racism-related stress (in standard
deviations from the mean) for the African American participant sample.
Figure 13: Somatic distress as a function of the frequency 108
of positive health behaviors and racism-related stress (in standard
deviations from the mean) for the Asian American participant sample.
Figure 14: Family member closeness as a function of the 110
frequency of positive health behaviors and racism-related
stress (in standard deviations from the mean) for the European
American participant sample.
Figure 15: Family member closeness as a function of ethnic 111
identity and racism-related stress (in standard deviations from
the mean) in predicting family member closeness for the Asian
American participant sample.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
vii
Figure
Figure 16: Friendship closeness as a function of ethnic identity
and the frequency of positive health behaviors (in standard deviations
from the mean) for the Asian American participant sample.
Figure 17: Perceived quality of life/general health as a function
of the frequency of drinking until feeling ill and racism-related
stress (in standard deviations from the mean) for the Latino
American participant sample.
Figure 18: Family member closeness as a function of the
frequency of drinking until feeling ill and racism-related stress
(in standard deviations from the mean) for the Asian American
participant sample.
Figure 19: Perceived quality of life/general health as a
function of ethnic identity and racism-related stress (in standard
deviations from the mean) for the African American participant sample.
Figure 20: Perceived quality of life/general health as a
function of smoking frequency and racism-related stress (in standard
deviations from the mean) for the Latino American participant sample.
Figure 21: Family member closeness as a function of smoking
frequency and racism-related stress (in standard deviations from
the mean) for the Asian American participant sample.
Figure 22: Anger as a function of the frequency of situational
attributions and the frequency of perceived racism (in standard
deviations from the mean) for the Asian American participant sample.
Figure 23: Aggressive ideation as a function of the frequency
of situational attributions and the frequency of perceived racism (in
standard deviations from the mean) for the European American
participant sample.
Page
114
116
120
123
125
129
134
137
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure
viii
Page
Figure 24: Aggressive ideation as a function of ethnic 138
identity and the frequency of perceived racism (in standard
deviations from the mean) for the European American
participant sample.
Figure 25: Aggressive ideation as a function of the frequency 139
of situational attributions and the frequency of perceived racism
(in standard deviations from the mean) for the Asian American
participant sample.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ABSTRACT
The role of racism in affecting the psychological and physical wellness of
people of color has received increasing scholarly attention in recent decades (see
Clark et al., 1999 for a review). The principal objective of the current analysis
was to investigate health-related cognitive, affective, and behavioral individual
difference factors as potential moderators in the relationship between racism and
dimensions of quality of life in an ethnically-diverse sample from integrative,
functional-developmental and acculturative stress perspectives. More
specifically, the interaction of ethnic identity, trait hostility, an anger rumination
coping style, health values, health-preventive behavior participation, and the
frequency of maladaptive drinking and smoking behavior with indicators of
racist victimization in predicting psychological, interpersonal, and physical
components of life satisfaction were assessed. A second aim based in dynamic
constructivist theory (e.g. Hong et al., 2000) was to evaluate the willingness of
endorsing situational factors as explanations for the actions of others in
mitigating the anger and aggressive thinking resulting from perceptions of
racism in a series of everyday, ambiguous scenarios. Several results emerged in
support of both the proposed rejection-protection and cognitive flexibility
hypotheses, highlighting the importance of clarifying the influence of ethnicity-
related stressors in scaffolding health-relevant beliefs and behaviors. These
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
X
experiences in turn may have significant implications for simultaneously
promoting and compromising the well-being of ethnic minority individuals.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
And I always feel this...with people-that whenever they're being nice to
me, all the time really, underneath they're only assessing me as a criminal and
nothing else... they're quite incapable of accepting me as anything else.
-E. Goffman, Stigma: Notes on the Management of Spoiled Identity
INTRODUCTION
The research literature has advanced racism as an indelible socio-cultural
stressor for persons of color (Clark et al., 1999; Harrell, 2000; Harrell et al.,
1998) that manifests itself in a myriad of forms, a chameleon that wears many
“faces” or “masks” (Essed, 1990; Feagin, 1991). Stressful events attributed to
racism may range in severity from physically aggressive, violent assaults as in
hate crime victimization (Dunbar, 2001; Jacobs & Potter, 1998; Levin &
McDevitt, 1993), to undisguised verbal aggression and harassing hate speech
(Jacobs et al., 1998), to ethnicity-specific cognitive worries (e.g. stigma
consciousness, stereotype confirmation concern, stereotype threat: Contrada et
al., 2001; Pinel, 1999; Steele et al., 1995) to the covert subtlety of chronic,
“everyday” forms of modern ethnic discrimination (e.g. Adams & Dressier,
1988; Essed, 1990; Feagin, 1991; Feagin & Sikes, 1994; Swim et al., 1998;
Williams, 2000). The last of these forms of racism reportedly takes place in
common, daily, interpersonal contexts such as in public places by service
providers (Feagin, 1991; Feagin et al., 1994), as ethnic profiling by law
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2
enforcement agents within the community (Feagin, 1991; Harris, 2002), and by
employers and colleagues in the working environment (Feagin et al., 1994).
Reported rates of racism-related stress have been documented as high as 98-99%
in certain participant samples (e.g. Bowen-Reid & Harrell, 2002; Landrine &
Klonoff, 1996).
Several investigations have demonstrated racism-related stress to be
linked to significant impairments in both mental (e.g. depression, anxiety, low
self-esteem, PTSD symptoms, life satisfaction, perceived health) as well as
physical wellness (e.g. enhanced cardiovascular reactivity, frequency of health
care visits, low gestational age at delivery, incidence of and risk for serious
chronic illness) among ethnic minority individuals (e.g. Armstead et al., 1989;
Blascovich et al., 2001; Contrada et al., 2001; Klonoff, Landrine & Ullman,
1999; Korte, 2000; Kwate, Valdimarsdottir, Guevarra, & Bovbjerg, 2003;
Landrine & Klonoff, 1996; Loo et al., 2001; McNeilly et al., 1995; Sanders
Thompson, 1996; Williams, Spencer & Jackson, 1999). Although this emerging
body of research evidencing important relationships between racism experiences
and health outcomes among people of color is promising, several important
issues remain to be addressed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
3
Firstly, few investigations have fully examined the more complex role
of individual difference factors (e.g. ethnic identity, health values, health
behaviors, attributional style, and anger regulation tendencies) in modulating the
impact of racism on dimensions of well-being. Hence, it is of interest to begin to
clarify what factors may serve to attenuate or exacerbate the relationship
between racism experiences and aspects of well-being. More specifically, as an
extension of the rejection-identification model (see Branscombe et al., 1999) and
in conjunction with a functional-multicultural perspective on cognitive, affective
and behavioral development, little work has investigated how individual
differences in hostility and anger rumination coping may demonstrate
counterintuitive stress-buffering relationships within the specific context of
racism and dimensions of quality of life among certain ethnic minority groups.
Secondly, there is a dearth of research that has explored the relationship
of racism to self-reported indicators of health and wellness within more
ethnically-diverse samples. The few studies which have done so, have been
limited in the number of ethnic groups for comparison (e.g. African Americans
only or as compared to Caucasian samples). This approach however, fails to
take into account possible differences in the relationships among racism,
individual differences, and quality of life within and between multiple ethnic
minority groups. This is an important factor to consider in light of the growing
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4
ethnic diversity extant across the U.S. Thus it is unclear if relationships between
these variables may vary depending on both individual difference factors and
ethnic group membership. How might ethnic group membership shape the
nature of the relationships among racism-related stress, individual differences in
coping and wellness? Thirdly, scientific endeavors in the area of racism and
well-being have failed to examine a broader range of life satisfaction dependent
variables, most notably absent are identifying the potential positive links
between racism and interpersonal components of quality of life.
Finally, from an attributional standpoint, the literature has yet to identify
certain cognitive factors (e.g. contextual/situational attributions) that may
modify the positive relationships among perceived racism, anger and aggression
that occur in everyday, ambiguous interpersonal situations. Could scholars in
the area of hostile attributions, affect and behavior be overlooking the
contribution of perceptions of racism in shaping reports under conditions of
attributional ambiguity when the participant is a member of a stigmatized ethnic
minority group? In other words, could perceptions of racism be considered an
ethnicity-specific form of hostile attribution and if ethnicity differences in this
research emerge, might it be partially explained by perceptions of racial
maltreatment?
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
5
Specific Aims
The consistent perception and prevalence of racism-based stressors,
predominantly investigated among African American individuals are robust
findings within the literature (Swim et al., 1998). However, additional work is
needed in further clarifying possible underlying cognitive, affective and
behavioral individual difference factors involved in influencing modern racism’s
less objectively detectable presence as a significant contributor to the stress
process and ultimately to the emergence of or protection against adverse health
outcomes within ethnically diverse populations. In reference to this overarching
research goal and informed by functional-developmental, rejection-
identification, and socio-cultural perspectives on cognition, emotion regulation
and behavior, five such candidates are suggested to play significant roles in this
wellness-eroding and/or wellness-buffering process. These include ethnic
identity, health values/health behaviors, trait hostility, anger rumination coping
and a willingness to consider contextual factors in everyday ambiguous
situations. With this proposition in mind, the primary aims of the current project
include:
1) To test and compare the potential moderating roles of trait
hostility, anger rumination, health behaviors, and ethnic
identity in the relationship between racism and specific
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
6
dimensions of quality of life (including interpersonal factors
such friendship closeness) among samples of African
American, European American1 , Asian American and Latino
American young adults.
2) To assess the possible interaction of racism, health values
and/or ethnic identity in predicting the frequency of
participating in wellness-enhancing behaviors within an
ethnically-diverse participant sample.
3) To investigate the potentially significant positive link between
perceived racism and hostile attributions in response to
everyday, ambiguous situations.
4) To evaluate and compare the potential interaction of the
willingness to attribute situational factors, the frequency of
perceived racism and/or ethnic identity in accounting for
significant variation in both anger and aggressive ideation
within the context of daily, ambiguous interpersonal stressors
among four different ethnic groups.
5) To explore more complex and specific relationships among
ethnic identity and other individual differences in hostility,
1 Terms used to describe various ethnic groups will be used interchangeably throughout the text.
For example, participants of European heritage will also be referred to as Caucasian Americans.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
7
anger rumination, health values, health behaviors and contextual
attributional style in affecting the links between racism and well
being.
In general, these findings may contribute to informing the
development of more culturally-sensitive models detailing the impact of racism
on health and wellness among a variety of ethnic minority groups.
The Unifying Theme: Dynamic Constructivism and a Biopsychosocial Model
of Racism-related Stress and Health
Dynamic constructivism was first introduced in the literature by Hong
and colleagues (2000) as a useful theoretical framework for interpreting the
“frame-switching” phenomenon that emerged from their cognitive-experimental
studies. Frame-switching referred to the tendency of bicultural Hong Kong
Chinese young adults to utilize collectivistic or individualistic attributions for the
ambiguous behaviors of others depending on the implicit cultural priming
paradigm (i.e. Chinese or Western) they were exposed to (e.g. Hong et al.,
2000). A dynamic constructivist approach advocates scientists in the fields of
ethnic minority health and cultural psychology to carefully consider how
members of diverse groups may make meaning (or interpretations, attributions)
based on contextual factors (e.g. cultural history, cultural cues within a specific
environment). Consequently, this conceptualization would argue that the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
8
traditional relationships among cognitive, affective, and behavioral variables
yielded from primarily Caucasian samples would not necessarily remain
consistent when assessed in other ethnic or racial samples.
The modern theoretical shift away from mind-body dualism has given
birth to the fields of clinical health psychology, behavioral medicine and
psychoneuroimmunology (Smith & Nicassio, 1995). This paradigmatic
transformation has further given rise to a proliferation of collaborative and
integrative empirical efforts among behavioral scientists traversing the
disciplines of psychology and medicine. As such, it has become increasingly
important within the racism and wellness literature to combine perspectives,
research methodologies, and levels of analysis in order to demonstrate the
relationships among biological, psychological and social underpinnings linking
for example, racism stressors to the health (e.g. risk for chronic illness, health
beliefs, health behaviors) and quality of life of ethnic minority populations.
In accordance with these contemporary scientific aims, Clark and
colleagues (1999) synthesized a cognitive, perception-based, biopsychosocial
model describing the interactions among appraisals of individual racism
stressors and aspects of psychological and physical well-being for African
American cohorts. Previous research demonstrates how specific categorical and
individual difference factors such as ethnic or racial identity, SES and skin color
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
9
play integral roles in predicting attributions of racial discrimination and
subsequently influence coping responses to perceived racism stressors (see Clark
et al., 1999 for a review).
Their model further asserts how one copes with perceived racial slights
in turn modulates the psychological and physiological impact of the initial
precipitating event and ultimately these variables could influence health
outcomes such as depression, anger and physical disease among African
American individuals. This person-centered, appraisal framework may be
particularly amenable to examining the cumulative effects of chronic, everyday
modern racism “microstressors” on various health-related components for racial
and ethnic minorities. It also provides a foundation for testing the influence of
other health-related individual difference factors and coping styles on buffering
or exacerbating the affect of racism on dimensions of well-being for people of
color. Clark and colleagues’ (1999) framework was used as the basis for testing
prospective moderators in this study in order to expand the existing literature to
include an examination and comparison of these factors in a wider range of
ethnic minority samples.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
10
A Functional Developmental Perspective on Racism Attributions and Ethnic
Identity as Wellness-Protecting Factors
The presence of modern, everyday racism is quite elusive and more
difficult to detect than its more conspicuous predecessors (e.g. lynching) and
contemporaries (e.g. hate speech). This intangibleness in turn sets the stage for
what has been introduced in the literature as "attributional ambiguity" (Crocker,
Voelkl, Testa, & Major, 1991). Thus, when an ethnic minority individual is
confronted with an exchange that may be racially motivated, there are several
causes one may attribute in attempting to explain the reasons behind such
situations. Miller and Major (2000) discussed how one of the primary means
racially stigmatized groups use in coping with and managing possible incidents
involving discrimination in order to protect self-esteem is to attribute negative as
well as positive outcomes to racism (see also Crocker & Major, 1989; Miller &
Kaiser, 2001).
A little over a decade ago, work by Crocker and colleagues (1991)
indicated how self-esteem following the reception of negative and positive
feedback by an unseen fellow undergraduate of European heritage was
dependent on the visibility of racial group membership. More specifically, when
their African American sample was aware that the nonexistent "other student"
could see them from behind a one-way mirror as they completed a batch of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
questionnaires, they were more likely to attribute any negative feedback given
by the student they were paired with to racial prejudice than the Caucasian
group. In so doing, this visible African American group achieved higher levels
of self-esteem than did the African American students who were not visible to
the student they were matched with (Crocker et al., 1991). They also reported
significantly greater positive affect than did the Caucasian students after
receiving criticism.
The Crocker et al. (1991) study was unique in that it also emphasized
how racial prejudice could be attributed for positive as well as negative results.
With regard to positive feedback, visibility of race also played an interesting
part. When the African American group could be seen, they tended to minimize
the positive opinions of the other student to a greater extent than did the group
who could not be seen. In this way, when race is a visible factor in evaluative
situations, it is likely that people of color may experience attributional confusion
or uncertainty for the rationale associated with positive or negative outcomes in
interracial contexts. As such, this attributional ambiguity has also been proposed
to be a potential stressor for ethnic minorities (Crocker et al., 1991).
The self-preserving benefits of an externalizing attributional style among
African American groups has been further supported by research with
adolescents (Hillman, Wood, Sawilowsky, 1992) and with other young adult
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
12
populations (e.g. Crocker, Luhtanen, Broadnax, & Blaine, 1999; Major, 1995).
Major and fellow researchers for example, found that on average, the
performance self-esteem of their European American cohort was more
dependent on feedback type than was a similar indicator in their group of
African American participants (Major, Spencer, Schmader, Wolfe, & Crocker,
1998). White participants' self-esteem was more likely to be higher following
positive feedback and lower after receiving negative feedback while the self
esteem of the Black participants remained the same regardless of feedback.
This phenomenon of disidentifying or disengaging one's self-esteem
from assessments made within the academic atmosphere has further implications
for the psychological functioning of ethnic minority individuals. Based on
stereotype threat research with people of color, authors have speculated that
perceived racial bias overlaying performance evaluations is an important
mechanism involved in the process of reducing the value and worth placed on
achieving academically (Steele, 1997). In this way, self-esteem is divested of
any relation to academic performance. This may be especially apparent among
specific, high risk groups of ethnic minority individuals who are performing
poorly (Major, 1995).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
13
Contrary to the findings of the Crocker group who espouse claims of the
adaptive, self-protective function of making attributions to discrimination,
research by other authors submits a different view (Ruggerio & Taylor, 1995;
Ruggerio & Taylor, 1997). Rugger io and Taylor (1997) used a novel
experimental approach to examine when ethnic minority individuals are likely to
attribute poor performance to racial prejudice under increasing levels of
ambiguity. In their protocol, participants were assigned to conditions varying in
the probability of receiving feedback by a racist confederate.
Using the explicitly defined base rate data provided, it was observed that
unless participants were absolutely certain of the tendency of their judge to be
racist, both African American American and Asian American undergraduates
often refrained from attributing negative grading to racism. This minimization
yielded benefits for perceived control and social self-esteem at a cost to
performance self-esteem. An opposite pattern of results accompanied the
tendency to claim racial discrimination as the cause for failure (Ruggerio et al.,
1997).
Similarly, the Personal-Group Discrimination Discrepancy (PGDD)
literature provides yet another depiction of how members of historically
oppressed groups tend to respond when asked about their encounters with
racism (see Foster & Matheson, 1999 for a review). This robust discrepancy
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
14
describes how persons from stigmatized minority groups are often more likely to
report the existence of group discrimination than of personal discrimination (see
Foster et al., 1999). For instance, results from a study conducted by Ruggerio
(1999) indicated that African American young adults exhibited quicker response
times when making assessments for group racism in a number of settings than
for personally-experienced racism across the same variety of settings.
Some scholars however, have questioned the validity of the dominant
PGDD phenomenon and have further extended this literature by illustrating how
individual differences in ethnic identity (or the strength of affiliation or sense of
belonging to one’s ethnic group) influences the endorsement of this discrepancy
(Operario & Fiske, 2001). More specifically, data revealed that the gap
between personal and group risk for reporting racist maltreatment narrowed the
more one tended to identify with their ethnic group. While the classic PGDD
finding emerged among ethnic minority individuals espousing lower levels of
ethnic group identity (Operario et al., 2001).
In addition, it is likely that variation in affiliation with one’s ethnic
minority group played a role in the tendency to either minimize (e.g. lower
ethnic identity) or emphasize (e.g. higher ethnic identity) racial maltreatment
within the academic setting. In this way, more specifically, ethnic identity may
buffer against the negative psychological consequences of racism only for those
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
15
who endorse stronger ethnic group affiliation. However, it may be argued that
these same individuals may be at higher risk for the social consequences linked
to attributing racism for failures (i.e. greater hostility and mistrust towards non
ethnic minority peers and colleagues).
These disparate attributions to racism findings may also be explained by
individual differences in perceived locus of control. With regard to the
attributional ambiguity results, it has been argued that the tendency to make
external attributions to racial prejudice is helpful in preserving self-esteem (e.g.
Crocker et al., 1989; Crocker et al., 1991). In effect, one is able to relinquish
responsibility for negative outcomes by labeling the cause of an event as racist
and thus not as a result of their own lack of skills or ability.
However, this predilection for removing personal controllability could
predict greater feelings of helplessness and/or resigned acceptance over time,
particularly if one has regular contact with people with whom they do not share
a similar ethnic background (see Shorey et al., 2002 for a discussion of the
relationship between perceptions of racism and diminished perceptions of
interpersonal control and lowered self-esteem in a Hispanic population).
Interestingly, when external attributions to race are made for positive events,
this phenomenon may also contribute to more long-term dangers to well-being
(e.g. self-doubt about the validity of performance capabilities). On the other
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
16
hand, what is gleaned from the racism minimization and PGDD evidence is that
in order to believe one has control over their own outcomes, one may tend to
dismiss or distance oneself from perceiving situations as involving racism
because skin color is a factor that cannot be controlled while performance and
self-presentation are factors that can be (Kaiser & Miller, 2001; Ruggerio et al.,
1997). However, these same individuals who tend to minimize their experience
of racism may be more vulnerable to encountering interpersonal conflict with
members of their own ethnic group. Once again, it is left to future investigators
in this area to more closely examine the relationships among ethnic identity,
specific components of self-esteem and perceived control within the attributional
ambiguity, racism minimization and PGDD literatures.
Threats to self-esteem in the short-term may be buffered by attributions
to racism (Crocker et al., 1991), but it may also be the case that confidence in
long-term self-efficacy when interacting in interracial environments may be
harmed (Ruggerio et al., 1997). As mentioned previously, racism appraisals
may also accentuate hostility, mistrust and aggression within the interracial
climate (Branscombe et al., 1999; see Major, Kaiser, & McCoy, 2003 for
results indicating how attributions to sexism did not mitigate feelings of
hostility). These effects could produce further risk of decline in physiological
and psychological dimensions of well-being for ethnic minority populations.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
17
However, it is also likely that racism attributions and frequent racism
experiences serve to create stronger bonds and feelings of closeness among
members within various ethnic minority communities. Thus, the racism
attribution and larger racism, stress and health literatures remain to more fully
address the complexities surrounding the interpersonal costs and benefits
resulting from appraising an event as racist from a functional developmental
perspective.
Building upon theory and findings discussed thus far, several personality
factors have been implicated in the literature as protecting against the corrosive
psychological effects of racism (see Clark et al., 1999 and Swim et al., 1998 for
comprehensive reviews of personality as well as other factors including coping).
However, ethnic or racial identity is considered to be one of the most widely
studied individual difference factors modulating the adversities of racism and
acculturative stress among people of color. Similar to the attributional ambiguity
literature, recent evidence has suggested that elements of ethnic identity may be
upshots of or a means of coping with the psychological and interpersonal
processes set in motion by racial discrimination and negative interracial
exposure (Branscombe et al., 1999; Contrada et al., 2001; Sanders Thompson,
1991; Wright & Littleford, 2002).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
18
For example, the rejection-identification model following from
traditional social identity theory posits that members of various ethnic minority
groups will tend to endorse stronger affiliation with their minority group as a
result of negative experiences with or perceived low probability of being
accepted by the dominant ethnic group (see Branscombe et al., 1999 for a
discussion). Thus exposure to racist events is itself asserted to be an important
part of the process of ethnic and racial identity development. This theme is also
evident in Cross’ (1971, 1991) well-known Nigrescence stage model of racial
identity formation in which encountering racial discrimination facilitates
questioning and transforming one’s orientation towards the dominant culture.
In support of these theories, ethnic identity was shown to mediate and/or
buffer the relationship between past experiences with racism and aspects of
psychological well-being in African American adult (Branscombe et al., 1999),
adolescent (Wong, Eccles & Sameroff, 2003) and school-age Mexican American
(Romero et al., 2003) participant groups. In another recent investigation, the
frequency of negative interracial experiences and perceived racial bias
meaningfully predicted aspects of ethnic identity within an ethnically-diverse
participant sample (Wright et al., 2002). This model has received additional
support among older adult (Gartska, Schmitt, Branscombe & Hummert, 2004),
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
19
international student (Schmitt, Spears & Branscombe, 2003) and gender diverse
samples (Schmitt, Branscombe, Kobrynowicz & Owen, 2002).
However, some research has failed to discern a significant effect of
racial identity in mitigating the negative impact of racism on psychological
health for people of color (Littleford et al., 1998). Other investigations have
shown aspects of racial or ethnic identity to be risk factors fostering the
development of racism-related psychological distress (Landrine et al., 1996) and
acculturative stress (Thompson et al., 2000) among certain ethnic minority
groups. The inconsistent findings regarding the role of ethnic and racial identity
in the field of racism and wellness may be a symptom of larger methodological
issues which highlight the need to enact more standard and unified methods of
measuring the complex constructs of ethnic identity and perceived racism.
In sum, the tendency to attribute racism may function to promote
psychological well-being for people of color in certain interracial contexts.
Similarly, the strength of ethnic group affiliation may serve to buffer against the
adverse effects of racism experiences on health and wellness for members of
particular ethnic minority groups. However, there may be a tradeoff of
enhancing vulnerability for other psychological (e.g. anger reactivity) and
interpersonal costs (e.g. hostility). Additionally, the arms of racism in impairing
the quality of life for persons of color may be farther reaching than perhaps first
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
20
anticipated. Within an acculturation framework, the following section will
address the idea that differential rates of participation in beneficial health
behaviors and endorsing positive health values by members of certain ethnic
groups may be in part due to the distal effects of racism.
Ethnicity, Health Values and Health Behaviors: An Acculturative Stress
Perspective on Moderating the Effects of Racism on Quality of Life
There is an expanding literature within the field of health psychology that
is providing evidence of the benefits of well-being promoting behaviors on
enhancing aspects of quality of life and sense of life satisfaction (e.g. Levy,
2003). For instance, one study found that individuals at later stages of exercise
change reported greater perceived quality of life than persons at earlier stages of
change (Laforge, Rossi, Prochaska, Velicer, Levesque, McHorney, 1999). A
similar positive link between quality of life estimates and health behavior
involvement emerged in an earlier investigation which included US Navy
personnel (Woodruff & Conway, 1992). Another analysis found that quality of
life was more strongly reflective of psychological versus physiological well
being (Vaez & Laflamme, 2003).
The relationship between stress and maladaptive health practices is well-
established. Greater stress levels are typically associated with less exercise/more
sedentary behavior, poorer dietary choices and increased cigarette smoking (e.g.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
21
Hudd, Dumlao, Erdmann-Sager, Murray et al., 2000; Ng & Jeffery, 2003;
Ogden & Mitandabari, 1997; Pak, Olsen & Mahoney, 1999; Steptoe, Wardle,
Pollard & Canaan, 1996; Wardle & Gibson, 2002). Given the impact of stress
on health practices, in recent years, from a psychosocial vulnerability
perspective, identifying inter-ethnic differences in adherence to positive health
habits (e.g. exercise, nutrition) and involvement in negative health behaviors
(e.g. alcohol and tobacco usage) has received increasing scholarly attention (e.g.
Berrigan, Dodd, Troiano, Krebs-Smith, & Ballard, 2003; Courtenay,
McCreary, & Merighi, 2002; Felton, Parsons, Misener & Oldaker, 1997; Pate,
Heath, Dowda, & Trost, 1996; Wise, Carmichael, Belar, Jordan & Berlant,
2001). For example, Lee and contributors found that self-reported dietary
intake, physical activity and smoking habits obtained in a large sample of 12-21
year olds varied depending upon ethnicity. The Black participants endorsed
poorer nutrition and lower smoking frequency and the Hispanic participants
tended on average to acknowledge healthier dietary practices, lower rates of
physical exercise and lower smoking frequency than their inter-ethnic peers (Lee
& Cubbin, 2002).
In another study, African American participants indexed higher mean
levels of tobacco, drug and alcohol consumption than European American
participants in the study sample (Harris, 2004). However, a few investigations
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
22
have failed to determine the existence of ethnic group differences in health
behaviors when other demographic (e.g. age, gender), SES and/or body mass
index (BMI) variables were controlled for (e.g. Appel, Harrell, & Deng, 2002;
Felton, Parsons, & Bartoces, 1997; Gallant & Dorn, 2001; Kim, Bramlett,
Wright, & Poon, 1998). Several authors assert that results in this area are often
confounded with status variables which have been consistently shown to affect
rates of health-preventive behavior participation through such mechanisms as
adequate access to appropriate health care (Hopp & Herring, 1999).
A few studies have also examined the specific relationship of racism
experiences to negative health habits in African American samples (e.g. Kwate,
et al., 2003; Landrine et al., 1996). In the earlier work, analyses revealed that
on average, a group of African American smokers tended to report higher levels
of racism-related experiences than did their non-smoking counterparts. Their
group of smokers also tended to appraise the racism they encountered as more
stressful than the non-smokers (Landrine et al., 1996). Additionally, Kwate and
colleagues learned that the frequency of past year racism events was
significantly associated with alcohol intake and the amount of cigarettes smoked
within a group of urban-dwelling African American women (Kwate et al.,
2003).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
23
It is also of interest to describe the findings of a more recent analysis that
also identified acculturative status and discriminatory experiences in
significantly predicting the level of engagement in health-enhancing behaviors in
a group of HIV-positive, immigrant Latino gay men (Bianchi, Zea, Poppen,
Reisen, & Echeverry, 2004). This interesting finding was further mediated by
coping style. More specifically, the socio-cultural variables of interest predicted
less active coping tendencies which in turn impacted the practice of beneficial
health behaviors (Bianchi et al., 2004).
Health beliefs and the stigma associated with HIV-positive status were
shown to significantly predict the attrition rate in an urban, community mental
health clinic (Reece, 2003). Moreover, a fatalistic belief orientation about illness
and inaccurate knowledge about cancer significantly predicted the length of
delay in care following an abnormal Pap smear among women of Asian and
Latino heritage (Nelson, Geiger, & Mangione, 2002). In contrast, Harris’ recent
investigation did not document significant racial group differences in health
values between her African American and Caucasian participants although a
hardy cognitive style was shown to interact with the level of health value to
predict personal distress in the African American group only (Harris, 2004). In
the same study, the relationship between the specific commitment component of
hardiness and health habit participation was moderated by the level of health
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
24
value endorsed in the ethnic minority sample (Harris, 2004). Other studies have
also not established ethnic group differences in health beliefs as predictive of
health-promoting behaviors (e.g. Thomas, Fox, Leake, & Roetzheim, 1996).
Although the findings in this literature are somewhat mixed, the current
study asserts (based on the findings obtained by Johnson, 1998 and Harris,
2004) that the role of individual differences may be a significant element
involved in clarifying variation in the relationships among health values and
positive health behaviors in certain ethnic minority groups. More specifically, in
addition to general coping style and health belief frameworks, the present study
encourages considering the utility of an acculturative stress-ethnic identity
perspective in understanding ethnic group differences in health beliefs or valuing
health-promoting behaviors (e.g. exercise, diet, nutrition) in affecting ethnic
group differences in participating in these behaviors (Hopp et al., 1999). For
example, one study found modest relationships among stage of racial identity
development, health value and adaptive health practices in a group of African
American participants (Johnson, 1998).
Proclamations and recommendations for changing health habits such as
physical activity level, smoking, drinking, drug use and overeating are often
disseminated by mainstream, middle class cultural vehicles such as physicians
and the media. These sources may be suspiciously deemed as part of the “White
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
25
establishment” by members of some ethnic minority communities. As a result of
a long history of institutionalized racism, these individuals may then fail to
seriously consider positive health behavior change. Additionally, for many
ethnic minority populations, following through on such health suggestions may
entail radically modifying lifestyle variables that have been traditionally
associated with ethnic group affiliation (e.g. diet) or once again dismissed as
what “White folks do”.
In order to enhance the efficacy of health behavior modification among
some ethnic minority groups, a health value acculturation process must take
place over time. Towards this end, it will be important to get the message across
in culturally-sensitive ways (e.g. using ethnic spokespersons, appealing to
family values, enlisting the assistance of local community organizations and
churches: Neighbors, Braithwaite & Thompson, 1995) that do not assume that
all individuals place a high value on longevity goals or on good long-term health
promoting practices. This may be particularly relevant among lower income
ethnic minority groups who tend to be more proximally focused on making it
through each day rather than focusing on preventing the emergence of more
distal health risks. Thus the present investigation hopes to start to shed light on
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
26
the roles of racism experiences and ethnic identity (akin to acculturation level)
in predicting health values and the practice of positive health habits within an
ethnically-diverse participant sample.
Hostility, Ethnicity and Health: A Functional Developmental Perspective on
Risk and Resilience
Hostility is described as a set of rigidly held interpersonal beliefs or a
dominant, cognitive, relational schema (Guyll & Madon, 2003) related to
affective (e.g. anger and its cousins: annoyance, resentment, contempt) and
behavioral (e.g. aggression) vulnerability (see Barefoot, 1992 for a review).
Smith (1994) defined hostility as: "a devaluation of the worth and motives of
others, an expectation that others are likely sources of wrongdoing, a relational
view of being in opposition toward others, and a desire to inflict harm or see
others harmed" (p. 26). In more colloquial terms, hostility is personified in the
suspicious, mistrustful and brooding cynic.
Based on an exhaustive compilation of the literature, there are at least
four primary pathways conceptualizing how hostility works to bring about
significant declines in physical health (Miller et al., 1996 and Smith, 1992 for
discussions of theoretical, measurement and other methodological issues). These
include health behavior (e.g. smoking, exercise), psychosocial vulnerability
(e.g. exposure to stress, interpersonal strain and availability of social support
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
27
buffers), psychophysiological reactivity (e.g. anger-induced cardiovascular and
endrocrinergic responsivity) and transactional formulations (i.e. a confluence of
psychosocial and physical reactivity models) each with varying amounts of
empirical support (see Miller et al., 1996).
However, very little research has tested these various models in
ethnically-diverse samples. The few exceptions have primarily evaluated the
existence of differential relationships between hostility and cardiovascular
disease or diabetes biological vulnerability factors in European American and
African American populations only (e.g. Fang & Myers, 2001; Harburg, Erfurt,
Hauenstein, Chape, Schull et al., 1973; Hughes et al., 2003; Jain, Dimsdale,
Roesch, & Mills, 2004; Saab, Llabre, Schneiderman, Hurwitz, Barry et al.,
1997; Surwit et al., 2002). The results of these analyses have been mixed.
Moreover, within the aforementioned psychosocial vulnerability model of
hostility, there is a paucity of systematic research that has given sufficient
attention to the role of socio-cultural factors in the creation, function and
meaning of a hostile interpersonal cognitive style and to differential stress
exposure among ethnic minority populations. Even less is known about the
interpersonal effects of hostility in general and specifically in ethnically-diverse
samples (e.g. Brondolo, Rieppi, Erickson, Bagiella et al., 2003; Cui, Rand,
Bryant & Elder, 2002; Davidson, Prkachin, Lefcourt & Mills, 1996; Gambone
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
28
& Contrada, 2002; Holman & Jarvis, 2003; Katz & Woodin, 2002). Therefore,
from a functional developmental perspective it is of interest to underscore both
the consistent finding of the link between ethnicity and hostility (e.g. Barefoot,
Peterson, Dahlstrom, Siegler, Anderson et al., 1991; Durel, Carver, Spitzer,
Llabre, Weintraub et al., 1989; Finney, Stoney & Engebretson, 2002; Harburg
et al., 1973; Hughes et al., 2003; Jain et al., 2004; Saab et al., 1997;
Scherwitz, Perkins, Chensney, & Hughes, 1991) and the robust positive
relationships indicated between racism and hostility (Branscombe et al., 1999;
Landrine et al., 1996) and between ethnic identity and hostility (Wright et al.,
2002). Interestingly, scholars in both literatures have not fully addressed the
importance of integrating findings in the areas of racism, ethnicity, hostility and
wellness. Thus, one principal hypothesis the current investigation poses is that
ethnic group differences in endorsing trait hostility and in the different patterns
of relationships evidenced between hostility and various outcome variables
across different racial groups may in part stem from variation in racism
exposure. Just as racially aversive experiences are involved in the process of
ethnic identity formation, they also likely play an important role in the
development of hostile beliefs within certain ethnic minority groups (see Price &
Glad, 2003 for a discussion of the impact of exposure to an abusive home
environment on the manifestation of a hostile attributional style in young
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
29
children; see Keltikangas-Jarvinen & Heinonen, 2003 for longitudinal evidence
of the predictive power of childhood parental behavior on subsequent hostility in
adulthood).
In this way, it is critical for researchers in this area to understand the
complex nature of hostility outside of typical European American study
populations. More specifically this may entail recognizing how hostility may
both function and be conceptualized as an adaptive, socially-sanctioned and
psychologically protective cognitive style among ethnic minority groups within
the specific context of racism-related stress and still remain a significant factor
implicated in enhancing risk for various serious physical health conditions. What
is construed by the larger, mainstream culture as a dysfunctional way of
thinking, the interpersonal themes inherent in hostile beliefs may in contrast be
considered a quite appropriate belief system to hold among people of color due
to the ever-present potential for encountering racism (see Hemmings, 2002 and
Miller, 2002 for a discussion of hostility from a context-dependent, dynamic
constructivist perspective; see Vontress & Epp, 1997 for a discussion of the
implications of “historical hostility” on the therapeutic relationship exhibited by
some African American clients entering mental health treatment).
Additionally, this cognitive style may further be associated with a greater
sense of cohesiveness or affiliation with members of one’s own ethnic group
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
30
particularly in the face of the ubiquitous threat of racial discrimination. Thus,
hostility may be considered another consequence of the rejection-identification
process that demonstrates both beneficial and detrimental effects on dimensions
of health and quality of life for ethnic minority individuals (see Branscombe et
al., 1999). More specifically, a rejection-protection process may be involved.
The next section will describe a specific form of anger coping that like
trait hostility may be associated in general with both subjective and objective
negative health indicators but may also like ethnic identity be linked to enhanced
psychological and interpersonal wellness among ethnic minority communities
within the specific context of racism-related experiences.
Anger Rumination and Health: Potential Cultural Influences on the Impact
of Anger Coping Style within the Context of Racism and Quality of Life
Anger is often the dominant emotion evoked by experiences of unjust
racist victimization (Armstead, 1989; Best, 2001; Moghaddam, Taylor, Ditto,
Jacobs & Bianchi, 2002) whose expression and experience are shaped by
cultural background and other socio-environmental factors (e.g. Chon, 2002;
Cole, Bruschi & Tamang, 2002; Ferrari & Koyama, 2002; Fields, Reesman,
Robinson, Sims, Edwards et al., 1998; Kim & Zane, 2004; Kino, 2000;
Koevecses, 2000; Ohbuchi, Kumagai & Atsumi, 2002; Ramirez, Fujihara, van
Goozen& Santisteban, 2001; Russell, 1989; Tanaka-Matsumi, 1995). This
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
31
affective state has also been shown to have negative relationships with aspects of
well-being (e.g. Diong & Bishop, 1999; Mahon, Yarcheski & Yarcheski, 2000)
and life satisfaction (e.g. Hong & Giannakopoulos, 1994) and has further been
indicated as an integral component involved in the development and maintenance
of a hostile relational schema which is implicated in the weakening of physical
health status.
How individuals cope with anger when provoked has been identified as a
possible vulnerability factor for sustained cardiovascular responsivity,
physiological debilitation (see Burns & Katkin, 1993) and lipid reactivity (see
Finney et al., 2002). More specifically, anger inhibition or anger suppression is
a widely-cited avoidant or passive anger coping style associated with elevated
physiological activity, hypertensive risk in response to stress in general and to
racism-related stress specifically (e.g. Armstead et al., 1989; Finney et al.,
2002; Harburg, Erfurt, Hauenstein, Chape, Schull etal., 1973; Krieger, 1990;
Moghaddam et al., 2002).
However, this concept does not fully capture the potential cognitive
processes activated in the aftermath of an anger-arousing event. The chronically-
potentiated physiological reactivity observed in certain populations while
recovering from an experimentally-induced anger manipulation may be the
result of thoughts and imagery about the event that are replaying through one's
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
32
mind. Such persistent cognitive effort could then maintain and prolong
physiological arousal as well as exacerbate feelings of anger (Rusting & Nolen-
Hoeksema, 1998). In this way, the content and process of anger rumination
itself may then become an additional stressor for the individual.
In light of this, researchers have conceptualized and promoted the anger
rumination construct as a potential mediator between anger suppression and
accentuated cardiovascular responding and as an important element preserving
hostile beliefs (Sukhodolsky, Golub, & Cromwell, 2001). Classic self-focused
rumination within the depression and anxiety literatures often takes the forms of
repeatedly attempting to understand why one is depressed or anxious, reviewing
memories and incidents consistent with negative mood, trying to come to some
meaning for one's adversities and catastrophizing over the possible ramifications
of mood disturbances (Lyubomirsky, Caldwell, & Nolen-Hoeksema, 1998).
Authors describe rumination as a maladaptive coping response that worsens as
opposed to alleviates negative mood, paralyzing instead of energizing persons to
take active steps to ameliorate symptoms (Lyubomirsky et al., 1998).
Ruminative processes in general have been useful in understanding gender
differences in the manifestation of depressive symptoms (Nolen-Hoeksema,
Larson, & Grayson, 1999), in eludicating the etiology of mood-congruent
memory recall biases among dysphoric populations (Lyubomirksy et al., 1998),
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
33
in investigating the augmentation of phobic anxiety (Davey, 1995) and in
predicting longitudinal PTSD status following motor vehicle accidents (Ehlers,
Mayou, & Bryant, 1998).
Anger rumination is presented as a stable affect-focused coping style. It
is regarded as a more specific mode of ruminative emotion regulation. An early
measure in this area was constructed by Caprara (1986) entitled the Dissipation-
Rumination Scale. The theory underlying this inventory specified dissipators as
individuals who tended to let go of anger while ruminators were more likely to
hold grudges and continuously brood over the wrongs done to them,
accentuating and extending feelings of anger (Caprara, 1986). Persons
exhibiting the slow dissipating, ruminating style were also proposed to have a
higher probability of seeking revenge on those blamed for the provocation
(Caprara, 1986). Anger rumination as indexed by this questionnaire was found
to positively covary with aggression (see Caprara, 1986; Collins & Bell, 1997).
More recently, Sukhodolsky and fellow investigators integrated themes
from the work of Nolen-Hoeksema, Caprara and others in formulating the
Anger Rumination Scale (Sukhodolsky et al., 2001). This measure is composed
of four factors: angry afterthoughts, thoughts of revenge, angry memories and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
34
understanding of causes. It was designed in order to assess individual
differences in the duration of anger experiences as a result of repetitively and
actively reprocessing interpersonal anger events (Sukhodolsky et al., 2001).
The influence of anger rumination on indices of physiological reactivity
and health appears to be dependent on the experimental paradigm utilized. In a
between-subjects investigation, Caprara and colleagues were not able to
differentiate normal controls from those with symptoms of coronary heart
disease based on rumination processes (Caprara, Barbaranelli, Colombo, Politi,
& Valerio, 1995). However, hostility and anger rumination were observed to
be among the strongest predictors of cardiovascular responsivity during a
mentally challenging laboratory stressor (Bermudez & Perez-Garcia, 1996).
Furthermore, the tendency to mentally rehearse distressing events was found to
accompany heightened cortisol secretion in response to completing an
examination among a group of student nurses (Roger & Najarian, 1998). It
would seem that experimentally creating a stressfiil environment may enhance
anger rumination processes rather than simply evaluating its association to
health outcomes within an unprovocative context. It is also possible that the
hospital environment, in which the participants from the Caprara study were
recruited, was an equally stressful situation for both groups, thus nullifying any
effect of anger rumination interacting with health status. Finally, a recent
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
35
analysis found that an anger rumination response style significantly interacted
with avoidant and assertive styles to predict variation in resting and ambulatory
blood pressure levels (Hogan & Linden, 2004).
In effect, rumination as an effortful cognitive process aids in keeping the
anger-eliciting memory "alive" and accessible in memory while extending
negative affect and potentially physiological arousal (see Topalli & O’Neal,
2003 for a discussion). However, it is possible that such rumination may distort
the accuracy of memories for anger events. This may occur when an individual
disproportionately focuses attention on the most provocative aspects of a
situation, thus strengthening those affective-recall associations. Among the
psychological costs perhaps of this coping style may involve a weakening of
attention paid to disconfirming or more innocuous information tied to the event
in memory along with a reinforcement of hostile schemas for future interactions
(see Topalli et al., 2003 for a review of evidence linking anger arousal to
cognitive phenomena including a narrowing of attention).
Following from this analysis, the current study proposes active anger
rumination coping to be a primary culprit in enhancing retrospective accounts of
racism and in modulating the impact of racism on dimensions of quality of life
even when trait hostility is controlled for. More importantly for the purposes of
this investigation, it is of interest to expand our awareness of the ethnic and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
36
cultural influences on the construal, function and relationship of anger
rumination to components of quality of life within the context of racism events.
As proposed in the previous section, although this emotion regulation style in
general has been shown to covary with or predict negative health and well-being
indicators, it may serve as a culturally-appropriate response and potential
psychosocial buffer against the realities of racism. This affect-focused form of
coping may further enhance interpersonal closeness within certain ethnic
minority groups specifically when racism-related events are involved. The
development of this anger coping style may be an additional consequence of the
current rejection-protection hypothesis which is an extension of the rejection-
identification model explaining the role of racism in the formation of racial and
ethnic identity in a variety of stigmatized groups.
Once again if viewed from behind a broader, multicultural lens, the
experience and expression of anger may be construed differently depending on
the prevailing socio-cultural influences and reinforcement contingencies extant
in the environment. The common stereotype of the “hostile, angry, brooding
Black man or woman” (e.g. Millner, Burt-Murray, Miller, 2004) when viewed
from a mainstream perspective appears anomalous and dysfunctional but when
viewed within the historical context of racial oppression from the Black
community’s standpoint, may be deemed the normatively adaptive response for
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
37
surviving in a continuously race-conscious world. Hostility and anger
rumination may like ethnic identity be considered the experiential-cultural glue
that creates a sense of in-group cohesion when racist acts are involved (e.g. the
unifying Civil Rights Movement in response to the legacy of slavery and
pervasive segregationist policies). The sad irony though is that what may be
beneficial in certain respects may also at the same time contribute to worsening
physical health in particular ethnic minority populations.
Perceived Racism, Hostile Attributions, Anger and Aggression: The
Potential Moderation of Contextual Attributions at the Situational Level of
Analysis
Dodge and co-investigators incorporated attributional causality and
behavioral responding for use with clinical populations as a basis for their social
information processing model (Dodge, 1980; Dodge, 1986). The socio-cognitive
phenomenon that has reliably distinguished aggressive from non-aggressive
cohorts was coined the "hostile attribution bias" (Nasby, Hayden, & DePaulo,
1979). This area of research has recognized that the process by which
individuals arrive at causality is not a perfect, rational science. Rather, there are
significant intervening factors (e.g. a hostile cognitive style) that may lead some
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
38
groups of individuals to have a tendency to process attributional information in a
biased manner such that they expect the worst in others particularly in situations
involving motivational uncertainty.
For example, when aggressive children and adolescents are confronted
with interpersonal scenarios in which the motivation underlying the behavior of
a peer is ambiguous, this group has been shown to infer hostile intentions on the
part of the peer to a greater extent than their non-aggressive counterparts
(Dodge & Somberg, 1987; Steinberg et al., 1983; Waas, 1988). The difference
between groups however is neutralized when the intentions of peers in the
provocations is clearly indicated as benign or hostile (see Steinberg et al., 1983
for a review). Authors have also observed that factors including situationally-
induced social anxiety (Dodge et al., 1987), being rejected by peers (Waas,
1988), experiencing social failure (Guerra, Huesmann, & Zelli, 1993) and
manipulating the relationship of the actor (e.g. best friend versus acquaintance
versus enemy; Ray & Cohen, 1997) influence attributing hostile intent when
encountering ambiguous social cues. While other experts have demonstrated that
children prone to prosocial behavior suppress the tendency to make hostile
interpretations when provoked by a peer (Nelson & Crick, 1999). This finding
was described as a "benign attributional bias" which may have important
implications for the formation of socially-appropriate behaviors and for reducing
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
39
interpersonal conflict (Nelson et al., 1999). The traditional paradigm utilized in
the hostile attribution bias literature involves presenting participants with a
series of vignettes that vary in the degree of ambiguity for the intentions of
another peer's behavior.
The research program developed by Graham and colleagues aimed to
more closely examine the mediation of the hostile attribution bias by affective
processes posited by Weiner's (1986) attribution-affect-action model (Graham,
Hudley, & Williams, 1992). A second goal was to broaden previous work by
focusing specifically on high risk, ethnic minority adolescent populations (e.g.
African American and Latino American). Consistent with hypotheses, anger was
cited as an important modulating factor linking the casual attributions and
proposed behavioral responses endorsed by the aggressive youth in this study.
Another important finding documented by this research group was that the high
degree of negative intentionality made by their aggressive African American
adolescent sample was resistant to a pre-experimental priming procedure
(Graham & Hudley, 1994; see Samples, 1998 for additional work in this area
with ethnic minority children). Further, these authors and others have provided
evidence for the influence of parental socialization or modeling as important
variables in the development of perceived hostile intentions in aggressive
children (Baden & Howe, 1992; see Graham, 1998 for a review).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
40
A handful of studies have examined analogous hostile attribution
processes in adult populations. Turkat and colleagues demonstrated how
perceived hostility under conditions of ambiguity discriminated individuals with
paranoid personality symptoms from controls (Turkat, Keane, & Thompson-
Pope, 1990). Trait hostility was found to covary with hostile attribution
tendencies among men and women differing in levels of job status (Flory,
Mathews, & Owens, 1998). Job prestige was a moderator of perceived hostile
intent and cardiovascular reactivity for the normotensive participants in this
sample. Specifically, within the technical and clerical positions, endorsement of
a hostile attribution cognitive set was a risk factor for high peak blood pressure
(Flory et al., 1998). In another recent analysis, the level of retaliatory
motivation (as indexed by scores on the Vengeance Scale) was shown to
augment the tendency to infer hostile motives on the parts of others in an
ambiguous interpersonal provocation paradigm (Topalli et al., 2003).
Research by Epps and Kendall (1995) identified two separable
components of the hostile attribution tendency with aggressive adults. These
included both an evaluation of purposefulness as well as attributing a hostile
affective quality to the cause of another's behavior. Their findings give further
support to the benign intent bias speculated by Nelson and Crick (1999) with
their sample of prosocial middle school children. More specifically, results
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
41
indicate that at lower levels of anger and aggression, individuals may attend
differentially to the hostile signals of others (Epps et al., 1995). The authors go
on to suggest that when confronted with a surfeit of hostile information in an
interpersonal exchange, non-aggressive adults may be sensitized to filter out
those cues and focus on even a modicum of non-hostile stimuli within the
context of the scenario. A reverse pattern was posited for those with more
aggressive information processing styles (Epps et al., 1995). This intriguing
phenomenon was replicated in a recent study examining aggression and hostile
attributions while driving (Matthews & Norris, 2002).
Hostile attributions elicited by ambiguous interpersonal situations have
not been systematically investigated in relation to attributions of causality, anger
and aggression within the specific context of perceived racism in ethnic minority
populations. Might attributions to racism be considered an ethnicity-specific type
of hostile attribution? This portion of the present study is informed by previous
work that has emphasized the cumulative impact of numerous everyday,
ambiguous interpersonal situations that ethnic minorities encounter as
microstressor events that may be attributed to racism (Essed, 1990; Feagin,
1991, Feagin et al., 1994; Swim et al., 1998). Similarly, from a functional-
developmental perspective, when examining hostile attribution biases in
ethnically-diverse samples as some authors have done previously, could
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
42
investigators in this area be overlooking how contextual factors such as prior
racism experiences influence participants’ responses particularly if the ethnicity
of the person in the vignette is not clarified? Perceptions of racism in these
cases may serve as a self-esteem protective albeit anger and aggression-
exacerbating function as discussed in a previous section.
Furthermore, the hostile attribution bias is a robust entity that has been
found with child, adolescent and adult populations varying in level of aggressive
proclivities. Similar to its self-report cousins, this protocol has been useful in
demonstrating the link between chronically-activated perceptions of hostile
intent under conditions of ambiguity and aspects of health status. However,
interestingly, this paradigm has yet to be implemented in studying the relation of
perceived hostile intentions to perceptions of negative outcomes attributed to
racist maltreatment and its impact on psychological factors (e.g. anger and
aggressive intent) in the aftermath of these events.
One exception that warrants mentioning are the findings of a study
conducted by Branscombe and colleagues (Branscombe et al., 1999). This was
the one study found within the racism-related stress literature which formally
examined hostile attitudes towards the dominant group as a function of the
willingness to attribute racism in ambiguous interpersonal situations
(Branscombe et al., 1999). Their measure of hostility included items such as:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
43
"When I see White people on the street, I can't help but think bad things about
them" and "I use terms like 'white trash', 'redneck', or other names in
reference to White people" (Think of the caricature of the sitcom character
George Jefferson). This scale is clearly describing hostility directed outward
rather than assessing perceptions of hostility from others.
One limitation of this research is that it did not additionally evaluate a
general hostile cognitive style among their African American participants. It
may be the case that the respondents in their study hold a more characterlogical
hostile cognitive approach when interacting in situations that transcend race
differentials. Although a situational paradigm was utilized, this study did not
include an examination of the affective, automatic thoughts one has as a
situation involving possible racism unfolds nor how these cognitions are related
to inferences made about the intentions of others and to the affect (e.g. anger)
and cognition (e.g. aggressive ideation) reported in response to such an event.
Based on an integration of the hostile attribution bias and the perceived racism
bodies of research, it is important to the current project to take a more fine
grained process approach to the study of affect, cognition and intention when
ethnic minorities are faced with the uncertainty of modern racism.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
44
In conjunction with this, the corpus of work in perceived racism and
health has neglected to attend to how possible interethnic differences in
attributions to racism and its affective and somatic consequences may be related
to aspects of culture. For instance, published research has cited how
attributional style interacts with cultural background when making explanations
for interpersonal successes and failures (Anderson, 1999; Lee, & Seligman,
1997; Park & Kim, 1998) and in impacting expressed emotion and
schizophrenia relapse (e.g. Yang, 2003). Findings from cross-cultural studies
have further highlighted elements identifying how culture influences the
dispositional versus situational attributions one makes for the actions of others in
a given ambiguous scenario (Hong, Morris, Chiu & Benet-Martinez, 2000;
Knowles, Morris, Chiu, & Hong, 2001; Lee, Hallahan, & Herzog, 1996).
More specifically, as mentioned at the outset of this paper, based on a
dynamic constructivist perspective, bicultural young adults from Hong Kong
demonstrated a “frame-switching” phenomenon in which their average
attributional endorsements for ambiguous behavior were influenced by the
content of a cultural priming paradigm (Hong et al., 2000). When these
individuals received primes involving Western or American cultural symbols
they tended to attribute dispositional factors to a greater extent than contextual
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
45
factors as reasons for the behavior of others in an ambiguous vignette. An
opposite pattern of results emerged when participants were subliminally primed
with traditional Chinese cultural symbols (Hong et al., 2000).
With this evidence in mind, little empirical work however, has closely
examined the effects of considering situational or contextual factors as
alternative explanations when appraising an ambiguous situation as racist on an
individual’s anger experience and endorsement of aggressive intentions within
the racism-wellness literature. In light of this gap in the scientific discourse,
from a cognitive-behavioral (Beck, 1976; Ellis, 1962) and interpersonal
therapeutic perspectives, it is of particular interest to the present study to more
closely examine the psychological benefits of considering motives in addition to
racism as underlying another's behavior in potentially ameliorating the anger
intensity and aggressive thinking associated with perceptions of racism.
METHOD
Participants
Sixty-seven African American (56 females), 82 European American (65
females), 80 Asian American (66 females) and 71 Latino American (55 females)
undergraduate students ages 18-30 from the University of Southern California
community took part in this study. The groups did not differ significantly on the
ratio of male:female participants [Pearson’s X2 = 1.13, p (NS), d f = 3]. The
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
46
mean age was 19 for the entire sample. Student participants were American
born with English as a primary language. Participants were recruited through
the USC Psychology Department Subject Pool, through fliers posted around the
USC campus and through advertisements placed in the Daily Trojan.
Participants received either two hours of psychology course credit or $10 for
their time.
Measures
Demographic Questionnaire. This questionnaire requested standard
socio-demographic information from each participant including: age, gender,
ethnicity, year in school, parents' education levels, and annual household
income. It also included a one-item general measure of degree of ethnic group
affiliation (i.e. ethnic identity). This global measure of ethnic identity asked how
strongly the participant identified with his or her endorsed ethnic group on a
scale from 1 “not at all” to 7 “extremely”.
Ambiguous Interpersonal Vignettes. A series of 10 vignettes were
created, each depicting an ambiguous interpersonal interaction or event
involving a negative outcome that the participant imagined him or herself taking
part in. Vignette content was developed from and inspired by the following
measures: the Index o f Race-related Stress (Utsey et al., 1996), the Racism
Reaction Scale (Thompson et al., 1990), along with the vignettes used in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Al
Branscombe et al. (1999) and Inman et al. (1996) studies. After a larger pool of
vignettes was created, they were qualitatively rated by 5 research assistants on
dimensions such as authenticity and ambiguity.
Following imagining each 2-3 sentence long vignette, the participant was
first asked to rate the likelihood on a seven point scale (from 1 = not at all
likely to 7 = almost definitely) that they would have a particular thought in
immediate reaction to what they just imagined. There were three automatic
thought types each participant rated: humor/sarcasm, verbal hostility and
resigned acceptance. The verbal hostility items were adapted from the Hostile
Automatic Thoughts scale (Snyder et al., 1997). There were five sets of
automatic thoughts created, each used twice.
In the second part of these vignettes, participants rated the likelihood of
each of a series of seven statements on a seven point scale (from 1 = not at all
likely to 7 = almost definitely). Statement one asked for the likelihood that what
happened in the vignette was due to race or ethnicity (racism
attributions/perceived racism: alpha = .91). Statement two asked for the
likelihood that what happened in the vignette was due to more situational,
impersonal or constructive reasons (situational attributions: alpha = .76).
Statement three asked about the likelihood of perceived hostile beliefs of the
person in the vignette towards the participant. Statement four asked about the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
48
likelihood of perceived hostile intentions of the person in the vignette towards
the participant (hostile attributions: alpha = .87). Statement five asked about the
likelihood of generalized negative expectations for racist treatment in the future
in a similar situation. This is followed by two statements that inquire about the
likelihood that the participant would have physically aggressive and vengeful
thoughts in reaction to what happened (aggressive ideation: alpha = .92). These
items were also adapted from the Hostile Automatic Thoughts Scale (Snyder et
al., 1997).
In the third and final portion of the vignette questionnaire, participants
were requested to appraise how stressful the event would be for them if it
happened in real life on a seven point scale (from 1 = not at all to 7 =
extremely). Participants were also asked to rate how much what took place in
the vignette would anger them afterwards if it happened in real life (from 1 =
not at all to 7 = extremely) (post-event anger: alpha = .87).
In order to reduce the likelihood of other demographical confounds such
as gender, there were two versions of these vignettes, one female and one male.
Also, despite the number of variables collected from this questionnaire, only the
variables of interest were included in subsequent statistical analyses (i.e.
perceived racism, situational attributions, hostile attributions, aggressive
ideation and post-event anger). See Appendix for sample vignette.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
49
Schedule o f Racist Events (SRE; Landrine et al., 1996). This 18-item
self-report measure assesses an individual's past experiences (i.e. past year and
lifetime prevalence) with racial discrimination in a variety of settings (e.g.
treated unfairly by teachers and professors, by people in service jobs, by
neighbors, etc. because of your ethnicity) on a scale from 1 (never) to 6 (almost
all of the time). It also has participants rate how stressful they perceived each
racism incident to be on a scale from 1 (not at all) to 6 (extremely). Internal
consistency reliabilities have been documented as: .95 for past year events, .95
for lifetime events and .94 for stress appraisal (Landrine et al., 1996). This
measure was originally developed on an African American sample. For the
purposes of this study, item wording was changed to be appropriate for use with
any racial or ethnic group. In the analyses involving the evaluation of hostility,
anger rumination and health values as moderators, the frequency of lifetime
racist events was used as a measure of the cumulative impact of racism for the
individual. For the analyses involving testing the moderation by various positive
and negative health behaviors, the racism stress appraisal scale was used as the
index of racism impact based on a stress and coping perspective. The frequency
of past year racism scale was not included in the analyses for this study.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
50
The Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992).
This is a shorter, factor-analyzed version of the original Buss-Durkee Hostility
Inventory (BDHI; Buss & Durkee, 1957). This 29-item scale measures
dimensions of trait aggression which include subscales for Physical Aggression
(alpha = .85), Verbal Aggression (alpha = .72), Anger (alpha = .83) and
Hostility (alpha = .77). Respondents use a five-point Likert scale ranging from
1 (extremely uncharacteristic of me) to 5 (extremely characteristic of me) when
endorsing each item. Only the 8-item hostility scale was used in the subsequent
statistical analyses.
Anger Rumination Scale (ARS; Sukhodolsky et al., 2001). Anger
rumination tendencies are measured along four dimensions. These include:
Angry Afterthoughts (6 items) "Memories of even minor annoyances bother me
for a while", Thoughts o f Revenge (4 items) "I have difficulty forgiving people
who have hurt me", Angry Memories (5 items) "I think about events from a
long time ago and they still make me angry" and Understanding of Causes (4
items) "I think about the reasons people treat me badly". Responses are rated on
a frequency scale from almost never (1) to almost always (4). Overall internal
consistency reliability for the complete scale was found to be .93 and a stability
coefficient of .77 was cited (Sukhodolsky et al., 2001). The overall total score
was used as the index of a general anger rumination coping style for this study.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
51
Inventory o f College Students' Recent Life Experiences (ICSRLE; Kohn
et al., 1990). This factor analytically derived measure of takes a daily hassles
approach to perceived stress. Its items are specifically tailored to capture events
relevant to the college student's experience. Respondents are required to
evaluate the intensity of each event within a timeframe of the previous month on
a scale from 1 (not at all) to 4 (very much). The seven factors of this
questionnaire include: Developmental Challenge (e.g. "Struggling to meet the
academic standards of others", Time Pressure (e.g. "Not enough time for
sleep"), Academic Alienation (e.g. "Finding courses uninteresting"), Romantic
Problems (e.g. "Conflicts with boyfriend/girlfriend/spouse"), Assorted
Annoyances (e.g. "Having your contributions overlooked"), General Social
Mistreatment (e.g. "Being ignored") and Friendship Problems ("Being let down
or disappointed by friends"). The overall alpha internal consistency reliability
coefficients for the total scale were demonstrated as .88 for men and .89 for
women (Kohn et al., 1990). Only the overall score for this measure was used as
the index of general college life stress for this participant sample.
Health Behaviors Scales (HBS; Cassidy, 2000). These are a series of
short indices measuring the value of certain health practices (e.g. exercise),
perceptions of general health and frequency of illness experienced and frequency
of beneficial (e.g. exercise, nutrition) and detrimental (e.g. smoking, use of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
52
alcohol) health behaviors. They are adapted from Cassidy (2000). The internal
consistency reliabilities for these measures range from .51-.84. However, for
purposes of this study, several items were re-worded and the range of scale
responses were revised so as to make them continuous versus dichotomous (i.e.
yes-no). The health value scale achieved a coefficient alpha of .86. The
frequency of two negative health behaviors was used in this study as potential
moderators in the relationships between racism and dimensions of quality of life.
These included a one-item measure of the frequency over the past month of
drinking until feeling ill and a one-item assessment of smoking frequency within
the last four weeks. Finally, the 16-item positive health behavior scale was used
as a measure of the frequency of participating in positive health habits (alpha =
.80). See Appendix for full questionnaire.
Life Satisfaction Questionnaire-34 (LSQ-34; Carlsson et al., 1996). This
34-item questionnaire uses seven point scales to measure interpersonal,
individual and health-related dimensions of quality of life. It was originally
developed for use with breast cancer patients. The subscales of this inventory
consist of physical symptoms (e.g. "heart palpitations"), sickness impact (e.g.
"muscular weakness"), quality of everyday activities (e.g. "level of activity"),
socio-economic situation (e.g. "housing situation"), family relationship
closeness (e.g. "emotionally satisfying") and close friend relationship closeness
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
53
(e.g."meaningful"). Alpha coefficients for the subscales from this questionnaire
range from .67-.90. A general indicator of overall perceived health and quality
of life used in the present study consisted of five items. These included two
items from this measure and three items from the Health Behavior Scales (alpha
= .80). The sickness impact and physical symptom subscales from this
instrument were combined to form the somatic distress quality of life dimension
used in this study (alpha = .80). The family member and friendship closeness
variables were also used as dependent variables to examine the specific
relationships of racism, hostility, health behaviors and anger rumination on
these interpersonal quality of life dimensions.
Procedure
All participants were met by a representative (either the author or a
trained research assistant) of the USC Cognitive Studies in Clinical Science
Laboratory. Participants were seated in groups of 5-10 in a classroom in the
USC Human Relations Center. Each participant was first asked to provide
informed consent. In order to promote honest and accurate responding,
confidentiality of responses was assured. Following this, the procedure and
study instructions were reviewed with each participant. All measures were
administered in a fixed order (i.e. Demographic Questionnaire, LSQ-34, SRE,
HBS, Ambiguous Interpersonal Vignettes, ICSRLE, ARS, BPAQ). After
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
54
completing the batch of questionnaires, participants were fully debriefed and
thanked for their time and efforts. Prior to exiting the classroom, students
received either 2 hours of subject pool extra credit points or $10.
Statistical Analyses
For all statistical analyses the Bonferroni correction was used to control
for Type I error for each set of analyses (i.e. correlations, MANOVAs, multiple
linear regressions) given the large number of analyses performed. In addition,
the protected F stepped approach was used to determine whether or not to
further interpret regression findings. Pearson’s bivariate correlation analyses
were used to test for predicted associations among the dependent and
independent variables. MANOVAs were performed to identify any significant
ethnic group differences on the primary variables of interest. Finally, a series of
hierarchical linear regression models were conducted in order to evaluate
principal predictor and moderator hypotheses. The SPSS version 10.0 statistical
software package was utilized to perform all analyses for this study. More
specifics regarding these analyses are described throughout the subsequent
results section.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
55
Hypotheses
Based on the prior review of the relevant literatures and the aims
described throughout, the current project anticipates the following findings:
I. Significant Associations
a. The frequency of lifetime racist experiences and racism-related
stress will be positively associated with ethnic identity, hostility,
anger rumination, negative health behaviors, somatic distress,
general college life stress, perceived racism, post-scenario anger and
aggressive ideation in the total sample. These same variables are
expected to be negatively related to general perceived health/quality
of life, situational attributions, health values and the frequency of
participating in positive health behaviors for the entire sample.
b. Anger rumination and hostility are anticipated to be positively
related to negative health behaviors, general college life stress,
perceived racism, post-vignette anger, aggressive ideation, somatic
distress but negatively related to situational attributions, health
values, positive health behaviors, general perceived health/quality of
life, family member closeness and friendship closeness for the total
sample.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
56
c. Ethnic identity is also anticipated to be positively associated with
family member closeness and friendship closeness for the full
participant sample.
d. Health values were expected to positively covary with the frequency
of taking part in health-promoting behaviors.
e. At the situational level, it is predicted that perceived racism/racism
attributions will strongly positively covary with hostile attributions,
anger and aggressive ideation within the ambiguous interpersonal
scenarios for the full participant sample. In contrast, the frequency
of situational attributions is expected to covary with the same
variables in the opposite manner across participant groups.
II. Significant Ethnic Group Differences
a. It was anticipated that the African American participants would on
average tend to endorse the highest reports of prior lifetime racism
experiences, racism-related stress, perceived racism, trait hostility,
hostile attributions, post-scenario anger and aggressive ideation. This
ethnic group was predicted to yield the lowest endorsements of
health values, positive health behavior participation and situational
attributions within the ambiguous vignettes.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
57
b. The Asian American and European American participants were
expected to index higher mean levels of situational attributions than
the other two ethnic groups.
c. The European American participants were anticipated to endorse the
highest mean level of health values and positive health behavior
participation.
d. The ethnic minority participants were predicted to index higher mean
endorsements of ethnic group affiliation than the European American
participants.
III. Significant Predictors
a. The frequency of lifetime racist events and racism-related stress will
negatively predict general perceived health/quality of life and the
frequency of participation in positive health habits within the ethnic
minority participants while controlling for covariates.
b. The frequency of lifetime racist events and racism-related stress will
positively predict somatic distress, family member closeness,
friendship closeness, post-scenario anger and aggressive thinking
within the ethnic minority participant groups while controlling for
covariates.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
58
c. Ethnic identity will also positively predict family member closeness,
friendship closeness and general perceived health/quality of life for
the ethnic minority participants while controlling for covariates.
d. General college life stress, trait hostility and anger rumination
coping will positively predict somatic distress, post-scenario anger
and aggressive ideation for all ethnic groups.
e. The frequency of negative health habits (i.e. drinking until feeling
ill, smoking) are anticipated to positively predict somatic distress
while adjusting for covariates for all ethnic groups.
f. General college life stress, trait hostility, anger rumination coping
and negative health behaviors are anticipated to negatively predict
perceived health/quality of life, family member closeness, and
friendship closeness across ethnic groups.
g. The frequency of participation in beneficial health practices is
expected to positively predict general perceived health/quality of
life, family member closeness and friendship closeness for all ethnic
groups while controlling for covariates of interest.
h. The frequency of positive health behavior participation is anticipated
to negatively predict somatic distress across all ethnic groups while
adjusting for covariates.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
59
i. The frequency of perceived racism made in response to the
ambiguous vignettes would positively predict hostile attributions,
post-event anger and aggressive ideation particularly for the ethnic
minority participants while controlling for covariates.
j. The frequency of situational attributions made in response to the
ambiguous scenarios was expected to negatively predict hostile
attributions, post-event anger and aggressive thinking primarily for
the Asian American and European American participant groups
when important covariates were included in the regression model.
IV. Significant Moderators
a. Ethnic identity, hostility and anger rumination will each significantly
positively interact with the frequency of lifetime racist events for the
ethnic minority participants (i.e. particularly for the African
American group) in predicting dimensions of quality of life (i.e.
general perceived health/quality of life, friendship closeness and
family member closeness). In other words, as the frequency of
lifetime racist events increases, the quality of life dimension will also
increase at high levels of ethnic identity, hostility or anger
rumination.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
60
b. Ethnic identity, hostility and anger rumination will each significantly
negatively or positively interact with the frequency of lifetime racist
events or the level of racism-related stress for the ethnic minority
participant groups (i.e. particularly for the African American group)
in predicting variation in somatic symptom distress. Thus, at low
levels, these factors are expected to be buffers against the negative
cumulative impact of racism on the somatic quality of life
dimension. However, at high levels, it is likely that these variables
may exacerbate the relationship between one’s prior history of
racism and current self-reported physical symptoms.
c. Ethnic identity is expected to negatively interact with the frequency
of lifetime racist events to predict the frequency of participating in
health-promoting behaviors among the ethnic minority participant
groups (i.e. especially evident for the African American sample).
d. Ethnic identity is anticipated to positively interact with racism-
related stress to predict perceived family member and friendship
closeness among the ethnic minority participants (i.e. particularly for
the African American participants).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
61
e. Health values are anticipated to positively interact with the frequency
of lifetime racist events and ethnic identity in predicting variation in
the frequency of participation in health-promoting practices among
the ethnic minority participant samples (i.e. primarily evident in the
African American sample). More specifically, it is anticipated for
the ethnic minority participants that the frequency of health-
enhancing behaviors will positively regress on the frequency of
lifetime racist events at the combination of low ethnic identity and
high health value.
f. The frequency of participating in positive health behaviors will
positively interact with the degree of racism-related stress reported
in predicting general perceived health/quality of life, family member
closeness and friendship closeness for the ethnic minority
participants.
g. The frequency of taking part in positive health behaviors is also
anticipated to negatively interact with the degree of racism-related
stress endorsed in predicting somatic symptom distress for the ethnic
minority participant groups.
h. The frequency of taking part in negative health behaviors (i.e.
drinking until feeling ill, smoking) is expected to negatively interact
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
62
with the degree of racism-related stress to predict the variance in
general perceived health/quality of life, family member closeness
and friendship closeness. As racism-related stress increases, these
three quality of life dimensions are expected to decrease at high
levels of negative health behaviors or to increase at low levels of
negative health behaviors among the ethnic minority participants,
i. The frequency of taking part in negative health behaviors is
anticipated to positively interact with the degree of racism-related
stress endorsed to predict somatic distress scores for the ethnic
minority participants,
j. The frequency of situational attributions made in the ambiguous
vignettes will negatively interact with the frequency of perceived
racism to predict both anger and aggressive ideation. Thus, the
willingness to consider contextual alternatives may buffer against the
negative impact of perceived racism in everyday, ambiguous
situations. These moderation effects are primarily anticipated for the
Asian American and European American participant groups. It is
also possible that ethnic identity may combine in a three-way
interaction in buffering against anger and aggression in response to
these vignettes for the Asian American participants.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
63
k. There are no specific hypotheses regarding the interaction of ethnic
identity with any of the other moderating variables (e.g. hostility,
anger rumination, health behaviors) to predict quality of life
dimensions. However, any significant findings may assist future
research in further clarifying more specific models of the roles of
individual differences in affecting the relationship between racism
and quality of life within a variety of ethnic group samples.
RESULTS
Interrelationships Among the Independent and Dependent Variables
Table 1 provides the means, standard deviations and ranges for all of the
study variables examined. Pearson’s bivariate correlations were employed in
order to assess basic linear relationships among the variables being investigated
in this study. In general, relationships were demonstrated to be in the expected
directions for the total participant sample. As displayed in Table 2, the racist
events and racism stress-appraisal variables for the most part were found to be
strongly negatively associated with perceived health/quality of life, health values
and positive health behaviors. In contrast, significant positive relationships
emerged among the self-reported racism variables and the level of overall
college life stress, hostile beliefs, anger rumination tendencies and current
somatic distress for the entire study sample (see Table 2 for specific correlation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
64
Table 1: Descriptive Statistics for the Primary Study Variables for the Full
Sample
Variable Mean SD Range
Age 20.0 1.69 18-30
Education 14.2 1.14 12-17
Mother’s Education 14.6 4.26 0-24
Father’s Education 14.9 4.73 0-22
Ethnic Identity 5.59 1.22 1-7
Racism-related Stress 39.3 17.9 17-101
Lifetime Racist Events 32.5 12.0 17-77
College Life Stress 109 22.0 61-187
Quality of Life 21.6 3.60 9-29
Somatic Distress 29.9 9.81 13-66
Family Member Closeness 25.6 4.77 7-35
Friendship Closeness 26.7 4.01 13-35
Health Values 89.7 14.1 33-117
Positive Health Behaviors 40.4 9.18 18-71
Drinking Until Feeling 1 1 1 1.49 0.70 1-4
Smoking Frequency 1.31 0.87 1-5
Hostility 18.4 6.11 8-38
Anger Rumination 38.4 10.5 20-71
Perceived Racism 26.7 13.4 10-66
Situational Attributions 48.4 8.71 10-66
Hostile Attributions 26.3 11.6 10-68
Anger Reactivity 33.9 11.0 11-69
Aggressive Ideation 39.6 18.6 20-114
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 2: Correlations Among Primary Study Variables for the Full Sample
El LRE RRS GCS HV PHB DR SM QOL SOM FC FRC HOS AR PR SAT HAT . ANG AGGH
LRE .20 1.0 .77 .40 NS NS NS NS -.21 .32 NS NS .25 .26 .63 -..31 .54 .44 .24
RRS NS
—
1.0 .41 NS NS NS NS -.24 .34 NS NS .26 .26 .57 -.24 .47 .46 .24
GCS NS
— —
1.0 NS -.17 NS NS -.46 .56 NS NS .55 .51 .38 -.29 .38 .45 .35
HV NS
— — —
1.0 .64 NS -.17 .36 NS NS .21 -.18 NS NS .22 NS NS NS
PHB NS
— — — —
1.0 NS NS .51 NS .23 .30 -.27 -.21 -.17 .28 NS NS NS
DR NS
— — — — —
1.0 .30 NS NS NS NS NS NS NS NS NS NS NS
SM NS
— — — — — —
1.0 NS NS NS NS NS NS NS NS NS NS NS
QOL NS
— — — — — — —
1.0 -.43 .19 .28 -.45 -.38 -.27 .25 -.26 -.26 -.25
SOM NS
— — — — — — — —
1.0 NS NS .36 .39 .28 -.17 .26 .31 .17
FC NS
— — — — — — — — —
1.0 .43 NS NS NS NS NS NS NS
FRC NS
— — — — — — — — — —
1.0 NS NS NS NS NS NS NS
HOS NS
— — — — — — — — — — —
1.0 .65 .29 NS .35 .43 .37
AR NS
— — — — — — — — — — — —
1.0 .27 NS .32 .45 .39
PR .20
— — — — — — — — — — — — —
1.0 -.50 .81 .64 .59
SAT ■ -.25
— — — — — — — — — — — — — —
1.0 -.43 -.35 -.22
HAT NS
— — .. . . — — — — — — — — — — — —
1.0 .74 .59
ANG NS
— — — — — — — — — — — — — — — —
1.0 .63
AGGID NS
— — — — — — — — — — — — — — — — —
1.0
All correlations reported are significant at p < / = .001. NS = non-significant. EI=ethnic identity; LRE=lifetime racist events;
RRS=racism-related stress; GCS=general college stress; HV=health values; PHB=positive health behaviors; D R =
drinking frequency; SM=smoking frequency; QOL=overall life satisfaction; SOM=somatic distress; FC=family member closeness;
FRC= friendship closeness; HOS=trait hostility; AR=anger rumination; PR=perceived racism; SAT= situational attributions;
HAT=hostile attributions; ANG=anger reactivity; AGGID=aggressive ideation.
66
coefficients and levels of significance). Additionally it is of importance to note
that one’s prior history of racism experiences, the stress associated with those
events, hostility and anger rumination were also shown to significantly
positively covary with the frequency of perceived racism, hostile attributions,
the degree of anger response and aggressive ideation reported in the everyday,
ambiguous scenarios (see Table 2). Also as anticipated, an opposite pattern of
relationships was specified among those same variables and the frequency of
situational attributions endorsed in the ambiguous interpersonal vignettes (see
Table 2).
Ethnic Group Mean Differences
Demographic Characteristics. An Ethnicity (4) x Demographic (5)
MANOVA was conducted to identify potential ethnic group differences on status
and ethnic identity variables. The MANOVA yielded a significant main effect
term for ethnicity, F (15, 882) = 5.64, p < .01. Univariate ANOVAs
generated significant effects of ethnicity for mother’s education level [F (3, 296)
= 17.78, p < .01)], father’s education level [F (3, 296) = 11.23, p < .01]
and for ethnic identity [F (3, 296) = 7.70, p < .01]. Mean ages [F (3, 296) =
.10, p > .1] and years of education [F (3, 296) = .37, p > .1] did not differ
among the four ethnic groups. Due to differences in sample size among the four
ethnic groups, equal variances were not assumed and Dunnett’s T3 procedure
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
67
was used to perform post-hoc multiple comparisons among the observed means.
Table 3 displays the statistical characteristics for the demographic variables in
more detail including Cohen’s d as an indicator of effect size.
Additionally, a Chi-square analysis revealed significant ethnic group
differences for annual household income (Pearson’s X2 = 62.24, p < .01, d f =
15). As depicted in Table 4, a lower percentage of ethnic minority participants
had household incomes in the higher brackets (e.g. > $100,000) than the
European American participant group. Similarly, the Latino American
participant group in particular evidenced a greater percentage of annual
household incomes in the lower ranges (e.g. $10,000-$29,999) than was
observed among the other participant groups (see Table 4).
Racism, Personality, Stress and Health-related Measures.2 An Ethnicity
(4) x Measures (19) MANOVA yielded a significant multivariate term for
ethnicity, F (57, 840) = 3.50, p < .01). In accordance with original hypotheses
and previous research in this area, univariate ANOVAs indicated significant
ethnic group differences for the following variables: racism-related stress [F (3,
296) = 16.25, p < .01], the frequency of past-year racism events [F (3, 296)
2
This group of analyses was run both with and without controlling for mother’s level of
education as an SES covariate. Both sets of analyses yielded highly significant main effects of
ethnicity and a similar pattern of findings emerged. Thus, the results reported here in more
detail are from the MANOVA that was not adjusted for mother’s level of education. Post-hoc
comparisons were also performed using Dunnett’s T3 procedure for unequal variances.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3: Ethnic Group Differences on Demographic Variables
Variable Mean SD F (total) p-value d (group:
Age .103 NS N/A
A f Am (1) 20.0 1.99
Eu Am (2) 20.0 1.79
As Am (3) 19.9 1.52
Lat Am (4) 20.1 1.47
Education .492 NS N/A
A f Am 14.2 1.26
Eu Am 14.1 1.16
As Am 14.2 1.16
Lat Am 14.2 1.00
Mother’s Education 17.8 <.001
A f Am 15.1 3.14 .73 (1 > 4 )
Eu Am 16.0 2.66 .98 (2 > 4 )
As Am 15.3 3.12 .80 (3 > 4 )
Lat Am 11.6 6.11
Father’s Education 11.2 <.001
A f Am 14.6 3.75
Eu Am 16.1 3.79 .74 (2 > 4 )
As Am 16.1 3.80 .74 (3 > 4 )
Lat Am 12.4 6.31
Ethnic Identity 10.8 <.001
A f Am 6.19 1.16
Eu Am 5.36 1.09 .71 (2 < 1)
As Am 5.53 1.12 .56 (3 < 1)
Lat Am 5.35 1.35 .64 (4 < 1)
O n
00
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 4: Current Annual Household Income Percentages by Ethnic Group
Income Bracket <$9,999 $10,000- $30,000- $50,000- $75,000- >$100,000
$29,999 $49,999 $74,999 $99,999
Ethnic Group
African American 4.5% 7.5% 19.4% 20.9% 17.9% 29.9%
European American 3.7% 6.1% 14.6% 8.5% 13.4% 53.7%
Asian American 3.8% 10% 16.3% 18.8% 17.5% 33.8%
Latino American 1.4% 15.5% 23.9% 22.5% 11.3% 25.4%
O n
NO
70
= 11.64, p < .01], the frequency of lifetime racism events [F (3, 296) =
21.68, p < .01], the level of general college life stress [F (3, 296) = 5.07, p
< .01], perceived health/quality of life [F (3, 296) = 3.57, p < .05], somatic
distress [F (3, 296) = 4.07, p < .01], the endorsement of positive health values
[F (3, 296) = 4.26, p < .01], the frequency of participating in positive health
behaviors [F (3, 296) = 6.83, p < .01], the frequency of drinking until feeling
ill [marginal, F (3, 296) = 2.27, p < .1], and level of trait hostility endorsed
[F (3, 296) = 5 .35,p< .01]. Table 5 provides further details on the specific
ethnic group comparisons for these variables and also includes effect size
estimates (i.e. Cohen’s d) to clarify statistically significant group differences.
Overall, according to Cohen’s (1988) criteria, mean differences furnished effect
sizes ranging from medium to large.
The groups did not significantly differ on mean endorsements of family
member closeness [F (3, 296) = 1.11, p (NS)], friendship closeness [F (3, 296)
= 1.25, p (NS)], smoking frequency [F (3, 296) = 1.06, p (NS)] nor on degree
of anger rumination tendencies reported [F (3, 296) = 1.24, p (NS)]. Also
consistent with anticipated findings, univariate ANOVAs yielded significant
effects of ethnicity for all of the ambiguous interpersonal vignette variables (see
Table 5 for specific between group comparisons and effect size estimates).
Generally speaking, the African American participant group tended on average
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
71
to demonstrate the highest endorsements of perceived racism/racism
attributions, hostile attributions, anger and aggressive ideation in response to the
ten paper and pencil scenarios. Additionally as predicted, this same ethnic group
indexed the lowest mean frequencies for the willingness to consider situational
factors in the ambiguous scenarios (see Table 5).
Multiple Linear Regression Predictor and Moderator Effects3
Frequency o f Lifetime Racism Experiences, Hostility and Dimensions
of Quality o f Life. Four hierarchical linear regression models were used to test
the interaction of hostility and the frequency of lifetime racist events in
predicting perceived health/quality of life, somatic distress, family member
closeness and friendship closeness for each of the four ethnic groups. The same
model also incorporated ethnic identity in order to explore the possible three-
way interaction of this variable with hostility and lifetime racist events in
predicting the four dependent variables. The following variables were entered
first as covariates for each of the four regression models: mother’s level of
education, age, general college life stress and anger rumination.
3 Items from the DES-IV were originally intended to be included in these analyses as an
indicator of general negative affectivity. However, this variable was dropped from the
subsequent regression models for the following reasons: (1) the large number of covariates
already being included, (2) the high positive correlation of this index with trait hostility (r =
.73, p < .01) and with general college life stress (r = .63, p < .01), (3) and due to evidence
that the four ethnic groups did not significantly differ on this variable in univariate ANOVA
analyses [F (3, 296) = 2.12, p = . 1]. Trait hostility and general college life stress were
included in each of the following regression models to control for negative affectivity due to the
overlap in variance with this variable.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5: Ethnic Group Differences on Independent and Dependent Variables
Variable Mean SD F (total) p-value d (groups)
Racism-related Stress 16.25 <.001
A f Am (1) 48.6 19.1 1.25 (1 > 2 )
Eu Am (2) 29.7 11.0 .70 (3 > 2 )
As Am (3) 39.5 16.5 .50(1 > 3 )
Lat Am (4) 41.4 19.7 .76 (4 > 2 )
Lifetime Racist Events 21.7 <.001
A f Am 39.8 13.6 1.35 (1 > 2 )
Eu Am 25.6 7.33 .76 (3 > 2 )
As Am 31.7 8.81 .72(1 > 3 )
Lat Am 34.5 13.5 .85 (4 > 2 )
College Life Stress 5.07 < .01
A f Am 114.8 21.5 .60(1 > 2 )
Eu Am 102.4 20.9
As Am 108.9 21.3
Lat Am 113.3 22.5 .50 (4 > 2 )
Quality of Life 3.57 < .0 5
Af Am 21.8 3.48
Eu Am 22.4 3.61 .50 (2 > 3 )
As Am 20.6 3.93
Lat Am 21.5 3.09
Somatic Distress 4.07 < .01
Af Am 30.7 10.5
Eu Am 26.8 8.94
As Am 31.4 10.1 .50 (3 > 2 )
Lat Am 31.3 9.06 .50 (4 > 2 )
- j
to
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5 continued: Ethnic Group Differences on Independent and Dependent Variables
Variable Mean SD F (total) p-value d (group:
Family Member Closeness l.ii NS N/A
A f Am (1) 26.6 5.01
Eu Am (2) 25.3 5.38
As Am (3) 25.5 4.04
Lat Am (4) 25.3 4.52
Friendship Closeness 1.25 NS N/A
A f Am 27.4 4.09
Eu Am 26.8 3.94
As Am 26.4 3.89
Lat Am 26.1 4.15
Health Values 4.26 < .0 1
A f Am 89.8 13.1
Eu Am 94.0 13.7
As Am 87.5 15.0 .50 (2 > 3 )
Lat Am 87.0 13.5 .50 (2 > 4 )
Positive Health Behaviors 6.83 <.001
A f Am 39.9 9.06 .50 (2 > 1)
Eu Am 44.1 8.87
As Am 38.8 8.63 .60 (2 > 3 )
Lat Am 38.5 9.17 .60 (2 > 4 )
Drinking Until Feeling 1 1 1 2.27 < .1 0 N/A
A f Am 1.31 .656
Eu Am 1.55 .703
As Am 1.47 .726
Lat Am 1.60 .695
- j
u>
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5 continued: Ethnic Group Differences on Independent and Dependent Variables
Variable Mean SD F (total) p-value d (groups
Smoking 1.06 NS N/A
A f Am (1) 1.15 .584
Eu Am (2) 1.33 .930
As Am (3) 1.39 .974
Lat Am (4) 1.36 .903
Hostility 5.35 < .01
A f Am 18.7 6.08
Eu Am 16.3 6.00
As Am 19.8 6.00 .60 (3 > 2 )
Lat Am 19.1 5.88 .50 (4 > 2 )
Anger Rumination 1.24 NS N/A
A f Am 38.7 10.2
Eu Am 36.6 11.5
As Am 39.7 10.2
Lat Am 38.7 9.90
Perceived Racism 48.9 <.001
A f Am 38.7 12.5 2.1 (1 > 2 )
Eu Am 16.8 7.91 1.0 (2 < 3),
As Am 26.1 10.3 1.1 (1 > 3 )
Lat Am 27.5 13.1 .87(1 > 4 )
Situational Attributions 14.0 <.001
A f Am 43.6 8.25 1.1 (1 < 2)
Eu Am 52.2 6.92 .61 (3 < 2)
As Am 49.0 9.32 .62 (3 > 1 )
Lat Am 47.6 8.14 .50 (4 > 1 )
- 4
4^
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 5 continued: Ethnic Group Differences on Independent and Dependent Variables
Variable Mean SD F (total) p-value d (groups)
Hostile Attributions 22.8 <.001
A f Am (1) 34.7 11.9 1.4(1 > 2 ), .93 (1 > 3 ), .65 (1 > 4 )
Eu Am (2) 20.6 8.37
As Am (3) 24.7 9.30 .50 (3 > 2 )
Lat Am (4) 26.8 12.4 .60 (4 > 2 )
Anger 18.5 <.001
A f Am 41.2 12.4 1.2(1 > 2 ), .74(1 > 3 ), .64(1 > 4 )
Eu Am 28.8 8.55
As Am 33.0 9.78 .50 (3 > 2 )
Lat Am 34.0 10.1 .60 (4 > 2 )
Aggressive Ideation 5.80 < .01
A f Am 47.1 22.7 .64(1 > 2 )
Eu Am 34.8 14.7
As Am 39.6 16.9
Lat Am 38.2 18.3
-o
76
The primary predictors and interaction terms of interest were entered in
subsequent steps: hostility (HOS), ethnic identity (El), frequency of lifetime
racist events (LRE), LRE x HOS, LRE x El, HOS x El and LRE x HOS x EI.
The amount and significance of R2 A was observed at each step. Significant
interactions were graphically displayed and slopes were tested to assess if they
were significantly different from zero according to procedures described in
Aiken and West (1991).
Perceived Health and Quality of Life.4 The overall hostility model was
significant for the African American participant sample, R2= .57, F (11, 66) =
6.64, p < .01. Results revealed significant main effects of general college life
stress [R2A = .32, F (1, 63) = 31.4, p < .01, B = -.1] and hostility [R2A =
.16, F (1, 61) = 21.06, p < .01, B = -.22] only. It should be noted that
neither anger rumination nor the frequency of lifetime racist events was shown
4 To avoid redundancy in describing significant results in predicting each of the four main outcome
variables (i.e. perceived health/quality of life, somatic distress, family member closeness and friendship
closeness), the specifics regarding main effects for covariates including mother’s level of education, age,
college life stress and hostility will only be mentioned in the text detailing the first group of regression
models run that tested hostility as a moderator. In subsequent models testing the moderating effects of
anger rumination, positive health behaviors, drinking and smoking frequency, only significant R2 changes
for the aforementioned variables along with ethnic identity, racism-related experiences and for the
interactions among them will be described (at p < . O S level). In all of the analyses described in the
subsequent sections: X = the frequency o f lifetime racist events, racism-related stress OR the frequency of
perceived racism, Z = moderator (i.e. hostility, anger rumination, health values, frequency o f positive
health behaviors, frequency o f drinking until feeling ill, smoking frequency OR the frequency of situational
attributions made in the ambiguous vignettes) and W = ethnic identity.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
77
to be a significant predictor of this self-reported health-related variable for this
ethnic minority group while controlling for related negative affectivity variables.
Similarly, the overall predictive model explained a significant proportion
of the variance in perceived health/quality of life for the European American
participant group, R2 = .47, F (11, 81) = 5.76, p < .01]. As anticipated,
regression analyses yielded main effects of college life stress [R2 A = .28, F (1,
78) = 30.85, p < .01, B = -.04] and hostility [R2 A = .11, F (1, 76) = 14.74,
p < .01, B = -.23]. The interaction of the frequency of lifetime racist events
and ethnic identity was marginally significant [R2 A = .02, F (1, 72) = 3.14, p
< .1 ,5 = .11],
The overall model was significant for the Asian American sample in
predicting perceived health and quality of life, R2 = .38, F (11, 79) = 3.87, p
< .01. In addition to college life stress [R2 A = .21, F (1, 76) = 20.1, p <
.01, B = -.06] and hostility [R2A = .04, F (1, 75) = 4.42, p < .05, B = -.1]
specified as significant predictors, hostility also emerged as a significant
moderator in the relationship between the frequency of lifetime racist events and
level of perceived health/quality of life for this ethnic group [R2 A = .09, F (1,
71) = 10.7, p < .01, B = .03]. This effect was in partial support of original
predictions. A t-test of the simple slope confirmed a negative regression of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
78
perceived health and quality of life on the frequency of lifetime racist events at
low levels of hostility, t (-2.17, p < .05). The t-test of the simple slope at high
levels of hostility was marginally significant, t (1.98, p < .1). Thus as Figure 1
illustrates, for the Asian American sample, perceived health/quality of life
decreases as the frequency of lifetime racist events increases at low levels of
hostility. An opposite pattern emerged for high levels of hostility (see Figure 1).
Interestingly, the overall regression model was only shown to be a
marginally good fit for the Latino American sample in predicting perceived
health/quality of life, R2 = .24, F (11, 70) = 1.7, p < .1. The level of college
life stress was found to be the only significant predictor for this group, R2 A =
.1, F (1, 67) = 7.84, p < .01, B = -.04.
Somatic Distress. For the African American sample, the overall hostility
model represented a good fit in accounting for a significant proportion of the
variance in current somatic distress, R2 = .39, F (11, 66) = 3.2, p < .01. Age
was shown to negatively predict somatic distress for this ethnic group [R2 A =
.07, F (1, 64) = 5.11, p < .05, B = -.8]. In contrast, main effects of college
life stress [R2A = .14, F (1, 63) = 11.27, p < .01, B = .11], anger
rumination [R2A = .06, F (1, 61) = 5.18, p < .05, B = .22] and the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
79
Figure 1: Perceived quality of life/general health as a function of hostility and
the frequency of lifetime racist events (in standard deviations from the mean) for
the Asian American participant group.
• £ )
a
fi 25-
2
4)
I 20-
o
O'
•a
£
8 15-
tm
ft
-1 0 1
Frequency of Lifetime Racist Events
— Hi HOS
- - Lo HOS
frequency of lifetime racist events [R2 A = .08, F (1, 59) = 7.0, p = .01, B =
.23] were revealed as significant positive predictors of somatic complaints.
These findings were consistent with original predictions.
This group of predictors was also demonstrated to be a significant
predictive model for the European American participant sample, R2 = .55, F
(11, 81) = 7.7, p < .01. Main effects of college life stress [R2 A = .45, F (1,
78) = 65.7, p < .01, B = .21] and hostility [marginal, R2A = .02, F (1, 76)
= 2.93, p < .1, B = .04] were indicated as significant positive predictors of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
80
somatic distress from this model. A marginally significant interaction between
the frequency of lifetime racist events and hostility was also generated for this
ethnic group [R2 A = .03, F (1, 73) = 3.95, p < .1 ,5 = -.04],
Although the entire model explained a significant proportion of the
variance for the Asian American sample, R2 = .45, F (11, 79) = 5.12, p <
.01, this was primarily due to the contribution of general college life stress [R2
A = .40, F (1, 76) = 49.52, p < .01, 5 = .30]. A similar result emerged for
the Latino American sample in predicting somatic distress. The overall model
accounted for 29% of the variance F (11, 70) = 2.23, p < .05 and the sole
significant positive predictor was the level of college life stress reported [R2 A =
.21, F (1, 67) = 17.9, p < .01, 5 = .2].
Family Member Closeness. The overall model was not found to be a
good fit for the African American sample in predicting the level of family
member closeness, R2 = .18, F (11, 66) = 1.10, p (NS). The tendency to
ruminate on anger events was identified as the only significant negative
predictor of this interpersonal dependent variable, [marginal, R2 A = .02, F (1,
62) = 2.86, p < .1, 5 = -.12]. Interestingly, neither hostility nor the degree of
college life stress emerged as meaningful negative predictors of family member
closeness for this ethnic group. Further, in contrast to predictions, the frequency
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
81
of lifetime racist events was not shown to be a meaningful positive predictor for
this interpersonal quality of life dimension.
Similarly to the African American group, the variance in family member
closeness was not significantly accounted for by the overall hostility model for
the European American participant group, R2 = .16, F (11, 81) = 1.22, p
(NS). Further, the degree of college life stress was the sole, marginally
significant negative predictor of family member closeness [R2 A = .04, F (1, 78)
= 3.49, p < .1, B = -.007]. Here again it is of interest to note how neither
trait hostility nor anger rumination was shown to significantly negatively predict
this interpersonal variable.
In contrast to the other two groups, the overall model was found to be
marginally significant for the Asian American participants, R2 = .22, F (11, 79)
= 1.76, p < .1. Age [R2A = .06, F (1, 77) = 5.29, p < .05, B = -.71] and
college life stress [R2 A = .06, F (1, 76) = 5.24, p < .05, B = -.06] were the
only two main effects that emerged for this sample in negatively predicting the
degree of family member closeness. Finally, although the overall predictive
model was not shown to explain a significant proportion of the variance in
family member closeness for the Latino American participant group, R2 = . 19,
F (11, 70) = 1.3, p (NS), ethnic identity [R2A = .04, F (1, 64) = 2.95, p <
.1, B = .9] and the interaction of hostility and the frequency of lifetime racist
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
82
events [R2A = .04, F (1, 62) = 1.57, p < .1, B = -.1] were yielded as
marginally significant predictors. It is of interest to note how neither hostility
nor anger rumination were specified as unique negative predictors of family
closeness for the European American, Asian American and Latino American
participant samples.
Friendship Closeness. The overall predictive model was found to be
significant for the African American sample in explaining a significant
proportion of the variance in friendship closeness, R2 = .28, F (11, 66) = 2.0,
p < .05. Demographic variables including mother’s level of education [R2 A =
.06, F (1, 65) = 4.14, p < .05, B = .21] and age [marginal, R2A = .04, F (1,
64) = 2.90, p < .1, B = .5] were shown to be positive predictors of this
interpersonal variable for the African American participants. Interestingly and
not in the expected direction, the frequency of lifetime racist events was
demonstrated as a significant negative predictor of friendship closeness [R2A =
.07, F (1, 59) = 5.0, p < .05, B = -.1]. Additionally, hostility emerged as a
marginally significant moderator of the relationship between the frequency of
lifetime racist events and friendship closeness for this participant group [R2 A =
.04, F (1, 58) = 3.11, p < .1, B = -.01].
This entire group of predictors did not provide a good model fit in
predicting variation in friendship closeness among the European American
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
83
participants, R2 = .18, F (11, 81) = 1.4, p (NS). However as anticipated, a
main effect of trait hostility [R2 A = .05, F (1, 76) = 4.0, p < .05, 5 = -.13]
and anger rumination [marginal, R2 A = .04, F (1, 77) = 3.35, p < .1 ,5 = -
.72] emerged to negatively predict this interpersonal dimension of quality of
life. The overall model predicted 28% of the variance in friendship closeness for
the Asian American participant group, F (11, 79) = 2.35, p < .05. However,
in contrast to original hypotheses, only main effects for age [R2 A = .06, F (1,
77) = 5.28, p < .05, 5 = -.72] and college life stress [R2A = .14, F (1, 76)
= 13.18, p < .01, 5 = -.06] were specified as significant negative predictors.
Interestingly, age was demonstrated to be a positive predictor of friendship
closeness among the African American sample which is in contrast to the sign of
the beta coefficient specified for age for the Asian American sample. Finally,
the general model did not provide a good fit in predicting friendship closeness
among the Latino American participants R2 = .08, F (11, 70) = .46, p (NS).
No significant main effects or interaction effects emerged for this group based
on the hostility model.
Frequency o f Lifetime Racism Experiences, Anger Rumination and
Dimensions o f Quality o f Life. In order to assess the moderating effects of
anger rumination and ethnic identity in the relationship between the frequency of
lifetime racism events and four domains of quality of life, a series of four
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
84
hierarchical regression models were conducted separately for each of the ethnic
group participant samples. Mother’s level of education, age, general college life
stress and hostility were entered in the first steps of the model. Anger
rumination, ethnic identity, the frequency of lifetime racist events, and all
combinations of two- (i.e. LRE x AR, LRE x El and AR x El) and three-way
interaction terms (i.e. LRE x AR x El) were entered in subsequent steps of this
model.
Perceived Health and Quality of Life. The overall anger rumination
model accounted for 57 % of the variance in perceived health and quality of life
scores for the African American sample, F (11, 66) = 6.71, p < .01. Similar
to the hostility model described previously, main effects of general college life
stress (B = -.07) and hostility (B = -.28) emerged as significant negative
predictors for this participant group. This model also provided a good fit for the
European American sample’s scores on this variable, R2 = .46, F (11, 81) =
5.5, p < .01. Once again, only trait hostility (5 = -.23) and general college life
stress (B = -.03) were indicated as significant negative predictors from this
model for the level of perceived health and quality of life.
The anger rumination model was also shown to explain a significant
amount of the variance in perceived health/quality of life scores for the Asian
American participant sample, R2 = .35, F (11, 79) = 3.3, p < .01. In addition
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
85
to main effects of college life stress (B = -.05) and hostile beliefs (B = -.11),
anger rumination was identified as a meaningful moderator of the relationship
between the frequency of lifetime racist events and perceived health/quality of
life [R2 A = .04, F (1, 71) = 4.64, p < .05, B — .01]. Consistent with
expected findings, Figure 2 graphically illustrates how as the frequency of
lifetime racist events increases, there is a marginally significant increase in
perceived health/quality of life at high levels of anger rumination for the Asian
American group (t = 1.94, p < .1). The simple slope at low levels of anger
rumination was not shown to be different from zero [t = -1.3, p (NS)] although
at low frequency of lifetime racist events, perceived health/quality of life is
higher at low levels of anger rumination than at high levels of anger rumination
(see Figure 2).
The overall predictive model was only marginally significant for the
Latino American participant sample, R2 = .26, F (11, 70) = 1.85, p < .1.
Similarly to the aforementioned hostility model, the level of college life stress
(B = -.03) was the only significant predictor found for perceived health/quality
of life in this ethnic group.
Somatic Distress. Forty-one percent of the variance in level of somatic
distress was explained by the general anger rumination regression model for the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
86
African American participants, F (11, 66) = 3.5, p < .01. Identical to the
results yielded from the previously described hostility model and in accordance
Figure 2: Perceived quality of life/general health as a function of anger
rumination and the frequency of lifetime racist events (in standard deviations
from the mean) for the Asian American participant group.
1 25 -
13
u
4 )
a
4 >
§ 2 0 -
S
U
* S
a
£
m
15 -
0 1 1
Frequency of Lifetime Racist Events
ANGRUM
- - Lo
ANGRUM
with experimental expectations, the following main effects were identified as
significant positive predictors of somatic symptom distress for this ethnic group
sample: general college life stress (B = .13), anger rumination (B = .25) and
the frequency of lifetime racist events (B = .11). Age once again was specified
as a meaningful negative predictor of degree of somatic complaints (B = -.9). It
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
87
is interesting to note here how like in the prior model, trait hostility did not
account for a significant proportion of the variance in somatic distress for this
group.
The overall model was also shown to be a good fit in predicting 52% of
the variance in somatic distress scores among the European American
participants, F (11, 81) = 6.85, p < .01. Here again consistent with the prior
moderator model tested and in contrast to the African American sample, trait
hostility (B = .22) remained a marginally-significant predictor of this self-
reported health-related variable within the Caucasian sample. A robust main
effect of college life stress (B = .23) was also sustained in this predictive
model. In a similar pattern of results generated by the hostility regression model
for this variable, although the overall model was significant for the Asian
American group, R2 = .45, F (11, 79) = 5.07, p < .01, it was primarily
accounted for by a main effect of college life stress (B = .31). Building upon
the findings indicated in the hostility moderator model discussed previously, the
interaction of ethnic identity and anger rumination predicted a marginally
significant amount of the variance in somatic distress scores for the Latino
American participants [R2 A = .04, F (1, 60) = 3.73, p < .1, B = .17]. The
overall model was a good fit for this group, R2 = .30, F (11, 70) = 2.28, p <
.05, and college life stress (B = .20) remained the only significant main effect
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
88
in accounting for a significant proportion of the variance in the Latino American
participants’ somatic symptom distress scores.
Family Member Closeness. The full anger rumination regression model
provided only a marginally significant fit for the African American participants’
ratings on family member closeness, R2 = .28, F (11, 66) = 1.96, p < .1.
Consistent with the hostility model and original hypotheses, anger rumination (B
-.13) emerged as the sole marginally significant main effect for this model in
negatively predicting this interpersonal quality of life dimension. Also somewhat
in support of anticipated results, anger rumination significantly moderated the
relationship between the frequency of lifetime racist events and family member
closeness for this ethnic group [R2 A = .07, F (1, 58) = 4.72, p < .05, B =
.01]. T-tests of the simple slopes against zero verified a marginally significant
negative regression of perceived family member closeness on the frequency of
lifetime racist events at low levels of anger rumination coping (t = -1.84, p
< .1: see Figure 3). As displayed in Figure 3, as the frequency of lifetime racist
experiences increases, the level of family member closeness decreases at low
levels of anger rumination. However, it is also important to highlight that at a
lower frequency of racism experiences, family member closeness is significantly
higher for those who hypothetically exhibit low levels of anger rumination
coping (see Figure 3).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
89
Outside of the context of racism experiences, the general interaction of
anger rumination and ethnic identity also emerged as a meaningful predictor of
the variance in family member closeness for this participant group [R2 A = .09,
F (1, 56) = 6.9, p < .05, B = -.12]. As depicted in Figure 4, perceived family
member closeness significantly decreases with increasing levels of anger
rumination coping at high levels of ethnic identity only (t = -3.1, p < .01) for
Figure 3: Family member closeness as a function of anger rumination and the
frequency of lifetime racist events (in standard deviations from the mean) for the
African American participant group.
35
Vi
? 25 -
4)
1
I 20-
Hi
ANGRUM
- - Lo
ANGRUM
I 15 -
10
-1 0 1
Frequency of Lifetime Racist Events
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
90
this African American sample. There was no significant regression of anger
rumination on family member closeness at low levels of ethnic identity [t = -
.685, p (NS): see Figure 4].
This overall anger rumination model also explained a marginally
significant proportion of the variance in perceived family member closeness
scores for the European American participants, R2 = .22, F (11, 81) = 1.82, p
< .1. In addition to a main effect of college life stress (marginal, B = .01), the
Figure 4: Family member closeness as a function of ethnic identity and anger
rumination (in standard deviations from the mean) for the African American
participant group.
35
I 15-
- ■ Lo ETHID
Hi ETHID
10
-1 0 1
Anger Rumination
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
91
three-way interaction among the frequency of lifetime racist events, anger
rumination and ethnic identity was specified as a significant predictor of family
member closeness in this ethnic group [R2 A = .1, F (1, 70) = 9.42, p < .01,
B = -.03]. Figure 5 graphically displays this moderating effect.
T-tests of the simple slopes against zero confirmed both the marginally
significant negative regression of perceived family member closeness on the
frequency of lifetime racist events at low anger rumination and low ethnic
identity (t = -1.97, p < .1) and a marginally significant negative regression
between the same variables at the combination of high anger rumination and
high ethnic identity (t = -1.95, p < .1). Although at higher frequency of racism
experiences, higher levels of self-reported family member closeness were
indicated at both high anger rumination-low ethnic identity and low anger
rumination-high ethnic identity than for the other two hypothetical groups,
neither of these simple slopes was found to be significantly different from zero
(see Figure 5).
The full anger rumination predictive model was not a significant fit for
the Asian American participant’s family member closeness scores, R2 = .21, F
(11, 79) = 1.63, p (NS). Similarly to the hostility model, only the covariates of
age (B = -.75) and college life stress (B = -.05) emerged as significant negative
predictors of this interpersonal variable. This model also only explained 21% of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
92
Figure 5: Family member closeness as a function of ethnic identity, anger
rumination and the frequency of lifetime racist events (in standard deviations
from the mean) for the European American participant group.
« 35-
4 >
s 30 -
v
U
1 2 5 -
S _
20 -
15 -
-1 0 1
Frequency of Lifetime Racist Events
Hi AR-Hi ETHID
- - Hi AR-Lo ETHID
Lo AR-Hi ETHID
Lo AR-Lo ETHID
the variance in perceived family member closeness among the Latino American
participants, F (11, 70) = 1.39, p (NS). Despite this a main effect of ethnic
identity (marginal, B = .85) and the interaction of anger rumination and ethnic
identity [R2 A = .05, F (1, 60) = 4.17, p < .05, B = -.1] were specified as
significant predictors from this model for this participant group. Figure 6 shows
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
93
how family member closeness is higher at high versus low levels of ethnic
identity at lower levels of anger rumination. Neither simple slope was
demonstrated to be significantly different from zero.
Figure 6: Family member closeness as a function of ethnic identity and anger
rumination (in standard deviations from the mean) for the Latino American
participant sample.
35
8 30
d
4>
©
u
p fi
a
S
25
20
S 15
10
-1 0 1
Anger Rumination
— Hi ETHID
- - -L o ETHID
Friendship Closeness. Unlike for the family member closeness quality
of life dimension, variation in perceived friendship closeness was significantly
accounted for by the overall anger rumination model for the African American
participant sample, R2 = .38, F (11, 66) = 3.09, p < .01. Similar main effects
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
94
emerged from this regression analysis as were obtained from the previous
hostility predictive model: mother’s level of education (B = .16), age
(marginal, B = .47) and the frequency of lifetime racist events (B = -.17).
Ethnic identity was also identified as a significant moderator in the relationship
between anger rumination and perceived friendship closeness for this ethnic
group [R2 A = .14, F (1, 56) = 11.48, p < .01]. Although neither simple slope
was significantly different from zero, this unanticipated interaction revealed that
friendship closeness is higher at high levels of anger rumination for low levels
of ethnic identity versus high levels of ethnic identity (see Figure 7).
Additionally, the three-way interaction of the frequency of lifetime racist
events x anger rumination x ethnic identity was yielded as a significant predictor
of friendship closeness for the African American participants [R2 A = .05, F (1,
55) = 4.44, p < .05, B = -.006]. In support of a priori expectations, perceived
friendship closeness sharply decreases with increasing frequency of lifetime
racist experiences at the combination of low anger rumination and low ethnic
identity among this group (t = -4.1, p < .01: see Figure 8). A similar,
marginally significant negative regression between levels of reported friendship
closeness and the frequency of lifetime racist experiences also emerged for the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
95
Figure 7: Friendship closeness as a function of ethnic identity and anger
rumination (in standard deviations from the mean) for the African American
participant sample.
30 -
4 >
a
I 25 -
u
a
| 2 0 -
0 )
£
15 -
1 1
Anger Rumination
— Hi ETHID
- - Lo ETHID
line reflecting the hypothetical high anger rumination-low ethnic identity group
(t = -1.74, p < .1) although the rate of decrease is not as high as for the
former group (see Figure 8).
In contrast to the results obtained from the African American sample, the
anger rumination predictive model did not account for a significant proportion
of the variance in friendship closeness among the European American
participants, R2 = .18, F (11, 81) = 1.41, p (NS). This regression model
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
96
Figure 8: Friendship closeness as a function of anger rumination, ethnic identity
and the frequency of lifetime racist events (in standard deviations from the
mean) for the African American participant group.
35 -
15 -
1 1
Frequency of Lifetime Racist Events
Hi AR-Hi ETHID
- - -H i AR-Lo ETHID
Lo AR-Hi ETHID
Lo AR-Lo ETHID
generated the same main effects of the previous predictor model which included
specifying trait hostility (B = -.14) and anger rumination (marginal, B = -.1) as
negative predictors of this interpersonal variable for the Caucasian American
group. The overall model was a good fit for the Asian American sample by
accounting for 26% of the variance in friendship closeness, F (11, 79) = 2.15,
p < .05. However, none of the variables of interest for this moderator model
(e.g. anger rumination, ethnicity-related indicators) provided incremental
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
97
predictive power beyond the main effects of college life stress (B = -.05) and
age (B = -.65). Finally, this model like the hostility model was did not provide
a good fit for explaining the variance in friendship closeness among the Latino
American participant group, R2 = .08, F (11, 70) = .44, p (NS). No significant
predictors or interactions were generated from this model.
Frequency o f Lifetime Racism Experiences, Health Values and Positive
Health Behaviors. A hierarchical linear regression model was performed to
assess health values and ethnic identity as moderators in the relationship
between the frequency of lifetime racist events and the frequency of
participating in positive health behaviors. The following covariates were entered
in the first steps of the predictive model: mother’s level of education, age,
general college life stress and trait hostility. Subsequently, health values (HV),
ethnic identity, the frequency of lifetime racist events and all two-way (i.e. LRE
x HV, LRE x El and HV x El) and three-way interaction (i.e. LRE x HV x El)
terms were entered in later steps of the model.
The full model was shown to be a good fit in explaining a significant
proportion of the variance in the frequency of participation in positive health
behaviors for the African American sample, R2 = .57, F (11, 66) = 6.61, p <
.01. Health values emerged as a meaningful positive predictor [R2 A = .35, F
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
98
(1, 60) = 36.29, p < .01, B = .5] for this participant group. Additionally, in
support of a priori hypotheses, ethnic identity was found to be a significant
moderator of the relationship between the frequency of lifetime racist events and
the frequency of positive health behaviors [R2 A = .06, F (1, 57) = 6.95, p =
.01, B = -.12]. T-tests conducted to ascertain whether or not the simple slopes
were significantly different from zero indicated a positive regression of the
frequency of positive health behavior participation on the frequency of lifetime
racist events at low levels of ethnic identity (t = 2.67, p = .01) only. In other
words, the frequency of participating in positive health behaviors tends to
increase with increasing frequency of lifetime racist events at low levels of
ethnic identity for the African American group (see Figure 9).
A significant three-way interaction was also generated from this model in
predicting the frequency of participating in positive health behaviors [R2 A =
.08, F (1, 55) = 9.75, p < .01, B = -.01], As Figure 10 illustrates, the
frequency of lifetime racist events positively regresses on the frequency of
positive health behaviors at high health value and low ethnic identity (t = 3.4, p
< .01) and negatively regresses at high health value and high ethnic identity (t
= -2.06, p < .05) based on the African American sample’s data. In further
clarification of this unanticipated finding it would appear that the combination of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
99
low ethnic identity and high health values enhances positive health habit
participation as the frequency of lifetime racist events increases. In contrast,
Figure 9: The frequency of positive health behaviors as a function of ethnic
identity and the frequency of lifetime racist events (in standard deviations from
the mean) for the African American participant sample.
-1 0 1
Frequency of Lifetime Racist Events
Hi ETHID
- - Lo ETHID
although the combination of high ethnic identity and high health value is
associated with high levels of positive health behaviors at low frequency of
lifetime racist events, participation for this hypothetical sample decreases with
increasing frequency of lifetime racist events (see Figure 10).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
100
Figure 10: The frequency of positive health behaviors as a function of health
values, ethnic identity and the frequency of lifetime racist events (in standard
deviations from the mean) for the African American participant group.
* 50 -
1 45 -
« 35 -
C m
i 30 -
0 >
I 25-
0 1 1
Hi HVAL-Hi
ETHID
- - Hi HVAL-Lo
ETHID
Lo HVAL-Hi
ETHID
Lo HVAL-Lo
________ ETHID
Frequency of Lifetime Racist
Events
The overall health value model was demonstrated to be a good fit for the
European American sample as well, R2 = .54, F (11, 81) = 7.48, p < .01. In
accordance with expectations, main effects of college stress [R2 A = .05, F (1,
78) = 4.40, p < .05, B = .-.04] and hostility [R2 A = .06, F (1, 77) = 5.8, p
< .05, B = -.2] were specified as significant negative predictors of the
frequency of participation in positive health behaviors. Health values did emerge
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
101
as anticipated as a significant positive predictor of positive health habit
participation for the European American participants [R2 A = .36, F (1, 76) =
.53.7, p < .01, 5 = .2].
The overall health value model explained 43 % of the variance in the
frequency of taking part in positive health behaviors for the Asian American
sample, F (11, 79) = 4.7, p < .01. Analyses yielded main effects for level of
general college life stress [marginal, R2 A = .04, F (1, 76) = 3.65, p < .1, 5
= -.11] and health values [R2 A = .25, F (1, 74) = 28.7, p < .01, B = .27] in
the expected directions. Interestingly, the frequency of lifetime racist events [R2
A = .03, F (1, 72) = 4.0, p < .05, 5 = .2] emerged as a significant positive
predictor of positive health habits for this ethnic group sample.
Fifty-three percent of the variance was accounted for by the overall
health value model in predicting the frequency of participating in positive health
habits for the Latino American group, F (11, 69) = 6.0, p < .01. As
anticipated, health value was demonstrated to be a significant positive predictor
[R2A = .35, F (1, 64) = 44.95, p < .01, B = .44]. Main effects of mother’s
level of education [marginal, R2 A = .04, F (1, 70) = 3.06, p < .1 ,5 = -.1],
age [R2A = .06, F (1, 68) = 4.83, p < .05, 5 = .41] and ethnic identity
[marginal, R2A = .04, F (1, 65) = 2.85, p < .1, 5 = -2.5] were also
indicated as meaningful predictors of positive health behavior participation for
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
102
this participant group. It is of interest to highlight the notable differences in the
patterns of status variables, ethnicity-related factors, stress variables and
hostility that were identified among the four ethnic groups in predicting the
frequency of taking part in positive health behaviors.
Racism-related Stress, Health Behaviors and Dimensions o f Quality o f
Life. A series of twelve hierarchical linear regression models were performed
to assess the possible moderating effects of positive health behaviors (4),
drinking until feeling ill frequency (4) and smoking frequency (4) in the
relationship between racism-related stress appraisal and selected dimensions of
quality of life. Separate regression models were conducted for each of the three
health behavior variables and for each of the four ethnic groups. The covariates
were entered in the first steps of the model (i.e. mother’s education level, age,
college life stress, hostility). Then the specific health behavior variable (e.g.
positive health behavior participation, PHB) was entered followed by ethnic
identity, racism-related stress (RS) and all of the two- (e.g. RS x PHB, RS x El
and PHB x El) and three-way (e.g. RS x PHB x El) interactions of ethnic
identity and racism-related stress with each of the three health behavior
variables, each in a separate model. In other words, the potential predictor and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
103
interaction effects of the three health behavior variables were not tested in the
same model. Rather, three separate models were performed to investigate the
unique effects of each.
Frequency of Positive Health Behaviors.
Perceived Health and Quality of Life. Seventy-one percent of the
variance in general perceived health/quality of life scores among the African
American participants was accounted for by the full positive health behavior
regression model, F (11, 66) = 12, p < .01. In conjunction with the degree of
college life stress (B = -.07) and a hostile cognitive style (B = -.2), the
frequency of positive health behaviors [R2A = .15, F (1, 60) = 28, p < .01, B
= . 16] also emerged as a significant main effect in accordance with prior
experimental predictors. Similarly, fifty-five percent of the variance in this
overall self-reported health-related variable was explained by the total positive
health behavior model for the European American group, F (11, 81) = 7.74, p
< .01. The frequency of participating in positive health behaviors was also
identified as a meaningful predictor for this ethnic group’s general perceived
health/quality of life scores [R2A = .12, F (1, 75) = 19.02, p < .01, B = .1].
College life stress (B = -.02) and trait hostility (B = -.22) remained robust
main effects in this model.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
104
The overall positive health behavior model was also a significant
predictor of general health/quality of life assessments for the Asian American
sample, R2 = .48, F (11, 79) = 5.77, p < .01. Interestingly, for this group of
participants, racism-related stress [R2A = .03, F (1, 72) = 4.16, p < .05, B =
-.05] was specified as a significant main effect in addition to college life stress
(B = -.03), hostile beliefs (B — -.08) and the frequency of participating in
positive health habits [R2A = .15, F (1, 73) = 19.52, p < .01, B = .24).
Further, once again unlike the other three participant groups, trait hostility did
not meaningfully contribute to the variance in perceived health/quality of life
scores among the Latino American participants.
The overall model was indicated as a significant predictor for this ethnic
group, R2 = .41, F (11, 70) = 3.73, p < .01 which yielded main effects of
college life stress (B = -.05) and the frequency of positive health behaviors [R2
A = .19, F (1, 64) = 19.03, p < .01, B = .2], A significant, unanticipated
interaction of ethnic identity and the frequency of positive health behaviors also
emerged from analyses in this sample [R2 A = .04, F (1, 60) = 4.61, p < .05,
B = .06]. As illustrated in Figure 11 and confirmed by t-tests of the simple
slopes against zero, a positive regression of general perceived health/quality of
life on the frequency of positive health behaviors exists at both high and low
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
105
levels of ethnic identity for this group. However, the slope for the hypothetical
high line is steeper (see Figure 11).
Somatic Distress. The overall frequency of positive health habits model
was shown to be a good fit for the variation in somatic symptom distress among
the African American participants, R2 = .45, F (11, 66) = 4.15, p < .01. Age
Figure 11: Perceived quality of life/general health as a function of ethnic
identity and the frequency of positive health behaviors (in standard deviations
from the mean) for the Latino American participant sample.
t£
a
3
o
" O
* 3
C J
35
15
10
-1 0
1 ■ Hi ETHID
- - Lo ETHID
Frequency of Positive Health Behaviors
( .B = -.34), college life stress (B = .14) and racism-related stress (B = .11)
were significant main effects generated by this regression analysis. Additionally,
the frequency of positive health behaviors interacted with racism-related stress
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
106
to predict levels of somatic distress [R2 A = .09, F (1, 58) = 8.9, p < .01, B
= .02]. However, in contrast to original predictions, the frequency of positive
health behaviors did not buffer against the effects of racism-related stress on
somatic complaints within this ethnic group. On the contrary, as Figure 12
displays, somatic distress positively regresses on racism-related stress at high
versus low levels of positive health behavior participation (t = 4.41, p < .01).
Figure 12: Somatic distress as a function of the frequency of positive health
behaviors and racism-related stress (in standard deviations from the mean) for
the African American participant sample.
40
cc 2 0 -
" " Lo PHBEH
Hi PHBEH
15 -
10
-1 0 1
Racism-related Stress
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
107
Although 50% of the variance in somatic distress scores was explained
by the full model in the European American group [F (11, 80) = 6.30, p <
.01), this was largely due to the contributions of trait hostility (B = .27) and
college life stress (B = .22). Fifty-one percent of the variance in somatic
symptom distress was accounted for by the full positive health behavior model
in the Asian American group, F (11, 79) = 6.36, p < .01. In addition to the
main effect of college life stress (5 = .31), the frequency of positive health
behaviors was identified as a meaningful moderator in the relationship between
racism-related stress and somatic complaints in this sample [R2 A = .04, F (1,
71) = 5.16, p < .05, B = -.02]. In contrast to the direction of the same
interaction found within the African American group, in this case and in
accordance with predictions, as the degree of racism-related stress increased,
somatic distress also increased at low participation in positive health habits (t =
2.63, p < .05: see Figure 13). Finally, the overall model was only marginally
significant in explaining the variance in somatic symptoms in the Latino
American sample, R2 = .26, F (11, 70) = 1.86, p < .1). Only college life
stress (B = .16) was specified as a significant predictor from this model within
the Latino American participant group. Interestingly, the frequency of positive
health behavior participation failed to emerge as a significant negative predictor
of this health-related variable in any of the participant samples.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
108
Family Member Closeness. The full positive health habit participation
model did not provide a good fit to predicting the variance in perceived family
member closeness among the African American participants, [R2 = .16, F (11,
66) = 1.0, p (NS)]. However, consistent with expectations, this positive coping
variable was indicated as a significant positive predictor of family member
closeness for this ethnic group [R2 A = .07, F (1, 60) = 4.68, p < .05, B =
.22]. In contrast, 25% of the variance in this interpersonal quality of life
dimension was accounted for by the overall regression model in the European
Figure 13: Somatic distress as a function of the frequency of positive health
behaviors and racism-related stress (in standard deviations from the mean) for
the Asian American participant sample.
45
40
35
30
25
20
15
10
-1 0 1
Racism-related Stress
Hi PHBEH
- - Lo PHBEH
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
109
American group, F (11, 81) = 2.11, p < .05. As in previous analyses
investigating the variance explained in perceived family member closeness
within this ethnic group, college life stress (B = -.04) remained a marginally
significant negative predictor. Analyses also yielded the frequency of positive
health habits as an additional main effect for this group [marginal, R2 A = .04,
F (1, 75) = 3.55, p < .1 , B = .30],
Also, somewhat as anticipated, a significant interaction between racism-
related stress and the frequency of participating in positive health behaviors
emerged [R2A = .06, F (1, 73) = 5.65, p < .05, B = .01]. Figure 14
illustrates the negative regression of family member closeness on racism-related
stress at low levels of positive health behaviors (t = -2.07, p < .05). The slope
of the hypothetical line at high levels of positive health behaviors was not
significantly different from zero [t = 1.47, p (NS): see Figure 14]. Further, the
three-way interaction among racism-related stress, the frequency of positive
health behaviors and ethnic identity was specified as a marginally significant
predictor of the variance in perceived family member closeness among the
Caucasian participants [R2A = .03, F (1, 70) = 3.08, p < .1 ,5 = -.02].
Twenty-six percent of the variance in perceived family member closeness
was accounted for by the overall regression model in the Asian American group,
F (11, 79) = 2.22, p < .05. Age (B = -.74) and college life stress (B = -.04)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
110
remained robust negative predictors of this interpersonal variable within the
Asian American sample. Additionally, in accordance with predictions, ethnic
identity emerged as a meaningful moderator of the relationship between racism-
related stress and perceived family member closeness [R2 A = .07, F (1, 70) =
6.03, p < .05, B = .07]. As Figure 15 illustrates and as confirmed by t-tests of
the simple slope against zero, family member closeness positively regresses on
Figure 14: Family member closeness as a function of the frequency of positive
health behaviors and racism-related stress (in standard deviations from the
mean) for the European American participant sample.
35 -
1 0 1
Racism-related Stress
Hi PHBEH
- - Lo PHBEH
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
I ll
racism-related stress at high levels of ethnic identity (t = 2.04, p < .05). The
hypothetical slope of the low ethnic identity line was not found to be
significantly different from zero [t = -1.03, p (NS)].
The full positive health habit regression model was shown to explain a
significant amount of the variance in family member closeness among the Latino
American participants, [R2 = .3, F (11, 70) = 2.27, p < .05]. The following
main effects were generated from this analysis: ethnic identity (marginal, B =
Figure 15: Family member closeness as a function of ethnic identity and racism-
related stress (in standard deviations from the mean) in predicting family
member closeness for the Asian American participant sample.
30 -
S
? 2 5 '
I
I 20-
-1 0 1
Racism-related Stress
Hi ETHID
- - -LoETHID
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
112
1.15), racism-related stress (marginal, B = .01), the frequency of positive
health behaviors [R2A = .07, F (1, 64) = 5.34, p < .05, B = .22], The
frequency of participating in positive health habits was also identified as a
marginally significant moderator of the relationship between racism-related
stress and this interpersonal quality of life component [R2 A = .05, F (1, 62) =
3.94, p < .1 ,5 = -.01].
Friendship Closeness. Thirty-one percent of the variance in this quality
of life dimension was accounted for by the full positive health behavior model in
the African American participant sample, F (11, 66) = 2.22, p < .05. This
analysis yielded main effects of mother’s education level (B = .3), age (B =
.19) and the frequency of taking part in positive health habits [R2A = .12, F (1,
60) = 8.93, p < .01, B = .23] as robust positive predictors of this
interpersonal variable. The overall model was also shown to be a good fit in
predicting variance in perceived friendship closeness among the Caucasian
participant group, R2 = .32, F (11, 81) = 3.0, p < .01. A hostile cognitive
style (B = -.1) and the frequency of positive health behaviors [R2A = .17, F (1,
75) = 17.47, p < .01, B = .19] emerged as the sole unique predictors from
this model.
Thirty percent of the variance in perceived friendship closeness ratings
with the Asian American participant sample was explained by the full regression
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
113
model, F (11, 79) = 2.66, p < .01. Consistent with previous models, age (B =
-.61) and college life stress (B = -. 06) remained meaningful negative predictors
of this interpersonal dimension of quality of life. Additionally and unexpectedly,
ethnic identity was identified as a significant moderator of the relationship
between the frequency of positive health habit participation and friendship
closeness in this ethnic group [R2 A = .06, F (1, 69) = 6.24, p < .05, B = .1].
This analysis indicated a positive regression of friendship closeness on the
frequency of positive health behaviors at high levels of ethnic identity (t = 2.66,
p = .01) only (see Figure 16). Interestingly, the overall regression model failed
to predict a significant amount of the variance in friendship closeness for the
Latino American group, R2 = .1, F (11, 70) = .61, p (NS). Neither significant
main effects nor interaction effects were specified.
Frequency of Drinking Until Feeling 1 1 1 .
Perceived Health and Quality of Life. Approximately 65% of the
variance in perceived health and quality of life was explained by the drinking
frequency model for the African American participant sample, F (11, 66) =
9.1, p < .01. As anticipated, drinking until feeling ill emerged as a significant
main effect in negatively predicting this health-related variable [R2 A = .05, F
(1,61) = 7.4, p < .01, 5 = -2.5] in conjunction with the covariates of college
life stress (B = -.06) and trait hostility (B = -.23). The interaction of racism-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
114
Figure 16: Friendship closeness as a function of ethnic identity and the
frequency of positive health behaviors (in standard deviations from the mean)
for the Asian American participant sample.
1
15 -
0 1 1
— Hi ETHID
- Lo ETHID
Frequency of Positive Health Behaviors
related stress and ethnic identity was also identified as a marginally significant
predictor of perceived health and quality of life for the African American group
[R2A = .02, F ( l, 56) = 3.86, p < .1, B = -.02],
Similarly, although this overall model was shown to be a good fit for the
European American sample, R2 = .45, F (11, 81) = 5.17, p < .01, it was
primarily due to the variance accounted for by the main effects of college life
stress (B = -.02) and levels of hostile beliefs (B = -.28). Neither anger
rumination nor any interactions with racism-related stress or ethnic identity were
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
115
specified as meaningful predictors for this model. Once again, despite
explaining a significant proportion of the variance in perceived health/quality of
life, R2 = .32, F (11, 79) = 2.96, p < .01, college life stress (fl = -.06) and
hostility (B — -.12) were the only robust predictors generated for the Asian
American participant group.
The general drinking frequency model was not a good overall fit for the
Latino American group’s scores on perceived health/quality of life, R2 = .22, F
(11, 70), p (NS). However, both a significant main effect of college life stress
(B = -.05) in conjunction with the interaction of racism-related stress with the
frequency of drinking until feeling ill [R2 A = .05, F (1, 62) = 4.13, p < .05,
B = .05] were specified as significant predictors of perceived quality of life for
this ethnic group sample. In the unanticipated direction, a t-test of the simple
slopes indicated a marginally significant negative regression of perceived
health/quality of life on levels of racism-related stress at low levels of drinking
until feeling ill (t = -1.77, p < .1; see Figure 17). The simple slope based on
the hypothetical data at high levels of drinking until feeling ill was not found to
be different from zero [t = .80, p (NS)]. Curiously, although perceived health
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
116
and quality of life is higher at low levels of racism-related stress for low versus
high levels of drinking until feeling ill, it tends to decrease with increasing
racism-related stress appraisals within this ethnic minority group (see Figure
17).
Figure 17: Perceived quality of life/general health as a function of the frequency
of drinking until feeling ill and racism-related stress (in standard deviations from
the mean) for the Latino American participant sample.
24 -
1
1 22 '
1 2 1 -
§ 20 -
5 19 -
i i s -
a >
6
P n
16 H
1 0 1
Racism-related Stress
Hi DRINK
- - Lo DRINK
Somatic Distress. The overall drinking frequency model significantly
predicted somatic distress for the African American sample, R2 = .35, F (11,
66) = 2.67, p < .01. However, in contrast to experimental hypotheses, neither
the frequency of drinking until feeling ill significantly predicted nor interacted
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
117
with racism-related stress and/or ethnic identity to account for the variance in
somatic symptom distress for this ethnic group. Rather, consistent with the
findings yielded from the previously described moderator models for this group
in predicting somatic complaints, age (B = -.82), college life stress (B = .12)
and racism-related stress (B = .22) remained significant main effects for this
model.
The drinking frequency model was also found to be a good fit for the
European American sample R2 = .5, F (11, 81) = 6.32, p < .01. However,
this was largely due to the contribution of the main effects of covariates
including college life stress (B = .23) and trait hostility (B = .24).
Interestingly, none of the variables of interest (i.e. drinking until feeling ill,
ethnic identity, racism-related stress) provided incremental predictive power for
explaining variation in somatic distress for this participant group.
Similarly, the overall model was found to be significant for the Asian
American sample R2 = .46, F (11, 79) = 5.36, p < .01. College life stress was
the only significant positive predictor of the level of somatic symptom distress
(B = .3) in this ethnic group. Twenty-nine percent of the variance was
explained for by this regression model for the Latino American participants, F
(11, 70) = 2.22, p < .05. A main effect of college life stress was specified as a
robust positive predictor of somatic distress (B = .17). A marginally significant
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
118
three-way interaction of drinking until feeling ill x racism-related-stress x ethnic
identity emerged from the analyses [R2 A = .05, F (1, 59) = 3.97, p < .1, B =
-.17].
Family Member Closeness. The overall drinking frequency model was
not a significant predictor of family member closeness for the African American
participants, R2 = .15, F (11, 66) = .85, p (NS). However, a marginally
significant two-way interaction was yielded from this model. More specifically,
the frequency of drinking until feeling ill moderated the relationship between
racism-related stress and family member closeness [R2 A = .05, F (1, 58) =
3.37, p < .1, B = -.04]. Similarly, this overall regression model did not
provide a good fit for predicting scores of family member closeness for the
European American sample, R2 = .13, F (11, 81) = 1.0, p (NS). Only a main
effect of college life stress (B = -.01) was indicated as a significant predictor of
family member closeness for this group.
Twenty-eight percent of the variance in family member closeness was
predicted by the full drinking frequency model for the Asian American group, F
(11, 79) = 2.37, p < .05. Regression analyses yielded main effects for age (B
= -.9), college life stress (5 = -.05) and the frequency of drinking until feeling
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
119
ill [marginal, R2 A = .04, F (1, 72) = 3.9, p < .1 ,5 = 1.4]. Interestingly, the
beta coefficient for this last predictor was opposite of the anticipated direction.
Moreover, drinking until feeling ill emerged as a significant moderator
of the relationship between racism-related stress and family member closeness
for this ethnic group [R2 A = .05, F (1, 71) = 4.39, p < .05, B = -.06].
However, t-tests of the simple slopes against zero failed to reach significance for
either high or low levels of drinking frequency. Nonetheless, at low levels of
racism-related stress greater family member closeness is indicated at high levels
of drinking than at low levels of drinking among the Asian American
participants (see Figure 18).
Finally, family member closeness was not strongly predicted by the
overall drinking frequency model among the Latino American participants, R2
= .2, F (11, 70) = 1.36, p (NS). Despite this and similar to the results
generated by previous moderator models for this group in predicting family
member closeness, main effects of ethnic identity (marginal, B = .63) and
racism-related stress (B = .06) were indicated. Again, it is important to
underscore how racism-related stress positively predicts family member
closeness within this ethnic group which is once again in accordance with a
priori expectations.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
120
Figure 18: Family member closeness as a function of the frequency of drinking
until feeling ill and racism-related stress (in standard deviations from the mean)
for the Asian American participant sample.
35 -
1 30 -
G
I 2 5 '
S
s i? 20 -
I*
15 -
-1 0 1
Racism-related Stress
1 Hi DRINK
- - Lo DRINK
Friendship Closeness. As with the previous dependent variable, the
overall drinking frequency model was not a good fit for the African American
sample in accounting for the variance in friendship closeness, R2 = .16, F (11,
66) = .94, p (NS). Robust positive predictors of this interpersonal variable
continued to be mother’s level of education (B = .3) and age (B = .48) for this
ethnic group. Only 18% of the variance in friendship closeness was accounted
for by the overall model in the European American group, F (11, 81) = 1.38, p
(NS). Trait hostility (B = -.21) remained a significant negative predictor of this
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
121
variable for the Caucasian participants. Additionally for this sample, a
marginally significant three-way interaction among the frequency of drinking
until feeling ill, racism-related stress and ethnic identity resulted [R2 A = .04, F
(1,70) = 3.33, p < .1 ,B = .14].
Although the full model was shown to be a marginally significant fit for
the Asian American participants in accounting for variation in friendship
closeness, [R2 = .23, F (11, 79) = 1.86, p < .1], this finding was mostly
explained by main effects of covariates for age (B = -.7) and college life stress
(,B = -.07). Once again it is interesting to note that the beta coefficients for the
African American group for age is in the opposite direction of the beta that
emerged for age for the Asian American group in predicting friendship
closeness across various regression models. Curiously, it would seem that for
the former ethnic group, this interpersonal dimension of quality of life increases
with increasing age but for the latter group, perceived friendship closeness
decreases with increasing age. Lastly, the amount of variance accounted for by
this drinking frequency model did not reach statistical significance for the Latino
American group, R2 = .13, F (11, 70) = .82, p (NS). As indicated in previous
regression models, no predictors or interaction effects significantly contributed
to explaining the variation in friendship closeness for this group of participants.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
122
Smoking Frequency.
Perceived Health and Quality of Life. The regression model testing the
moderation of smoking frequency was a significant fit in predicting the African
American participants’ variance in perceived health/quality of life scores, R2 =
.66, F (11, 66) = 9.71, p < .01. However, in contrast to expectations, the
main variable of interest (i.e. frequency of smoking behavior) was not a
significant predictor of perceived health/quality of life for this ethnic group.
Main effects of college life stress (B = -.05) and hostility (B = -.25) remained
robust negative predictors of this self-reported health-related variable for this
model.
Additionally, ethnic identity was identified as a meaningful moderator in
the relationship between racism-related stress levels and ratings of perceived
health/quality of life [R2 A = .04, F (1, 57) = 5.6, p < .05, B = -.03]. Figure
19 illustrates this interaction effect. As shown, at low levels of racism-related
stress, low ethnic identity is indexed by lower perceived health/quality of life
than at high ethnic identity. In contrast to experimental hypotheses, a t-test of
the simple slope against zero confirmed a marginally significant positive
regression of perceived health/quality of life on the degree of reported racism-
related stress at low ethnic identity (t = 1.67, p = .1: see Figure 19).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
123
Figure 19: Perceived quality of life/general health as a function of ethnic
identity and racism-related stress (in standard deviations from the mean) for the
African American participant sample.
m
1 0 1
Hi ETHID
- - Lo ETHID
Racism-related Stress
No relationship was found between the predictor and dependent variable for
high ethnic identity [t = -.911, p (NS)].
The smoking frequency regression model predicted a significant
proportion of the variance in overall perceived health/quality of life for the
European American sample, R2 = .46, F (11, 81) = 5.49, p < .01. However
this effect was largely accounted for by the contribution of the main effects of
college life stress (B = -.02) and trait hostility (B = -.3). A similar pattern of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
124
results emerged for the Asian American sample. The overall model explained
30% of the variance in perceived health/quality of life for this ethnic group, F
(11, 79) = 2.67, p < .01. College life stress (B = -.06) and hostile beliefs (B
= -.14) remained significant negative predictors in this model.
A marginally significant proportion of the variance in perceived
health/quality of life scores among the Latino American participants was
explained by the total smoking frequency model, R2 = .25, F (11, 70) = 1.79,
p < .1. As in other regression models used to predict this dependent variable
within this ethnic minority sample, college life stress (B = -.05) continued to
emerge as the sole negative predictor of this self-reported health-related
variable. Additionally, smoking frequency significantly interacted with the
degree of racism-related stress appraisals to predict perceived health/quality of
life for this group [R2 A = .05, F (1, 62) = 4.0, p = .05, B = .04]. Figure 20
illustrates the marginally significant negative regression of perceived
health/quality of life on racism-related stress at low levels of smoking frequency
(t = -1.87, p < .1). Similar to the unanticipated interaction obtained from the
drinking until feeling ill regression model for this sample and dependent
variable, as racism-related stress increases, perceived health/quality of life tends
to decrease at lower smoking frequency (see Figure 20).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
125
Figure 20: Perceived quality of life/general health as a function of smoking
frequency and racism-related stress (in standard deviations from the mean) for
the Latino American participant sample.
35
30
u 25
I
1
2 20
15
10 1 1 -------
-1 0 1
Racism-related Stress
— Hi SMOKE
- - Lo SMOKE
Somatic Distress. Thirty-six percent of the variance in the degree of
somatic complaints was predicted by the full smoking frequency model for the
African American group, F (11, 66) = 2.85, p < .01. Like previous regression
models for this dependent variable among the African American participants,
main effects of age (B = -.8), college life stress (B = .11) and racism-related
stress (B = -.8) were the only significant predictors specified from this analysis.
The overall regression model also explained a significant proportion of the
variance in somatic symptom distress among the Caucasian participants, R2 =
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
126
.53, F (11, 81) = 7.12, p < .01. However, neither smoking frequency nor any
of the ethnicity-related variables of interest was identified as a meaningful
predictor of this self-reported health-related factor. College stress and trait
hostility (B = .2) remained robust positive predictors of somatic complaints
within this group (B = .37). Analyses further generated a marginally significant
three-way interaction among racism-related stress, smoking frequency and
ethnic identity [R2 A = .02, F (1, 70) = 2.99, p < .1, B = .23] for the
European American group.
Among the Asian American participants, the variation in somatic
symptom distress was accounted for by the overall smoking frequency
regression model, R2 = .5, F (11, 79) = 6.43, p < .01. In addition to a
significant main effect of college life stress (B = .32), smoking frequency did
emerge as a meaningful positive predictor of current somatic complaints within
this ethnic group [R2 A = .03, F (1, 73) = 4.19, p < .05, B = 1.07]. Finally,
college life stress (B = .18) was the only significant effect specified from this
model in predicting somatic distress for the Latino American group which
largely accounted for the significance of the overall model within this participant
sample, R2 = .28, F (11, 70) = 2.13, p < .05.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
127
Family Member Closeness. The total smoking frequency regression
model accounted for a non-significant, 12% of the variance in perceived family
member closeness scores within the African American sample, F (11, 66) =
.67, p (NS). No significant predictor or moderator effects were yielded from
this analysis. For the European American participant group, the full model was
also not a good fit for explaining variation in family member closeness, R2 =
.13, F (11, 81) = .93, p (NS). College life stress (B = -.03) was identified
once more as the single marginally significant predictor for this model within
the Caucasian sample.
In contrast, the entire smoking frequency model did emerge to account
for a significant amount of the variance in perceived family member closeness
for the Asian American participants, R2 = .23, F (11, 79) = 2.27, p < .05.
Main effects for the covariates of age (B = -.8) and college life stress (B = -
.05) remained significant negative predictors of this interpersonal quality of life
component within the Asian American participant group. Regression analyses
also yielded a marginally significant interaction of racism-related stress and
ethnic identity in the prediction of family member closeness [R2 A = .04, F (1,
70) = 3.86, p < .1, B = .04],
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
128
In addition, the frequency of smoking behavior was specified as a
meaningful moderator in the relationship between racism-related stress and
family member closeness for this ethnic group [R2A = .0 8 ,F (l,7 1 ) = 7.15,p
< .01, B = 1], As displayed in Figure 21 and in support of anticipated
results, the level of perceived family member closeness increases with
increasing levels of racism-related stress at low frequency of smoking behavior
only (t = 2.27, p < .05). The amount of variance accounted for in family
member closeness by the full smoking frequency model did not reach statistical
significance for the Latino American participants, R2 = .18, F (11, 70) = 1.19,
p (NS). A main effect of racism-related stress (B = .05) was the only
marginally significant predictor specified in the model. It is important to
reiterate that in accordance with predictions, racism-related stress has emerged
as a positive predictor of this interpersonal variable among the Latino American
participant group.
Friendship Closeness. Only 17% of the variance in perceived friendship
closeness was accounted for by the full smoking frequency model, F (11, 66) =
1.0, p (NS) for the African American participants. Main effects of mother’s
level of education (B = .25) and age (marginal, B = .40) were the only
significant predictors of this variable for this ethnic minority group. The full
smoking frequency regression model was not a significant predictor of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
129
Figure 21: Family member closeness as a function of smoking frequency and
racism-related stress (in standard deviations from the mean) for the Asian
American participant sample.
a 30 -
5
0
3 25 -
b
a
a
1 2 0 -
*
S 15 -
0 1 1
— Hi SMOKE
- - Lo SMOKE
Racism-related Stress
friendship closeness among the European American participants, R2 = .15, F
(11, 81) = 1.1, p (NS). Trait hostility (B =-.26) continued to be the lone
significant effect for this dependent variable within the Caucasian American
sample.
Main effects of age (5 = -.7) and college life stress (B = -.07) were the
primary factors contributing to the significant predictive fit of the overall model
for the Asian American ethnic group, R2 = .27, F (11, 79) = 2.33, p < .05.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
130
Finally, the full smoking frequency model only predicted 10% of the variance in
perceived friendship closeness in the Latino American sample, F (11, 70) =
.57, p (NS). As in other regression models tested for this interpersonal quality
of life factor within this ethnic group, no significant main or interaction effects
were identified.
Perceived Racism, Hostility, Anger and Aggressive Ideation in
Everyday, Ambiguous Interpersonal Situations. One of the main objectives of
this study was to investigate the relationship between perceived racism and
hostile attributions within the context of ambiguous, everyday microstressor
events for each of the four ethnic groups included in this investigation. Towards
that end, the following covariates were entered in the first steps of the
regression model: mother’s level of education, age, college life stress, trait
hostility, ethnic identity and the frequency of lifetime racist events. Perceived
racism or the frequency of racism attributions were entered in the final step in
predicting the frequency of hostile attributions.
For the final set of analyses, the frequency of situational attributions was
evaluated as a moderator in the relationship between the frequency of perceived
racism/racism attributions and both level of anger and aggressive ideation within
the context of daily ambiguous interpersonal situations. Each of the two
regression models controlled for the following variables (which were entered
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
131
first): mother’s level of education, age, trait hostility, anger rumination coping,
ethnic identity and the frequency of lifetime racist events. The frequency of
racism attributions (PR), situational attribution frequency (SA) and all two- and
three-way interactions among them and ethnic identity were entered in later
steps in the model (i.e. PR x SA, PR x El, SA x El, PR x SA x El).
Perceived Racism as a Predictor of the Variance in Hostile Attributions
in Ambiguous Situations. In accordance with predictions, the frequency of
racism attributions made within the ambiguous interpersonal situations explained
a significant proportion of the variance in hostile attribution scores across all
ethnic groups. This effect was robust even when controlling for SES variables,
general college life stress, trait hostility, ethnic identity and prior exposure to
racism. For the African American participant sample, perceived racism (B =
.67, p < .01) predicted 35% of the variance in hostile attributions. Within the
European American group, the frequency of racism attributions (B = .8, p <
.01) accounted for 31% of the variance in the willingness to make hostile
attributions within the context of ambiguous, everyday microstressor events.
Similarly, 31% of hostile attribution scores were predicted by the frequency of
perceived racism (B = .65, p < .01) endorsed in the ambiguous vignettes for
the Asian American sample. Finally, the tendency to make racism attributions
(B = .81, p < .01) in the series of 10 ambiguous scenarios accounted for 40%
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
132
of the variance in hostile attribution endorsements among the Latino American
participants. Related to this, pearson’s bivariate correlation coefficients between
perceived racism and hostile attributions were specified as .81, .63, .70 and .84
for the African American, European American, Asian American and Latino
American participant groups respectively.
Situational Attributions as a Moderator between Perceived Racism and
Anger in Ambiguous Situations. The full model explained 55% of the variance
in anger ratings among the African American participants, F (13, 66) = 5.0, p
< .01. As predicted, the covariates of college life stress (B = .08), trait
hostility (B = .01) and anger rumination coping (B = .51) emerged as
significant positive predictors of this affective variable. Additionally as
expected, perceived racism [R2 A = .16, F (1, 58) = 19.35, p < .01, B = .52]
was identified as a significant main effect from this model.
The overall regression model predicted 40% of the variance in anger
endorsements within the European American group, F (13, 81) = 3.56, p <
.01. The degree of college life stress (B = .07), trait hostility (B = .22) and
perceived racism [R2 A = .09, F (1, 73) = 10.92, p < .01, B = .5] were the
only significant effects generated from this analysis. In contrast, in addition to
the full model being indicated as a meaningful predictor of the variance in anger
for the Asian American group, R2 = .5, F (13, 79) = 5.1, p < .01, several
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
133
main effects and interaction effects were yielded. College life stress (B = .03),
trait hostility (B = -.13) and anger rumination coping (B = .33) were identified
as significant predictors for this ethnic group’s variation in anger response.
Interestingly the beta coefficient for hostility was not found to be in the expected
direction.
Somewhat in accordance with expectations, the frequency of situational
attributions was shown to significantly interact with the frequency of perceived
racism in predicting anger levels for the Asian American sample [R2 A = .04, F
(1, 69) = 4.6, p < .05, B = .02]. As Figure 22 displays, at low frequency of
perceived racism, anger ratings are lower at high levels of situational
attributions (t = 2.68, p < .01). However, as the frequency of racism
attributions increases the level of anger also increases at high frequency of
situational attributions (see Figure 22). The three-way interaction among
perceived racism, situational attributions and ethnic identity was found to be
marginally significant in explaining anger response level in this group [R2 A =
.03, F (1, 66) = 3.67, p < .1, B = -.01], It is of interest to note that while
controlling for the specified covariates, perceived racism did not emerge as a
meaningful predictor of anger for the Asian American participants.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
134
Figure 22: Anger as a function of the frequency of situational attributions and
the frequency of perceived racism (in standard deviations from the mean) for the
Asian American participant sample.
Frequency of Perceived Racism
Hi SITATT
- - Lo SITATT
Sixty-eight percent of the variation in anger ratings was accounted for by
the full model, F (13, 70) = 9.13, p < .01 in the Latino American sample.
Similar to the findings obtained with the other ethnic groups, college life stress
(B = .02), trait hostility (B = .35) and the frequency of racism attributions [R2
A = .22, F (1, 62) = 35.26, p < .01, B = .48] were identified as significant
positive predictors of this post-event affective variable. However, unlike the
other participant groups, both the frequency of lifetime racist events (B = .04)
and ethnic identity (B = -.21) also emerged as unique main effects for
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
135
predicting the variance in the Latino American participants’ anger responses.
Moreover, the frequency of situational attributions was specified as a marginally
significant moderator in the relationship between perceived racism and anger [R2
A = .02, F (1, 60) = 3.28, p < .1, B = -.01]. Ethnic identity also marginally
interacted with the frequency of situational attributions to predict anger ratings
in this ethnic group [R2 A = .02, F (1, 59) = 3.82, p < .1 ,5 = -.12].
Interestingly, in contrast to anticipated findings, the frequency of situational
attributions failed to be identified as a unique negative predictor of anger for any
of the four participant groups.
Situational Attributions as a Moderator between Perceived Racism and
Aggressive Ideation in Ambiguous Situations. The full situational attribution
regression model was found to be a good fit in explaining 39% of the variation
in aggressive ideation endorsement among the African American participants, F
(13, 66) = 2.59, p < .01. College life stress (5 = .31), hostility (marginal, B
= .36) and perceived racism [R2 A = .1, F (1, 58) = 8.36, p < .01, B = .61]
were indicated as positive predictors of aggressive thought frequency in the
ambiguous situational vignettes. Ethnic identity marginally interacted with the
frequency of racism attributions to predict aggressive ideation for this ethnic
group [R2 A = .04, F (1, 54) = 3.47, p < .1, B = .38],
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
136
Fifty-two percent of the variation in the willingness to endorse
aggressive cognitions within the context of everyday, ambiguous events was
explained by the overall model in the European American participant group, F
(13, 81) = 5.77, p < .01. The following covariates were identified as
predictors from this model: college life stress (B = .08), trait hostility
(marginal, B = -.13), the tendency to ruminate on anger experiences (B = .35)
and the frequency of lifetime racist events (marginal, B = -5.48). The
frequency of perceived racism also emerged as a meaningful main effect in
predicting aggressive ideation within the Caucasian American group [R2 A =
.09, F (1, 73) = 11.16, p < .01, B = .6],
Additionally, consistent with predictions, this contextual attributional
style significantly moderated the relationship between the frequency of racism
attributions and aggressive thinking for this ethnic group: [R2 A = .07, F (1, 71)
= 9.62, p < .01, B = -.1]. Figure 23 illustrates the positive regression of
aggressive ideation on the frequency of perceived racism at low levels of
situational attributions (t = 3.92, p < .01). Analyses also yielded a significant
unanticipated interaction between ethnic identity and the frequency of perceived
racism in predicting aggressive ideation [R2 A = .04, F (1, 70) = 5.23, p <
.05, B = -.41]. As shown in Figure 24 and confirmed by a t-test of the simple
slope against zero, the frequency of aggressive ideation positively regresses on
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
137
Figure 23: Aggressive ideation as a function of the frequency of situational
attributions and the frequency of perceived racism (in standard deviations from
the mean) for the European American participant sample.
90
80
70
60
50
40
30
20
10
0
Frequency of Perceived Racism
- Hi SITATT
- - Lo SITATT
the frequency of perceived racism at high ethnic identity only (t = 3.44, p <
.01). Here the data suggest that low ethnic identity among the European
American participants may be a risk for significant aggressive thinking with the
increasing frequency of racism perceptions in ambiguous, everyday life
experiences.
The total predictive model was found to explain 29% of the variation in
aggressive thought frequency among the Asian American participants, F (13,
79) = 2.07, p < .05. College life stress (B - .02), hostility (B = .09) and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
138
Figure 24: Aggressive ideation as a function of ethnic identity and the frequency
of perceived racism (in standard deviations from the mean) for the European
American participant sample.
60 -
a 5 5 -
©
50 - 2
N 4
g 45 -
I 4 0 -
W >
« 3 5 -
30 -
0 1 1
Frequency of Perceived Racism
Hi ETHID
- - Lo ETHID
anger rumination (B = .62) were all specified as meaningful positive predictors
of this variable. Similarly to the results obtained for predicting anger, the
frequency of situational attributions significantly interacted with the frequency
of racism attributions to predict aggressive ideation [R2 A = .05, F (1, 69) =
5.0, p < .05, B = .04]. This result once again is partially consistent with
predictions for this ethnic group. Figure 25 graphically displays the marginally
significant positive regression of aggressive ideation on the frequency of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
139
Figure 25: Aggressive ideation as a function of the frequency of situational
attributions and the frequency of perceived racism (in standard deviations from
the mean) for the Asian American participant sample.
1 ------------------- r
-1 0 1
Frequency of Perceived Racism
Hi SITATT
- - Lo SITATT
perceived racism at high levels of situational attributions (t = 1.76, p < .1). It
is important to note that in accordance with hypotheses, high levels of situational
attributions index lower aggressive ideation than low situational attributions at
low frequency of perceived racism (see Figure 25).
Finally, although the overall regression model predicted a significant
amount of the variation in aggressive thinking among the Latino American
participant sample (R2 = .45, F (13, 70) = 3.67, p < .01), this was principally
the result of main effects in college life stress (B = .16), trait hostility (B =
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
140
.51) and perceived racism [R2 A = .18, F (1, 62) = 19.41, p < .01, B = .8].
Once again, the frequency of situational attributions failed to provide
incremental predictive power in explaining variation in aggressive ideation for
any of the participant samples.
DISCUSSION
There is a burgeoning corpus of research that has established significant
relationships between racism and components of psychological and physical
wellness in ethnic minority groups (e.g. Contrada et al., 2002; Klonoff et al.,
1999; Landrine et al., 1996). The current study’s correlational and regression
analyses provided further support to existing research in this area. As
anticipated, the frequency of lifetime racist events and the level of racism-
related stress reported were positively related to general college life stress,
somatic distress, trait hostility and an anger rumination coping style. While
racism experiences were negatively associated with overall quality of life
perceptions within this ethnically-diverse college age sample.
Additionally, previous research has indexed positive associations
between racism and negative health behaviors such as tobacco usage and alcohol
consumption (e.g. Landrine et al., 1996). However, this study was unique in its
aim to explore the links among an individual’s cumulative racist victimization
experiences, the negative appraisal of those life events, health values and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
141
wellness-promoting health behaviors. As predicted, the frequency of lifetime
racist events and racism-related stress were shown to negatively covary with
both the importance of taking part in health-enhancing behaviors (e.g. exercise,
regular doctor visits, making healthy nutritional choices) and the actual
frequency of participating in positive health habits. These correlational findings
tentatively suggest that aversive interpersonal experiences involving racism
could negatively impact the value placed upon and the level of involvement in
behaviors that improve health and well-being.
Interestingly and also in accordance with expectations, the stress
attributed to racism was identified as a positive predictor of family member
closeness within the Latino American participants in this investigation. This
finding suggests particular environmental conditions in which the content of
stress may be an important factor involved in the process of promoting a greater
sense of family cohesiveness within certain ethnic minority communities.
However, this intriguing finding will need to be replicated in future longitudinal
research. It will also be important to gain a greater understanding of ethnic
group variation in the relationship found between racism-related stress and this
interpersonal quality of life dimension. For example, the stress due to racism
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
142
failed to emerge as a meaningful predictor of family or friendship closeness in
the other participant groups when regression models were adjusted for SES,
college life stress and trait hostility.
Beyond achieving the basic objectives described thus far, the primary
purpose of this study was to provide a rigorous test of the potential roles of
health- and ethnicity-related individual difference factors in modulating the
relationship between racism and dimensions of quality of life while adjusting for
important psychosocial covariates. This analysis took an integrative approach in
emphasizing the influence of ethnicity and culture on cognition, affect regulation
and behavior from functional-developmental, social identity (i.e. rejection-
identification) and acculturative stress perspectives. More specifically, this
project sought to begin to clarify the complex relationships among racism
exposure and stress appraisal, hostility, anger rumination, ethnic identity, health
values, health behaviors and attributional style in explaining variation in self-
reported health and wellness variables within a multi-ethnic participant sample.
The remainder of this discussion will briefly address these issues in more detail.
Racism, Hostility and Quality of Life
Hostility is one of the most widely-studied personality dimensions that
has been reliably implicated in enhanced risk for the development of serious
chronic illness (e.g. hypertension, diabetes; Miller et al., 1996; Surwit et al.,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
143
2002) and acute forms of cardiovascular disease (e.g. myocardial infarction;
Miller et al., 1996). It is defined as a set of interpersonal beliefs or a cognitive
relational schema reflecting the themes of suspiciousness, negative expectations
about the motives of others and that others are likely responsible for harms
experienced by the individual who inflexibly holds these views (Smith, 1992).
Recent empirical work involved in exploring the long-term predictive
power of hostility on health status has alerted researchers to the importance of
studying this construct in younger, healthy cohorts (e.g. Siegler et al., 2003). In
so doing, clinical scientists within the field of health psychology have the
opportunity to target early prevention or intervention efforts among vulnerable
groups in order to help reduce risk for later disease onset. Thus, young adult
participants from college or university communities are no longer merely
“samples of convenience” in this era of clinical health research. Instead they are
considered a vital population to study how health-related personality variables
for example develop, are maintained over time and influence affective
experience and modifiable behavioral coping responses.
As described earlier, several models have been promoted to explain the
role of hostility in the process of compromising physical health (see Miller et
al., 1996 for a discussion) and that less is known about hostility’s involvement
in psychological and interpersonal functioning. As a result of this gap in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
144
literature, the psychosocial vulnerability model of hostility perhaps has received
the least empirical attention. According to this perspective, high hostility is
associated with enhanced stress exposure, low social support and has its
antecedents rooted in cultural and environmental causes (e.g. parental
socialization practices and behavioral reinforcement tendencies: Keltikangas-
Jarvinen et al., 2003; Sebej, 1989).
In light of this conceptualization and in conjunction with the rejection-
identification model of ethnic identity development (Branscombe et al., 1999), it
was of interest to the current study to investigate hostility within the context of
life events involving racism. In this case, the present project suggested that
“rejection-protection” or “rejection-rejection” may be more appropriate ways of
qualifying this proposed hostility-related phenomenon. Results consistent with
this hypothesis could then serve to provide both a theoretical rationale and
further empirical support to the existing body of research that has robustly
indexed significant ethnic group differences in this health-related personality
variable.
It was hypothesized that a hostile cognitive style in part evolves as a
consequence of the stress associated with encountering frequent negative
interracial experiences. This particular type of life event is more unique to the
experience of racial and ethnic minorities (although it was also important to
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
145
begin examining these issues in European American samples living in ethnically
rich areas such as Los Angeles) and thus increases the stress exposure potential
for people of color. In contrast to the general psychosocial vulnerability model
of hostility with regard to social support, it was predicted that within the specific
context of racism, hostility may be associated instead with greater interpersonal
benefits.
In this way, hostility is reffamed as an adaptive, culturally-normative,
cognitive coping strategy employed in the service of buffering the ill effects of
racism on multiple aspects of well-being (e.g. overall quality of life and the
sense of perceived closeness with family and with friends presumed to share a
common ethnic heritage). However, it was also anticipated that hostility would
also remain associated with greater vulnerability for experiencing anger and
aggressive ideation when situations are perceived to be racially-motivated. The
frequent experience of anger in response to racist maltreatment in turn may pose
a particular threat to the physical well-being for ethnic minority individuals.
Thus, hostility would likely not act as a protective factor in attenuating the link
between one’s lifetime history of racism and the level of somatic distress
endorsed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
146
Indeed, several results emerged in support of both the psychosocial
vulnerability model and the contemporary rejection-protection hypothesis of
hostility. Firstly, in accordance with previous research, significant ethnic group
differences in hostility were evidenced in this study. However, this is one of the
first investigations to examine this health-related personality variable in a wider
range of ethnic minority groups. In the current analysis, the Asian American
and Latino American participants on average tended to endorse higher levels of
hostile beliefs than the European American participant group. A marginally
significant trend in the same direction was indicated for the African American
versus the Caucasian sample. Secondly, as mentioned earlier, continuous
variables assessing lifetime history of racist victimization and racism-related
stress were shown to positively covary with trait hostility for the full sample
although mean endorsements on these ethnicity-related variables significantly
distinguished the ethnic minority groups from the European American sample.
These combined results suggest that members of certain ethnic minority
groups may be particularly susceptible to advocating beliefs reflecting a hostile
cognitive style in part stemming from greater exposure to specific negative
racism-related life events. Alternatively, ethnic group differences on mean
endorsements of a hostile relational schema could be confounded by inflated
reports of general college life stress among these groups. For example, the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
147
African American and the Latino American participants on average
acknowledged experiencing higher levels of stress associated with adjusting to
the college environment than the other two participant groups. However,
general college stress and the racism-related stress and lifetime history of racism
variables were all highly correlated in the entire participant sample.
Given this, it is likely that both the general measures of college life stress
and trait hostility may be contaminated by the experience of racism-related
stress and the impact of frequent racist events for ethnic minority individuals.
These are areas that will need to be further addressed in future research that
examines the effects of stress and hostility on various psychological and physical
health outcomes in ethnically diverse samples. It is important to note that
adjusting in some way for an individual’s history and/or appraisal of previous
racism exposure is rarely performed within these larger bodies of research.
Hostility was also shown to be a significant negative predictor of overall
perceived health/quality of life, somatic distress, family member closeness and
sense of friendship closeness. However, the existence of these relationships
varied by ethnic group membership. For instance, as endorsements of a hostile
interpersonal schema increased, general life satisfaction tended to decrease for
the African American, European American and Asian American participant
samples. This finding is consistent with the notion that perceived quality of life
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
148
is an indicator of positive affectivity and thus it would be expected that this
personality dimension which reflects vulnerability to experiencing negative
affect would index such a relationship.
Additionally, hostility interacted with the frequency of lifetime racist
events to predict reports of general life satisfaction among the Asian American
participant sample. This finding is what would have been predicted by the
rejection-protection racism buffering hypothesis (i.e. an extension of the
rejection-identification model of ethnic group affiliation: Branscombe et al.,
1999). It would appear that low levels of hostility may be a risk factor for
diminished perceptions of quality of life as the frequency of racist victimization
experiences accumulates for members of this ethnic group. A trend emerged
indicating an inverse relationship for high levels of hostility. Thus this
interaction effect suggests that a hostile belief system which involves an external
attribution style and negative interpersonal expectancies is an important variable
involved in mitigating the negative impact of racism on life satisfaction.
It is particularly interesting to note that this result was obtained from the
Asian American participants. Virtually no other research has examined the
relationship of hostility to aspects of well-being for members of this ethnic
group. In fact, this finding may be especially surprising given the cultural
stereotypes surrounding Asian individuals characterized as the submissive and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
149
amiable “model minority” (Mok, 1998). Clearly, this is a fascinating area to
further explore in future research especially due to the significant within group
heterogeneity extant among communities of Asian descent. It is possible that
more fine grain analyses may yield additional specificity for this hostility
phenomenon depending on other factors such as country of origin (e.g. Japan
versus China) and level of acculturation (e.g. first or second generation, etc.)
which likely impact the historical context of racial discrimination experienced in
the past (e.g. Japanese American internment during World War II) and how
current racism exposure is encountered and interpreted.
When adjusted for significant covariates, hostility was not identified as a
highly meaningful predictor of somatic symptom distress across all ethnic
groups although it was specified as a marginally significant main effect for the
Caucasian sample. Further, it did not significantly predict the sense of family
member closeness for any of the ethnic groups included in this study. Finally,
individual differences in a hostile cognitive style were only indicated as a
meaningful negative predictor of friendship closeness for the European
American participants.
In general, the intensity of recent college life stress was the most robust
main effect yielded from this first set of regression analyses for all ethnic
groups. It is possible that hostility evidenced an unremarkable effect on
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
150
components of interpersonal relationship satisfaction and other components of
well-being due to the shared variance already accounted for by current college
stress or in some cases anger rumination (which will be addressed further in the
next section of the discussion).
Alternatively, consistent with a dynamic constructivist perspective (e.g.
Hemmings, 2002; Miller, 2002) this could represent ethnic differences in the
function and construal of hostility which then could variably impact dimensions
of quality of life. This is suggested by the fact that out of the four ethnic groups,
hostility was specified as a significant effect most consistently for the European
American participants. Perhaps this is the case resulting from how this construct
is appraised and interpreted among “ethnic majorities”. In other words, hostility
as measured in various ethnic groups may not necessarily hold the same
meaning for each group. Researchers may be assessing ethnically different
constructs but utilizing the same term.
This phenomenon was most strikingly apparent in that a similar pattern
of relationships between hostility and elements of quality of life failed to emerge
for the Latino American participant group. Thus, although this ethnic group on
average documented higher elevations on this health-related construct in
comparison to the Caucasian sample, this failed to translate into explaining
significant incremental variation in their reports of general life satisfaction,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
151
somatic distress or interpersonal quality of life dimensions when current college
life stress, SES and demographic factors were controlled for. This intriguing,
ethnicity-specific result encourages subsequent analyses to further explore why
this is the case and what implications it has for understanding the role of
hostility in the psychosocial health of members of the Latino community. This
finding tentatively suggests once again that hostile beliefs within this ethnic
sample may not function in a similar fashion in comparison to other ethnic
groups in terms of predicting aspects of well-being. This hypothesis is further
supported by recent work that revealed greater warm engagement and less
hostility observed in the interactions among Latino versus European American
couples during a conflict resolution task (Fisher, Gudmundsdottir, Gilliss, Skaff,
Mullan et al., 2000).
Alternatively, the notable ethnic group distinctions in the links exhibited
between trait hostility and aspects of quality of life could stem from variation in
the evolution of hostility over time used as a coping mechanism when
encountering racist maltreatment. It is possible that as a whole, the families of
the Latino American participants may have had a shorter length of time living in
the United States than the African American and Asian American participants.
Thus, these latter two groups theoretically may have had a longer acculturation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
152
epoch during which time hostile beliefs have developed into a protective strategy
to mitigate the adverse effects of racism on psychological and relational
wellness.
Some evidence to support this has indicated significant generational
differences in both mean reports of hostility and in the strength of relationships
found between this variable and somatic symptom endorsement among a large
sample of Mexican Americans (Lee & Markides, 1989). More specifically, the
younger (and presumably most acculturated) cohort indexed the highest levels of
hostility and the strongest association between hostile irritability and physical
symptom reporting (Lee et al., 1989). These generational effects could reflect
the notion that with greater acculturation, greater hostility could manifest itself
due to psychological shifts towards a more individualistic versus collectivistic
way of thinking and interacting. It may be argued that hostility is in direct
conflict with interpersonal harmony which is a hallmark value evident among
more collectivistic cultural groups. It will be left up to future qualitative
ethnographic and quantitative work to continue to expand our knowledge of the
impact of culture in scaffolding the relationships between hostility and well
being.
Finally, in partial support of hypotheses, the level of one’s sense of
ethnic group affiliation was demonstrated as a marginally significant positive
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
153
predictor of family member closeness among the Latino American participants
in this analysis. Although indicated as a trend, this finding is in accordance with
emphasizing how the strength of ethnic group membership is involved in an
individual’s feelings of closeness with their family with whom they share a
common cultural heritage. However, in contrast to the original rejection-
identification model, the degree of ethnic identity endorsed did not meaningfully
interact with the cumulative frequency of one’s lifetime history of racism to
predict various aspects of life satisfaction in this overall study sample.
One exception that merits emphasizing was the significant moderation of
the link between racism-related stress and perceived family member closeness
by ethnic identity among the Asian American participants. A high level of ethnic
identity promoted greater relationship closeness on this interpersonal quality of
life dimension with increasing levels of stress stemming from racial
discrimination. Another related finding that provides limited support to the
rejection-identification model was that high ethnic identity indexed greater life
satisfaction ratings at low levels of racism-related stress than did low ethnic
identity for the African American participant group.
The reasons for this general dearth of findings with regard to an
individual’s psychological sense of belonging to one’s ethnic minority group
likely are a result of both the method of measurement used (i.e. a one-item
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
154
Likert scale) and due to the overlap in variance already accounted for by other
covariates in the regression models. In other words, there was little unique
variance remaining that this variable could predict given when it was entered in
the analyses. Once again, it was important for this investigation to provide a
stricter test of the role of ethnicity-related factors in well-being after controlling
for SES, demographic, general stress and personality variables which other
research in this area has often failed to do.
What this lack of findings also suggests is that ethnic identity for some
racial minorities (e.g. African American and Asian American) may be more
tightly coupled with other psychological (e.g. hostility), experiential (e.g.
racism-related stress) and status variables (e.g. mother’s level of education) than
for other groups (e.g. Latino American). Thus, future research would benefit
from including a more specific measure of ethnic identity (e.g. Phinney et
al.’s...Multi-group Ethnic Identity Measure or Contrada et al.’s 2001, Group
Membership Questionnaire) that assesses various psychological, affective,
interpersonal and behavioral components of this global construct. In so doing,
clinical scientists would begin to elucidate what particular dimensions of ethnic
identity are most relevant to buffering aspects of health and wellness against the
negative effects of racist victimization and whether these elements vary by
ethnic group membership.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
155
Racism, Anger Rumination and Quality of Life
Anger is deemed one of the “lethal” or “deadly” emotions (Johnson,
1990). In light of this, over the years there has been increasing scholarly
attention paid to clarifying ethnic group differences in anger coping as an
explanation for robust ethnic variation in cardiovascular disease morbidity and
mortality (e.g. Harburg et al., 1973; Dimsdale et al., 1986; Armstead et al.,
1989; Finney et al., 2002). Harburg and colleagues’ seminal work discussed the
importance of considering how the legacy of slavery has shaped the manner with
which African Americans in particular react to the experience of anger (Harburg
et al., 1973). Given the oftentimes life-threatening methods (e.g. lynchings)
enacted as a form of behavioral control in the service of preserving the racial
hierarchy in this country, Blacks have a long history of suppressing their
authentic and justified feelings of anger for fear of retaliation (Harburg et al.,
1973).
Ironically, this adaptive albeit avoidant response to navigating within a
hostile and oppressive environment has likely resulted in increased vulnerability
to later onset of serious physical illness within the African American
community. Less research has explored the link between anger coping and
health indicators in multiple ethnic minority groups although evidence has
accrued which underscores the influence of culture and stage of racial identity
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
156
formation on both anger experience and expression (e.g. Deffenbacher et al.,
1999; Ferrari et al., 2002; Fields et al., 1998; Kelly, 1999; Kim et al., 2004;
Kino, 2000; Koevecses, 2000; Ohbuchi et al., 2002; Perez-Rivera, 2002;
Ramirez et al., 2001; Steele et al., 2003; Tanaka-Matsumi, 1995).
Related to this, there has also been a recent scientific movement aimed at
further articulating why anger inhibition in general tends to be linked to
experiential markers of enhanced cardiovascular reactivity in response to stress
(e.g. Hogan et al., 2004). Authors have advanced anger rumination as a chief
candidate involved in this psychophysiological phenomenon (Hogan et al.,
2004). Anger suppression describes a very broad response style in which an
individual avoids openly expressing this negative affect. Anger rumination
though a related construct (Sukhodolsky et al., 2001) is distinguished as a
specific, active mental process whereby the individual continuously attempts to
make sense of an anger-provoking event by replaying the situation over in one’s
mind and the ruminative content often reflects themes of revenge or retaliatory
ideation (Sukhodolsky et al., 2001). Furthermore it is important to note that
anger rumination is not selectively associated with an anger suppression style. It
has also been shown to be highly positively correlated with a more aggressive
anger out response style (Sukhodolsky et al., 2001).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
157
In essence, anger rumination prolongs the psychological and
physiological intensity of an individual’s experience of anger since the original
eliciting event is transformed into a mental stimulus that may be repeatedly
brought to mind and relived. This anger regulation style is also believed to be
integral in the development, maintenance and elaboration of a hostile cognitive
schema which then has implications for how subsequent events are remembered
and interpreted. The present investigation is unique in that it is the first to
examine potential ethnic group variation in endorsing anger rumination within
an integrative theoretical framework rooted in the historical context of racism-
related victimization and cultural differences in the development, meaning and
function of anger expression.
As anticipated, the stress attributed to racism and the frequent
experiences of racism in one’s lifetime were both significantly positively related
to an anger rumination coping style for the full participant sample. Interestingly,
in contrast to original predictions and to the previously described hostility
findings, no meaningful ethnic group differences emerged on this emotion
regulation variable. This could be an artifact of both study design and how these
two inter-related constructs are conceptualized (i.e. core belief system versus
reactive coping process).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
158
Whereas hostility is viewed as a relatively inflexible cognitive style that
is resistant to situational influences, perhaps this null finding with regard to
anger rumination suggests that differences could emerge if an engaging,
experimental manipulation was presented for the participants to react to (e.g.
assessing anger rumination subsequent to exposure to a racially-provocative
laboratory stressor). In this way, hostility is a more “trait-like” cognitive set
that likely influences the emergence of anger rumination processes. However,
the tendency to ruminate on anger experiences may be more susceptible to
variation in contextual demands and the degree to which life events are deemed
a threat to self-worth.
Although average endorsements of anger rumination did not differ across
participant samples, a variety of regression findings involving this emotion-
focused coping style were shown to significantly differentiate among the ethnic
groups in this study. Within the Asian American participant group for example,
anger rumination significantly interacted with the frequency of lifetime racist
events to predict overall life satisfaction/general health. This finding was similar
to that obtained with trait hostility and provides additional support to the
rejection-protection hypothesis proposed in this analysis. Although at low
frequency of racist encounters, low anger rumination indexed higher quality of
life scores, as the amount of racism experienced increased, only high levels of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
159
anger rumination was marginally associated with improved perceptions of
general health and life satisfaction. Anger rumination was not related to quality
of life for the other three ethnic groups while trait hostility and general college
life stress were controlled for.
This was not however the case for reports of current somatic symptom
distress. Anger rumination as predicted was shown to increase with increasing
somatic complaints only within the African American participant sample even
when adjusted for trait hostility. This extends prior work that has demonstrated
positive links between anger suppression and physical health indicators for
members of the African American community. What the current finding
suggests is that not only might African Americans be more likely to suppress
anger due to historical factors such as slavery and the continued pervasiveness
of racism but it also underscores what active cognitive-affective processes might
be occurring under conditions of anger inhibition that could impact future risk
for cardiovascular disease. However, this intepretation is cautiously suggested in
light of the correlational nature of the data at only one time point.
Here it is also important to note once more that endorsing elevated levels
of hostile beliefs likely makes an individual susceptible to ruminating on anger
events as these variables were highly related in this study. However, this finding
also provides tentative evidence advocating the specific effect of this anger
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
160
regulation style in impairing elements of physical well-being for the African
American participant group. Only experimental and longitudinal analyses in the
future will aid in further establishing this coping style as a mediator in the
relationship between hostility and physiological outcome for this ethnic group.
Regarding family member closeness, brooding on anger more generally
was also shown to negatively predict this interpersonal quality of life dimension
for the African American students. This is also in accordance with original
hypotheses that this ruminative anger regulation style has significant
interpersonal costs. If someone has a tendency to be consumed with anger, this
preoccupation likely adversely affects how engaged an individual is with others
in their environment. Also, although not assessed directly in this study, high
levels of anger rumination have been associated with both increased levels of
aggressive behavior and an anger out response style in previous research (e.g.
Sukhodolsky et al., 2001). Thus this coping style may be more globally related
to deficits in assertive communication.
This phenomenon was further moderated by the level of ethnic identity
reported. The relationship between anger rumination and family member
closeness significantly declined among the African American participants at high
levels of ethnic group affiliation. A similar pattern of results albeit a non
significant trend, also emerged for the Latino American participants.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
161
Interestingly, this would seem to suggest that the strength of ethnic
identity could potentiate or exacerbate rather than attenuate the negative impact
of anger rumination on this interpersonal quality of life dimension. Here is an
example of how elevated endorsements of ethnic identity (in conjunction with a
specific form of anger regulation) may be a corrosive element in effecting
perceptions of the quality of interpersonal relationships.
This expands upon previous research that has shown how variation in
anger expression is associated with particular psychological stages of racial
identity development for African American individuals (Perez-Rivera, 2002).
More specifically, a predominantly Afrocentrically-oriented stage of racial
identity was positively linked to a more volatile, aggressive anger response style
(Perez-Rivera, 2002). Here again it may be the case that anger rumination and a
strong sense of African American identity may negatively effect interpersonal
health more generally due to dysfunctions in communication style.
However, also as predicted, anger rumination moderated the relationship
between one’s lifetime history of racism and perceived family member closeness
for the African American participant group. Thus, in the specific case of racism,
if an individual has a low tendency to ruminate on the anger associated with
these discriminatory events, there was a trend indicating lowered perceptions of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
162
family member closeness. This finding is consistent with the proposed role of
anger due to racism in enhancing the bonds experienced within ethnic minority
families.
This may be especially the case given the predominantly female
participant sample included in this study. Recent theory and research has
advanced the idea that social sharing is a form of “verbal rumination” and
relying on social support in times of stress tends to be impacted by gender.
Thus, it is possible that not only are these African American students replaying
the events in mind, having retaliatory thoughts and trying to make sense out of
the racially-motivated situation mentally, they may also be likely to openly
communicate with family members about what transpired.
It will be important for subsequent work to evaluate the relationships
among anger rumination, aggression and social sharing within the context of
racism-related stress in both correlational and experimental studies where the
objective coding of behaviors and affect may be conducted. This will assist in
further clarifying the interpersonal mechanisms involved in promoting family
cohesiveness and at the same time facilitating augmented vulnerability to
impaired physical wellness for members of certain ethnic minority communities.
Interestingly, the relationships among the frequency of lifetime racist
events, anger rumination and the sense of family member closeness was further
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
163
specified by the strength of ethnic identity for the European American
participants. What this unanticipated finding suggests is that although this ethnic
group reported the lowest frequency of negative life events involving racism,
how one copes with the anger elicited by these events and how important
ethnicity is to one’s identity combined to modulate the impact of racist
victimization on this interpersonal dimension of quality of life. At high levels of
racism experiences, high levels of family member closeness were indexed at
both high anger rumination and low ethnic identity and at low anger rumination
and high ethnic identity for this sample.
In contrast, the two extreme combinations of low anger rumination and
low ethnic identity and high anger rumination and high ethnic identity were
associated with low perceptions of family member closeness at a high frequency
of lifetime racism events. These results build upon the interactions previously
described that were found for the African American participant group. Thus it
would seem that among the Caucasian students, either high ethnic identity or
high anger rumination coping (but neither the combination of the two nor lack of
either element for reasons proposed earlier) may serve to foster greater feelings
of emotional closeness with family members when racially-motivated incidents
are involved. This finding is a unique contribution to the literature which has
largely neglected to explore the roles of ethnic identity, anger regulation and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
164
racism-related events on the impact of quality of life for those of European
American heritage. This is especially surprising given the ever increasing multi
ethnic and multi-cultural landscape of the larger American community. This
changing socio-cultural tapestry creates novel opportunities for majority citizens
to experience racist maltreatment from ethnic minorities and perhaps the college
or university campus is a particularly appropriate microcosm to examine such
transformations in inter-group relations over time.
In terms of friendship closeness, ethnic identity was shown to
significantly interact with an anger rumination coping style to predict this
interpersonal quality of life dimension in the African American participant
group. It would appear that similarly to the findings regarding family member
closeness, high versus low levels of ethnic identity are associated with lower
perceived friendship closeness at high levels of anger rumination for this ethnic
group. Additionally, as anticipated, both the level of anger rumination and the
strength of ethnic identity interacted to impact the relationship between the
frequency of lifetime racist events and perceived friendship closeness within the
African American participant sample only. Here it was shown that friendship
closeness negatively regresses on the amount of racism experienced in one’s
lifetime at the combination of low levels of ethnic identity and low anger
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
165
rumination coping (two factors that at high levels may tend to foster greater
interpersonal closeness when racist incidents are involved).
This theoretical sample is likely most vulnerable to compromised
friendship closeness with others of their own ethnicity when exposed to racism
since their ethnic group membership is not a salient part of their identity nor is
the tendency to experience anger resulting from frequent racial discrimination
(i.e. a culturally normative response to such events). Neither anger rumination
nor ethnic identity were significant moderating factors for the other ethnic
groups regarding affecting the link between lifetime history of personal racism
exposure and this relational quality of life component. This once again points to
the likely more specific functions of anger coping and ethnic identity in
galvanizing interpersonal relatedness particularly within the context of racial
discrimination for members of the African American community.
Racism, Health Values, Health Behaviors and Quality of Life
An acculturative stress model was proposed to be a useful framework for
understanding ethnic and cultural group differences in the frequency of taking
part in health-preventive behaviors. This hypothesis was based on the following
two premises: (1) frequent participation in wellness-enhancing behaviors and
high health value advocacy are considered aspects of mainstream, middle class
European American culture and (2) health values and positive health habits are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
166
in part impacted by socialization processes and acculturation experiences (e.g.
racial discrimination) that may vary based on ethnic group membership and
individual differences in ethnic identity. Several results emerged in support of
this hypothesis.
For example, as anticipated, the European American participants indexed
the highest average reports of regular positive health behavior participation in
comparison to the ethnic minority participant samples in this investigation. This
group also demonstrated higher mean health value endorsements than the Asian
American and Latino American groups. Similar to other recent work, the
average level of health value reported did not meaningfully differentiate the
European American from the African American group (Harris, 2004).
One possible reason for this null finding for this specific pairing could
reflect ethnic group variation in the level of acculturation to more general
dimensions of mainstream American culture. More specifically, perhaps the
African American participants although indicating high levels of ethnic minority
group affiliation, may be the most acculturated ethnic minority group in this
sample. Thus, it will be important for future studies to also explore not only the
role of ethnic identity but also include an assessment of how “majority identity”
may be involved in affecting reports of health values and health behaviors.
Scholars in the area of acculturation have promoted the concept of biculturalism
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
167
competence in which an individual may hold beliefs and be adept at
implementing behaviors (e.g. language, customs) associated with more than one
ethnic or cultural group (e.g. LaFromboise, 1993).
Ethnic identity was also shown to moderate the relationship between the
frequency of lifetime racist events and participation in positive health behaviors
for the African American participant group. As predicted by the acculturative
stress model, at low levels of ethnic identity (and perhaps high levels of
acculturation), more frequent participation in health-promoting behaviors occurs
with the increasing frequency of encountering racial discrimination. This finding
was further specified by the level of health value endorsement for this ethnic
group.
Consistent with hypotheses, high ethnic identity among the African
American group in combination with high health value was associated with
reductions in the frequency of taking part in beneficial health behaviors with
increasing lifetime history of racism. An opposite pattern emerged for the
combination of low ethnic identity and high health value. In this case, low levels
of ethnic identity for this minority group may be a protective factor against the
adverse effects of racist maltreatment in part by promoting greater value of and
involvement in positive health habits. Given the nature of the findings obtained
here, it will be important for subsequent projects to develop process models
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
168
through examining the relationships among racism experiences, ethnic identity
formation and health value/health behavior socialization and participation over
time especially with regard to members of the African American community.
Thus, the impact of racism for this ethnic group in particular may have more
specific effects on dimensions of health and wellness than previously
anticipated.
A second objective of this portion of the paper was to investigate
whether both constructive along with more health-compromising behaviors
would serve as moderators in the relationship between racism-related stress and
components of quality of life from both classic stress and coping (e.g. Lazarus
& Folkman, 1984) and more contemporary emotion regulation perspectives
(e.g. Gross, 2002). First, in accordance with hypotheses and previous research
in this content domain (e.g. Levy, 2003), the frequency of participating in
wellness-promoting behaviors was identified as a significant positive predictor
of general health and life satisfaction reports for all ethnic groups included. The
specific mechanism involved remains to be fully clarified but at least in part may
reflect improvements in positive affect or mood.
Additionally, this variable interacted with the level of ethnic identity to
predict perceived quality of life for the Latino American participants only. As
the frequency of taking part in positive health habits increased, quality of life
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
169
estimations also increased both at high and low levels of ethnic identity for this
group. However, the effect was stronger at higher endorsements of ethnic group
affiliation. Here results could be highlighting a health-preventive asset
associated with higher degrees of ethnic identity for members of the Latino
American community.
Frequent participation in beneficial health behaviors was not a
meaningful negative predictor of somatic symptom distress for any of the
participant samples as originally anticipated. However, it did significantly
interact with the degree of racism-related stress reported to predict variance in
this physical health-related quality of life dimension within select ethnic groups.
Interestingly and inconsistent with predictions, greater positive health behavior
activity did not attenuate the link between the stress attributed to racism and
somatic symptom distress for the African American sample.
Although at low levels of racism-related stress, infrequent participation
in health-promoting behaviors indexed higher somatic distress reports than at
high participation, as the stress resulting from racist encounters increased so did
somatic distress at high frequency of positive health habits. It is possible that a
high level of positive health behavior engagement is a more general indicator of
stress reactivity and/or somatic awareness for members of this ethnic minority
group. Only experimental and/or longitudinal research may help tease apart
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
170
whether these individuals are adaptively responding to physical complaints
instigated by racism-related stress through more frequent participation in health-
preventive behaviors.
An inverse pattern of findings emerged for the Asian American
participants with respect to these same variables. As expected, as racism-related
stress increased so did reports of somatic complaints at a low frequency of
positive health behaviors only. This would seem to suggest that infrequent
participation in health-preventive behaviors may make individuals of Asian
American heritage more vulnerable to the experience of physical symptoms
resulting from intense levels of stress due to racist victimization. Alternatively,
a low level of beneficial health habit participation may represent a more global
marker of somatic sensitivity and/or stress reactivity for members of this ethnic
minority group.
This is one of the first analyses to begin to describe the potential
interpersonal benefits of positive health behavior participation within an
ethnically-diverse sample. In general, constructive health habits were indicated
as significant positive predictors of perceived family member closeness for the
African American, European American and Latino American participants. More
specifically, a low level of positive health behavior activity (and possibly
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
171
positive affectivity) was associated with a diminished sense of family member
relationship satisfaction with increasing levels of racism-related stress among the
Caucasian participants.
Additionally, frequent participation in wellness-enhancing behaviors was
also demonstrated as a robust predictor of the variance in perceived friendship
closeness for both the African American and European American participant
groups. Within the Asian American sample, more regular positive health habit
participation was associated with a greater sense of friendship closeness at high
levels of ethnic group affiliation. This last finding highlights the relationship
between ethnic identity and a specific coping strategy (i.e. taking part in well
being improving behaviors) in fostering stronger relational bonds among this
Asian American participant group. Further clarifying the underlying affective
(e.g. changes in positive and/or negative mood, stress reactivity) and
interpersonal (e.g. adaptive health behavior participation as a form of social
activity) mechanisms involved in this process and how cultural factors may
influence their existence are fruitful areas to target in subsequent research.
The two principal maladaptive health behaviors assessed in the current
research included the frequency of drinking until feeling ill (i.e. a rudimentary,
subjective indicator of binge-like drinking) and the frequency of cigarette
smoking behavior. Unlike the number of findings obtained for health-enhancing
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
172
behaviors discussed previously, analyses including these negative health habits
yielded significantly fewer meaningful results. For instance, in contrast to
research that has indicated important ethnic group differences in the
endorsement of health-impairing behaviors (see Harris, 2004 for a discussion),
the present study failed to identify significant distinctions in the average level of
drinking and smoking frequency among the four ethnic groups investigated.
It should be noted that both 1-item scales included in this study solely
measured thc frequency of the behavior within the past month and neither the
actual amount regularly consumed on a daily or weekly basis or the duration of
consumption (i.e. substance use history) were assessed. It is possible that these
global instruments were not sensitive enough to detect potential ethnic group
differences on these constructs. Moreover, drinking and smoking frequency
were rarely indicated as main effects in predicting dimensions of quality of life
when status, demographic, general stress and hostility variables were controlled
for. This is likely due to the overlap in variance already accounted for by the
psychosocial covariates included in the regression analyses.
Despite this dearth of negative health behavior findings, and as predicted
the tendency to drink until one felt ill did explain a significant proportion of the
variance in life satisfaction/overall quality of life reports within the African
American participant group. It is unclear as to why this result emerged for this
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
173
ethnic group in particular. It may reflect possible variation in how different
ethnic groups appraise the appropriateness of alcohol consumption and the
degree to which alcohol use is integrated within the cultural habits and traditions
of various ethnic communities.
Perhaps frequent periods of drinking until feeling physically ill are
considered more normative, ego-syntonic behaviors among the European
American, Asian American and Latino American participants (possibly within
the specific context of the college environment). Whereas it is possible that such
behavior is not as socially sanctioned and therefore experienced as more ego-
dystonic for the African American participants. However, these are hypotheses
to be investigated further in future research.
Within the Latino American sample, frequently drinking until feeling
sick interacted with the level of racism-related stress experienced to predict
variance in perceived health/quality of life ratings. More specifically and
somewhat consistent with expectations, it was found that greater life satisfaction
was evident at low levels of racism-elicited stress for low frequency drinking
than regularly participating in more frequent drinking behavior. Interestingly,
this main effect did not remain at high levels of racism-related stress for
members of this ethnic group.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
174
In addition to the psychological consequences of binge-like drinking
described thus far, this wellness-compromising behavior also impacted
interpersonal aspects of life satisfaction. For the Asian American sample,
greater reports of perceived family member closeness at low levels of racism-
related stress were higher for high versus low frequency of drinking until feeling
ill. This unanticipated result raises the question of how such drinking behavior
may be correlated between the student participants and designated family
members, a relationship that was not evaluated in the current analysis. Thus,
although destructive in some respects with regard to physical health, drinking
until feeling ill may also be associated with social benefits for certain ethnic
groups within the context of ethnicity-relevant stress. This is certainly an area
that will need to be addressed more comprehensively in subsequent empirical
work.
With regard to cigarette smoking frequency, this variable significantly
moderated the link between racism-related stress and perceived quality of life
among the Latino American participants. Also in contrast to a priori hypotheses,
greater overall life satisfaction endorsements were indexed at high levels of
racism-related stress for high versus low smoking behavior for this ethnic
minority group. What this main effect could suggest is that more frequent
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
175
smoking functions as a method of improving mood (and as a result enhancing
perceived quality of life estimations) in reaction to highly aversive racism
experiences among the Latino American sample.
As predicted, the frequency of cigarette smoking did positively predict
variation in somatic symptoms within the Asian American participant group.
This health-destructive behavior was also shown to significantly interact with
racism-related stress to predict perceived family member closeness in this ethnic
minority sample. Also in accordance with expectations, perceived family
member closeness increased as the intensity of racism-related stress increased at
low smoking frequency. Due to the nature of these analyses, it remains to be
clarified why this is the case. This result could reflect variation due to a third
variable such as stress reactivity, negative affectivity or some other
temperamental component, communication style or simply social acceptance of
smoking by the designated family members reported by the Asian American
participants included in this project.
The essential objective addressed in this part of the study was to identify
individual differences in elements of positive and negative health habits as
moderators in the association between the stress resulting from racist
victimization and dimensions of quality of life. Clearly, future work will benefit
from exploring process models that assess racism’s impact on health via both
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
176
adaptive and maladaptive health behavior mechanisms and how these
relationships may be influenced by ethnicity and within group variation in ethnic
identity and level of acculturation. As such, the psychosocial covariates (e.g.
SES, general stress, hostility) controlled for in these moderator-focused
regression analyses may alternatively serve in future work to be the independent
and/or dependent variables in more complex, culturally-sensitive health behavior
models that explain relationships among these variables over time.
Perceived Racism, Hostility, Anger and Aggression in Ambiguous Situations
The tendency to infer hostile motivations on the parts of others in
everyday, ambiguous situations has received much attention in the socio-
cognitive psychological literature over the past fifteen years (e.g. Dodge, 1980).
This hostile attribution bias has been consistently demonstrated in a variety of
participant samples ranging in ages and to a lesser extent ethnicity. However,
the current project sought to address two gaps in this corpus of research that had
yet to be explored. This included (1) the possibility this cognitive
interpretational bias may be confounded with perceptions of racism when
examined in ethnically diverse groups and (2) that the willingness to consider
situational or contextual factors amidst the post-event processing phase (i.e. an
indicator of cognitive flexibility) may modulate feelings of anger and aggressive
ideation following such daily interpersonal encounters.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
177
This last proposition is consistent with standard Beckian cognitive
restructuring techniques that aim to reduce anger intensity through promoting a
more flexible attributional style that incorporates an openness to examining
alternative explanations for the actions of others. Additionally, this phenomenon
was further specified by cultural or ethnic background based on the frame
switching phenomenon discussed previously. It was anticipated that the
endorsement of situational attributions would be most prominent among the
Asian American participants as a reflection of a potentially more flexible,
bicultural cognitive style that emphasizes the importance of context in
influencing the emergence of either individualistic (e.g. dispositional) or
collectivistic (e.g. situation-specific) attributional styles. Here it was of interest
to begin to clarify how these processes may work within the naturalistic context
of perceived racism. Several findings were obtained that were in support of
original hypotheses.
First of all, significant ethnic group differences emerged with regard to
all of the variables investigated in this portion of the study including the
tendency to perceive racism and to attribute situational explanations to the
ambiguously-motivated behaviors of others in the ten vignettes. As predicted,
the African American participants evidenced the highest mean reports of a
willingness to attribute racism to the interpersonal scenarios while the Caucasian
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
178
students on average indexed the lowest reports of perceived racism.
Additionally, the European American and Asian American participants both
exhibited a greater tendency to endorse nondispositional factors as a rationale
for the negative outcomes than the African American and Latino American
participant groups.
However, these two ethnic groups did diverge in their mean reports of
inferring hostile intentions on the parts of others in the vignettes. In this case,
the Asian American participants like the other ethnic minority groups included
in the study were more likely to exhibit a hostile attribution bias in reaction to
the events than the European American students. The Caucasian students also
expressed the lowest anger and the least frequent aggressive thoughts in
response to the ambiguously-motivated situations. These results are not
surprising given the significant ethnic group variation in past history of racism
reported which in previous research was shown to impact reports of perceived
racial bias in ethnic minority samples (e.g. Branscombe et al., 1999).
In light of the anticipated ethnic group differences in the situational
variables, it was of interest to investigate the link between perceptions of racist
maltreatment and the tendency to impart hostile motivations on others in
everyday, ambiguous situations. While controlling for covariates including
status variables, stress, hostility and prior exposure to racism, the frequency of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
179
racism attributions made (i.e. perceived racism) emerged as a robust predictor
of the frequency of hostile attributions across ethnic groups with a greater
proportion of the variance accounted for in the African American and Latino
American participant samples. Moreover, correlational analyses revealed very
high positive associations between these two variables with coefficients ranging
from .63 in the Caucasian group to .84 in the Latino American group. Together,
these findings provide limited evidence which suggests perceptions of racism
may be considered an ethnicity-specific form of hostile attribution that
researchers in this area may need to pay closer attention to when investigating
this phenomena in ethnically-diverse samples. Further, it alerts researchers to
the importance of clarifying the ethnicity of the person in the vignettes which
could significantly impact hostile attributions, anger and aggressive intention
ratings on the parts of participants due to inferences of racism.
Finally, a few noteworthy interactions were identified involving the
tendency to attribute nondispositional factors or degree of ethnic identity in
influencing the relationships among perceived racism and post-event anger and
aggressive thinking. Within the Asian American group, a low contextual
attribution style indexed a high degree of anger regardless of the frequency of
perceived racism. However, at high levels of situational attributions,
significantly lower anger was evidenced at a lower frequency of perceived
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
180
racism, which was shown to increase to the level of anger endorsed at low
situational attribution reports with increasing racism attributions.
Thus, in partial support of hypotheses, the ability to endorse alternative
explanations for negative outcomes mitigated the link between low levels of
perceived racism and post-event anger in this ethnic group. However, the
willingness to consider situational influences on behavior was not an anger
buffer at higher tendencies of perceiving racism. What instead may be occurring
is an anchoring phenomenon whereby contextual factors exacerbate rather than
reduce the affective reaction when the individual has already committed to the
belief that what happened resulted from racism. Anything else may be construed
as a conjunctive feature or reason instead of as a truly alternative explanation. In
other words if someone rigidly believes that racial prejudice is the primary
rationale for another’s behavior, admitting that person may be having a bad day
does not mitigate the anger evoked by this initial belief. Instead it is likely
thought of as this racist person who is also having a bad day.
Regarding aggressive ruminations in the aftermath of these ambiguous
events, as expected, it was shown that among the Caucasian participants, a high
situational attribution tendency was associated with lower levels of aggression
regardless of the frequency of perceived racism. However, a low situational
attribution style was significantly linked with increasing aggressive ideation as
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
181
the frequency of attributing racism to the vignettes increased. This result was
mirrored by an interaction effect indicated for the level of ethnic identity in this
group. More specifically, a greater endorsement of physically aggressive,
retaliatory thinking was exhibited at low ethnic identity with increasing
perceptions of racism in the ambiguous vignettes.
These findings together suggest that within the European American
participant sample, low ethnic identity and a low situational attribution tendency
may reflect a more specific distress intolerance style when situations are
believed to involve racial discrimination, a type of life event that this group as a
whole has less exposure to. Alternatively, these dimensions may represent more
general deficits in an assertive communication style. Perhaps due to the lower
frequency of lifetime racist events indicated by members of this ethnic group
(and therefore potentially lowered expectations of receiving racist
maltreatment), the reaction to events perceived to be racially-motivated is
exacerbated based on limited cognitive flexibility and a lack of cohesive sense of
ethnic identity. This intriguing finding would have been predicted by the
rejection-identification model whereby strong ethnic group affiliation serves as a
buffer against the negative impact of perceived racism which in essence is a
direct attack against this self-concept dimension. Interestingly, the role of ethnic
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
182
identity in psychological health among European American groups has not
received as much scholarly attention as it has within ethnic minority samples.
Lastly, also in partial accordance with the cognitive flexibility
hypothesis, less aggressive rumination was reported at higher levels of
situational attributions when a lower frequency of perceived racism was
endorsed within the Asian American participant sample. There was a
marginally-significant trend similar to the interaction effect described for post
event anger in which aggressive ideation was shown to increase for the high
situational attribution group to the level indexed by the low situational
attribution group with the increasing frequency of racism attributions reported.
Here again, likely an anchoring effect is occurring that does not translate into
diminishing the likelihood of aggressive, revenge intentions reported at high
levels of contextual attributions when an individual has previously strongly
endorsed a high likelihood of perceived racism.
SUMMARY
Several findings from the present study provided support for all three
primary theoretical frameworks proposed that sought to examine the relationship
between racism and dimensions of quality of life within an ethnically-diverse
sample from multiple levels of analysis (i.e. through standard, retrospective
self-report instruments in conjunction with a more novel, prospective, Situation-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
183
based approach). More specifically, evidence emerged consistent with the
rejection-protection, the acculturative stress, and the cognitive flexibility
hypotheses. In addition, ethnicity-specific results pointed to both differences in
average endorsements on the primary variables of interest along with
highlighting variation in the moderation of relationships between aspects of
socio-cultural experience and elements of life satisfaction among the four ethnic
groups included. In general, these findings may serve to facilitate the continued
development of multicultural models of racism and well-being and in so doing
underscore important areas involved in enhancing health-promotion and disease-
prevention efforts among various underserved ethnic minority populations.
LIMITATIONS AND AREAS OF FUTURE RESEARCH
This investigation based in a multicultural, functional-developmental
framework provided an original empirical contribution and an integrative
theoretical perspective to the existing racism and wellness literature. Despite
this, it is also important to note the limitations of this project in order to set the
stage for future scientific work in this area. First of all, the study sample was
primarily composed of female participants. As investigators involved in
multicultural and ethnic minority research know well, it is often a challenge to
successfully recruit sufficient numbers of male participants. Thus, it will be
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
184
important for subsequent research to continue to expand and clarify whether the
models tested here may be further specified by unique ethnicity x gender
interactions.
Secondly, the design of this study was correlational in nature and data
was collected at only one point in time. Given that the results are discussed
within an acculturative-developmental framework, it is critical to test these
effects in experimental work and to assess changes in these relationships
longitudinally following several cohorts over time. This will continue to enhance
our understanding of the socio-cultural (i.e. both inter-ethnic and intra-ethnic)
mechanisms involved in health-related belief system/relational schema formation
(i.e. hostility), the expression and meaning of anger coping (e.g. rumination,
aggression), the value of regularly practicing beneficial wellness-promoting
behaviors (e.g. health socialization processes) and on the flexibility of
attributional style when encountering everyday, ambiguous situations that may
involve racism.
Thirdly, it was a chief objective to solely examine the relationship of
one’s personal exposure to racism to dimensions of well-being. However, there
are several ways in which an individual may encounter racist experiences that
not only contribute to the development of ethnic group affiliation but also likely
shape the creation and maintenance of hostile cognitive schemas that serve to
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
185
protect and prepare the individual for living in a racially-noxious environment.
These include a wide array of informal, vicarious means such as through the
media (e.g. news, music, movies), as a witness or bystander to the racist
victimization of another and for example through hearing traumatic stories or
cautionary tales passed down from grandparents, other relatives and friends over
generations (Harrell, 2000).
Related to this last vehicle, there is also an interesting construct that has
recently emerged within the larger discourse on racism and health. The concept
of racial socialization (Peters, 1985) describes a more formal process that
involves “the tasks Black parents share with all parents—providing for and
raising children...but include the responsibility of raising physically and
emotionally healthy children who are Black in a society in which being Black
has negative connotations” (p. 161). Racial socialization messages (including
racism awareness preparation) were found to impact aspects of ethnic identity
development (Anglin, 2004), acculturative stress (Thompson, Anderson &
Bakeman, 2000) and were noted to vary between the parents of African
American and Hispanic American children and teens (Hughes, 2003).
Assessing the relationships among racial socialization messages, ethnic
identity and hostility in predicting dimensions of well-being both cross-
sectionally and longitudinally in a variety of ethnic groups will be a very fruitful
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
186
area for later scholarly work to target. Within the context of the rejection-
protection theory posited in this study, it is possible that racial socialization
affects the links among hostility, racism and wellness through influencing
expectancies about racist victimization. More generally, it will be important to
investigate how the associations between both personal and vicarious
experiences of racism and quality of life may be modified by cultural differences
in cognition, affect and behavior.
The literature exploring ethnic differences in health values and health-
preventive behavior participation is in its infancy. This area is ripe for further
analysis of the role of ethnic and cultural variation in longevity beliefs which
may further impact health beliefs and engagement in wellness-promoting
behaviors. An additional important avenue to research includes identifying
relevant mediating variables involved in the relationships among racism, health
values and beneficial health behaviors (e.g. mistrust of healthcare professionals,
access to health care and availability of accurate health-related information).
Another caveat that merits attention is the possibility that the results
obtained regarding the ambiguous vignettes are artifacts of study design. The
correlations found among perceptions of racism, contextual attributions, hostile
attributions, anger and aggressive ideation could have been inflated by the fixed
sequence of item presentation. Future qualitative work (e.g. thought listing) will
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
187
aid in further clarifying the relationship between racism attributions and the
tendency to infer hostile or benign motives and how these factors modulate
anger intensity and retaliatory cognitions in a variety of ethnic groups. In
addition, a more direct examination of the effect of broader cultural dimensions
such as individualism and collectivism on the existence of an inflexible hostile
attribution style and/or a more flexible contextual sensitivity within the
heterogeneous Asian American community is warranted.
The instrument used to measure racist life events in this study didn’t
permit a more fine grain assessment of possible qualitative distinctions in the
experiences of racism among the four ethnic groups. This in turn could impact
the intensity of stress evoked by the event. For instance, one recent study found
that the type of racist encounter reported varied by ethnic group membership
(Contrada et al., 2001). The Asian American participants reported more
frequent overt aggressive incidents related to interracial conflict while the
African American participants in that study indicated greater covert perceptions
of social alienation attributed to racial discrimination (Contrada et al., 2001).
Subsequent research would benefit from evaluating the frequency of a fuller
range of racially discriminatory events (e.g. everyday microstressors v. hate
speech v. hate crime victimization) within an ethnically-diverse sample.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
188
A final limitation important to mention is that the interaction effects
yielded from the regression analyses discussed throughout were minimal in size
accounting for 10% or less of the variance. It was important for this study to
provide a more stringent test of moderation by adjusting for important covariates
which previous research largely has neglected. Thus, it is likely that these are
conservative estimates of these effects.
CONCLUSIONS AND IMPLICATIONS
In general, several results from this multidimensional analysis emerged
to lend support to the overall importance of expanding the conceptualization and
meaning of psychological/interpersonal-resilience factors within the specific
context of racism experiences while these same factors may still significantly
contribute to risk for impaired physical health. In essence, findings revealed
how particular cognitive and emotion regulation styles that traditionally have
been positively associated with stress vulnerability (see Miller et al., 1996 for a
review), interpersonal difficulties (e.g. Cui et al., 2002; Brondolo et al., 2003;
Holman et al., 2003; Katz et al., 2002) and objective markers of compromised
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
189
health status (e.g. Miller et al., 1996) may function as psychological and
interpersonal protective factors for ethnic minority individuals when life
experiences involve racist victimization.
Thus, results from this analysis more globally advocate adopting a
broader, more multiculturally sensitive perspective on how health-related
personality and behavioral variables are interpreted in research that includes
people of color. This is in support of recent theory and research that has also
strongly encouraged redefining personality traits within a dynamic constructivist
approach that takes into account the influence of the cultural context in shaping
beliefs and the development and construal of shared behavioral scripts (e.g.
regarding trait hostility: Hemmings, 2002; Miller, 2002).
Consideration of this novel perspective in turn has significant
implications for example in enhancing the effective delivery of mental health
services and promoting motivation for positive health behavior change among
groups for whom the historical context of racism plays a meaningful role in
affecting one’s openness towards interacting with and one’s willingness to
receive and carry out health recommendations by persons recognized as
representing the “White establishment” (e.g. the previously noted “historical
hostility” phenomenon in psychotherapy with African American individuals:
Vontress et al., 1997). Given the pervasive legacy of slavery and racism in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
190
United States, findings from this project also strongly urge health care providers
to become more creative (e.g. enlisting the help of churches and local
community organizations) in the methods of assisting various ethnic groups in
enhancing their level of acculturation into both valuing health in the long-term
and actively participating in health-preventive behaviors (e.g. breast self-exams,
colon screenings, diet and exercise). Helping to better educate health care
professionals in a variety of settings of the shades of meanings of such
constructs as hostility and anger expression for different ethnic groups could
work towards improving communication and reducing mistrust between health
care providers and ethnic minority individuals who are seeking treatment.
Additionally, the results obtained from this study lay the foundation for future
scholarly inquiry into further specifying the cognitive, affective, and behavioral
mechanisms involved in ethnic identity development within various cultural
groups which will assist in continuing to identify under what set of conditions or
circumstances individual differences in ethnic identity function as wellness-
enhancing versus wellness-debilitating factors.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
191
REFERENCES
Adams, J. P., Jr. & Dressier, W. W. (1988). Perceptions of injustice in
a Black community: Dimensions and variation. Human Relations, 41, 753-767.
Aiken, L. S. & West, S. G. (1996). Multiple regression: Testing and
interpreting interactions. Thousand Oaks, CA: Sage Publications.
Anglin, D. M. (2004). The predictive effects of racial socialization,
racial identity, and African self-consciousness on African American college
adjustment and psychological defenses. Dissertation Abstracts International, 64
(7-B), 3512.
Anderson, C. A. (1999). Attributional style, depression, and loneliness:
A cross-cultural comparison of American and Chinese students. Personality and
Social Psychology Bulletin, 25, 482-499.
Anderson, L. P. (1991). Acculturative stress: A theory of relevance to
Black Americans. Clinical Psychology Review, 11, 685-702.
Appel, S. J., Harrell, J. S., & Deng, S. (2002). Racial and socio
economic differences in risk factors for cardiovascular disease among southern
rural women. Nursing Research, 51, 140-147.
Armstead, C., Lawler, K., Gordon, G., Cross, J., & Gibbons, J.
(1989). Relationship of racial stressors to blood pressure responses and anger
expression in Black college students. Health Psychology, 8, 541-556.
Baden, A. D. & Howe, G. W. (1992). Mothers’ attributions and
expectancies regarding their conduct-disordered children. Journal o f Abnormal
Child Psychology, 20, 467-485.
Barefoot, J. C. (1992). Developments in the measurement of hostility. In
H. S. Friedman (Ed.). Hostility, coping, and health (pp. 13-31). Washington,
D.C.: American Psychological Association.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
192
Barefoot, J. C., Peterson, B. L., Dahlstrom, W. G., Siegler, I. C. et al.
(1991). Hostility patterns and health implications: Correlates of Cook-Medley
Hostility Scale scores in a national survey. Health Psychology, 10, 18-24.
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New
York: Harper & Row.
Bermudez, J. & Perez-Garcia, A. M. (1996). Cardiovascular reactivity,
affective responses and performance related to the risk dimensions of coronary-
prone behavior. Personality and Individual Differences, 21, 919-927.
Berrigan, D., Dodd, K., Troiano, R. P., Krebs-Smith, S. M., &
Ballard, R. (2003). Patterns of health behavior in U. S. adults. Preventive
Medicine: An International Journal Devoted to Practice & Theory, 36, 615-623.
Best, J. L. (2001). The relationship o f hardiness and racial identity
attitudes to the articulated thoughts and psychological impact o f simulated
racism encounters. Unpublished masters thesis. University of Southern
California.
Bianchi, F. T., Zea, M. C., Poppen, P. J., Reisen, C. A., & Echeverry,
J. J. (2004). Coping as a mediator of the impact of sociocultural factors on
health behaviour among HIV-positive Latino gay men. Psychology & Health,
19, 89-101.
Blascovich, J., Spencer, S. J., Quinn, D., & Steele, C. (2001). African
Americans and high blood pressure: The role of stereotype threat. Psychological
Science, 12, 225-229.
Bowen-Reid, T. L. & Harrell, J. P. (2002). Racist experiences and
health outcomes: An examination of spirituality as a buffer. Journal o f Black
Psychology, 28, 18-36.
Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999).
Perceiving pervasive discrimination among African Americans: Implications for
group identification and well-being. Journal of Personality and Social
Psychology, 77, 135-149.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
193
Brondolo, E., Rieppi, R., Erickson, S. A., Bagiella, E. et al. (2003).
Hostility, interpersonal interactions, and ambulatory blood pressure.
Psychosomatic Medicine, 65, 1003-1011.
Bullock, S. C., & Houston, E. (1987). Perceptions of racism by Black
medical students attending White medical schools. Journal o f the National
Medical Association, 79, 601-608.
Burns, J. W. & Katkin, E. S. (1993). Psychological, situational, and
gender predictors of cardiovascular reactivity to stress: A multivariate approach.
Journal o f Behavioral Medicine, 16, 445-465.
Buss, A. H. & Durkee, A. (1957). An inventory for assessing different
kinds of hostility. Journal o f Consulting Psychology, 21, 343-349.
Buss, A. H. & Perry, M. (1992). The aggression questionnaire. Journal
o f Personality and Social Psychology, 63, 452-459.
Caprara, G. V. (1986). Indicators of aggression: The dissipation-
rumination scale. Personality and Individual Differences, 7, 763-769.
Caprara, G. V., Barbaranelli, C., Colombo, G., Politi, A., & Valerio,
P. (1995). Emotional susceptibility, irritability and hostile rumination as
correlates of coronary heart disease. Personality and Individual Differences, 19,
569-575.
Carlsson, M. & Hamrin, E. (1996). Measurement of quality of life in
women with breast cancer: Development of a life satisfaction questionnaire
(LSQ-32) and a comparison with the EORTC QLQ-C30. Quality of Life
Research, 5, 165-174.
Cassidy, T. (2000). Stress, healthiness and health behaviours: An
exploration of the role of life events, daily hassles, cognitive appraisal and the
coping process. Counselling Psychology Quarterly, 13, 293-311.
Chon, K. K. (2002). Cultural aspects of anger. In: C. von Hofsten & L.
Backman (Eds.), Psychology at the turn of the millennium, vol. 2: Social,
developmental, and clinical perspectives (pp. 323-346). Florence, KY: Taylor &
Frances/Routledge.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
194
Clark, R.., Anderson, N. B., Clark, V. R., & Williams, D. R. (1999).
Racism as a stressor for African Americans: A biopsychosocial model.
American Psychologist, 54, 805-816.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2n d ed.). Hillsdale, NJ: Erlbaum.
Cole, P. M., Bruschi, C. J., & Tamang, B. L. (2002). Cultural
differences in children’s emotional reactions to difficult situations. Child
Development, 73, 983-996.
Collins, K. & Bell, R. (1997). Personality and aggression: The
dissipation-rumination scale. Personality and Individual Differences, 22, 751-
755.
Contrada, R. J., Ashmore, R. D., Gary, M. L., Coups, E., Egeth, J.
D., Sewell, A., Ewell, K., & Goyal, T. M. (2001). Measures of ethnicity-
related stress: Psychometric properties, ethnic group differences, and
associations with well-being. Journal o f Applied Social Psychology, 31, 1775-
1820.
Courtenay, W. H., McCreary, D. R., & Merighi, J. R. (2002). Gender
and ethnic differences in health beliefs and behaviors. Journal o f Health
Psychology, 7, 219-231.
Crocker, J., Luhtanen, R., Broadnax, S., & Blaine, B. E. (1999). Belief
in U.S. government conspiracies against Blacks among Black and White college
students: Powerlessness or system blame? Personality and Social Psychology
Bulletin, 25, 941-953.
Crocker, J. & Major, B. (1989). Social stigma and self-esteem: The self-
protective properties of stigma. Psychological Review, 96, 608-630.
Crocker, J., Voelkl, K., Testa, M., & Major, B. (1991). Social stigma:
The affective consequences of attributional ambiguity. Journal o f Personality
and Social Psychology, 60, 218-228.
Cross, W. E. (1971). Negro-to-Black conversion experience. Black
World, 20, 13-27.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
195
Cross, W. E. (1991). Shades of Black: Diversity in African-American
identity. Philadelphia: Temple University Press.
Cui, M., Conger, R. D., Bryant, C. M., & Elder, G. H. (2002).
Parental behavior and the quality of adolescent friendships: A social contextual
perspective. Journal of Marriage & Family, 64, 676-689.
Davey, G. C. L. (1995). Rumination and the enhancement of fear: Some
laboratory findings. Behavioural and Cognitive Psychotherapy, 23, 203-215.
Davidson, K. W., Prkachin, K. M., Lefcourt, H. M., & Mills, D. E.
(1996). Towards a psychosocial mediator model of hostility and CHD.
Psychology & Health, 11, 421-429.
Davison, G. C., Vogel, R. S., & Coffman, S. G. (1997). Think-aloud
approaches to cognitive assessment and the Articulated Thoughts in Simulated
Situations paradigm. Journal o f Consulting and Clinical Psychology, 65, 950-
958.
Deffenbacher, J. L. & Swaim, R. C. (1999). Anger expression in
Mexican American and White non-Hispanic adolescents. Journal o f Counseling
Psychology, 46, 61-69.
Diong, S. & Bishop, G. D. (1999). Anger expression, coping styles, and
well-being. Journal o f Health Psychology, 4, 81-96.
Dodge, K. A. (1980). Social cognition and children’s aggressive
behavior. Child Development, 51, 162-170.
Dodge, K. A. (1986). A social information processing model of social
competence in children. In M. Perlmutter (Ed.). Eighteenth Annual Minnesota
Symposium on Child Psychology (pp. 77-125). Hillsdale, NJ: Erlbaum.
Dodge, K. A. & Somberg, D. R. (1987). Hostile attributional biases
among aggressive boys are exacerbated under conditions of threats to self. Child
Development, 58, 213-224.
Dunbar, E. (2001). Counseling practices to ameliorate the effects of
discrimination and hate events: Toward a systemic approach to assessment and
intervention. The Counseling Psychologist, 29, 279-307.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
196
Durel, L. A., Carver, C. S., Spitzer, S. B., Llabre, M. M. et al.
(1989). Associations of blood pressure with self-report measures of anger and
hostility among Black and White men and women. Health Psychology, 8, 557-
575.
Earle, T. L., Linden, W., & Weinberg, J. (1999). Differential effects of
harassment on cardiovascular and salivary cortisol stress reactivity and recovery
in women and men. Journal of Psychosomatic Research, 46, 125-141.
Ehlers, A., Mayou, R. A., & Bryant, B. (1998). Psychological
predictors of chronic posttraumatic stress disorder after motor vehicle accidents.
Journal o f Abnormal Psychology, 107, 508-519.
Ellis, A. (1962). Reason and emotion in psychotherapy. New York:
Citadel Press.
Epps, J. & Kendall, P. C. (1995). Hostile attributional bias in adults.
Cognitive Therapy and Research, 19, 159-178.
Essed, P. (1990). Everyday racism: Reports from women o f two cultures.
Claremont, CA: Hunter House Inc. Publishers.
Essed, P. (1992). Alternative knowledge sources in explanations of racist
events. In M. L. McLaughlin, M. J. Cody & S. J. Read (Eds.). Explaining
one’ s self to others: Reason-giving in a social context (pp. 199-224). Hillsdale,
New Jersey: Lawrence Erlbaum Associates.
Feagin, J. R. (1991). The continuing significance of race: Anti-Black
discrimination in public places. American Sociological Review, 56, 101-116.
Feagin, J. R. & Sikes, M. P. (1994). Living with racism: The Black
middle-class experience. Boston: Beacon Press.
Felton, G. M., Parsons, M. A., & Bartoces, M. G. (1997).
Demographic factors: Interaction effects on health-promoting behavior and
health related factors. Public Health Nursing, 14, 361-367.
Felton, G. M., Parsons, M. A., Misener, T. R., & Oldaker, S. (1997).
Health-promoting behaviors of Black and White college women. Western
Journal o f Nursing Research, 19, 654-666.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
197
Ferrari, M. & Koyama, E. (2002). Meta-emotions about anger and
amae: A cross-cultural comparison. Consciousness & Emotion, 3, 197-211.
Fields, B., Reesman, K., Robinson, C., Sims, A. et al. (1998). Anger of
African American women in the South. Issues in Mental Health Nursing, 19,
353-373.
Finney, M. L., Stoney, C. M., & Engebretson, T. O. (2002). Hostility
and anger expression in African American and European American men is
associated with cardiovascular and lipid reactivity. Psychophysiology, 39, 340-
349.
Fisher, L., Gudmundsdottir, M., Gilliss, C., Skaff, M. et al. (2000).
Resolving disease management problems in European-American and Latino
couples with type 2 diabetes: The effects of ethnicity and patient gender. Family
Process, 39, 403-416.
Flory, J. D., Matthews, K. A., & Owens, J. F. (1998). A social
information processing approach to dispositional hostility: Relationships with
negative mood and blood pressure elevations at work. Journal o f Social and
Clinical Psychology, 17, 491-504.
Foster, M. D. & Matheson, K. (1999). Perceiving and responding to the
personal/group discrimination discrepancy. Personality and Social Psychology
Bulletin, 25, 1319-1329.
Fredrickson, B. L., Maynard, K. E., Helms, M. J., Haney, T. L. et al.
(2000). Hostility predicts magnitude and duration of blood pressure response to
anger. Journal o f Behavioral Medicine, 23, 229-243.
Friedman, R., Schwartz, J. E., Schnall, P. L., Landsbergis, P. A.,
Pieper, C., Gerin, W., & Pickering, T. G. (2001). Psychological variables in
hypertension: Relationship to causal or ambulatory blood pressure in men.
Psychosomatic Medicine, 63, 19-31.
Gallant, M. P. & Dorn, G. P. (2001). Gender and race differences in the
predictors of daily health practices among older adults. Health Education
Research, 16, 21-31.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
198
Gambone, G. C. & Contrada, R. J. (2002). Patterns of self- and other-
representation in trait hostility. Journal o f Social & Clinical Psychology, 21,
546-565.
Gartska, T. A., Schmitt, M. T., Branscombe, N. R., & Hummert, M.
L. (2004). How young and older adults differ in their responses to perceived age
discrimination. Psychology o f Aging, 19, 326-335.
Gentry, D. (1985). Relationship of anger-coping styles and blood
pressure among Black Americans. In M. A. Chesney & R. H. Rosenman
(Eds.). Anger and hostility in cardiovascular and behavioral disorder (pp. 139-
147). New York: McGraw-Hill.
Graham, S. (1998). Social motivation and perceived responsibility in
others: Attributions and behavior of African American boys labeled as
aggressive. In: J. Heckhausen & C. S. Dweck (Eds.). Motivation and self
regulation across the life-span (pp. 137-158). New York, NY: Cambridge
University Press.
Graham, S. & Hudley, C. (1994). Attributions of aggressive and
nonaggressive African-American male early adolescents: A study of construct
accessibility. Developmental Psychology, 30, 365-373.
Graham, S., Hudley, C., & Williams, E. (1992). Attributional and
emotional determinants of aggression among African-American and Latino
young adolescents. Developmental Psychology, 28, 731-740.
Gross, J. J. (1998). The emerging field of emotion regulation: An
integrative review. Review of General Psychology, 2, 271-299.
Guerra, N. G., Huesmann, L. R., & Zelli, A. (1993). Attributions for
social failure and adolescent aggression. Aggressive Behavior, 19, 421-434.
Guyll, M. & Madon, S. (2003). Trait hostility: The breadth and
specificity of schema effects. Personality and Individual Differences, 34, 681-
693.
Harburg, E., Erfurt, J. C., Hauenstein, L. S., Chape, C. et al. (1973).
Socio-ecological stress, suppressed hostility, skin color, and Black-White male
blood pressure. Psychosomatic Medicine, 35, 276-296.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
199
Harrell, J. P., Merritt, M. M., & Kalu, J. (1998). Racism, stress and
disease. In R. Jones (Ed.). African American mental health (pp. 247-280).
Hampton, VA: Cobb & Henry Publishers.
Harris, S. M. (2004). The effect of health value and ethnicity on the
relationship between hardiness and health behaviors. Journal o f Personality, 72,
379-411.
Harrell, S. P. (2000). A multidimensional conceptualization of racism-
related stress: Implications for the well-being of people of color. American
Journal o f Orthopsychiatry, 70, 42-57.
Harris, D. A. (2002). Profiles in injustice: Why racial profiling cannot
work. New York: The New Press.
Hemmings. A. (2002). Youth culture of hostility: Discourses of money,
respect, and difference. Qualitative Studies in Education, 15, 291-307.
Hillman, S. B., Wood, P. C., & Sawilowsky, S. S. (1992).
Externalization as a self-protective mechanism in a stigmatized group.
Psychological Reports, 70, 641-642.
Hogan, B. E. & Linden, W. (2004). Anger response styles and blood
pressure: At least don’t ruminate about it! Annals o f Behavioral Medicine, 27,
38-49.
Holman, T. B. & Jarvis, M. O. (2003). Hostile, avoiding, and validating
couple-conflict types: An investigation of Gottman’s couple-conflict types.
Personal Relationships, 10, 267-282.
Hong, S. & Giannakopoulos, E. (1994). Effects of age, sex, and
university status on life satisfaction. Psychological Reports, 74, 99-103.
Hong, Y., Morris, M. W., Chiu, C., & Benet-Martinez, V. (2000).
Multicultural minds: A dynamic constructivist approach to culture. American
Psychologist, 55, 709-720.
Hopp, J. & Herring, P. (1999). Promoting health among Black
American populations: An overview. In: R. M. Huff & M. V. Kline (Eds.),
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
200
Promoting health in multicultural populations: A handbook for practitioners (pp.
201-221). Thousand Oaks, CA: Sage Publications, Inc.
Hudd, S., Dumlao, J., Erdmann-Sager, D., Murray, D. et al. (2000).
Stress at college: Effects on health habits, heath status and self-esteem. College
Student Journal, 34, 217-227.
Hughes, D. (2003). Correlates of African American and Latino parents’
messages to children about ethnicity and race: A comparative study of racial
socialization. American Journal o f Community Psychology, 31, 15-33.
Hughes, J. W., Sherwood, A., Blumenthal, J. A., Suarez, E. C., &
Hinderliter, A. L. (2003). Hostility, social support, and adrenergic receptor
responsiveness among African-American and White men and women.
Psychosomatic Medicine, 65, 582-587.
Jackson, J. S., Brown, T. N., Williams, D. R., Torres, M., Sellers, S.,
& Brown, K. (1996). Racism and the physical and mental health status of
African Americans: A thirteen year national panel study. Ethnicity and Disease,
6, 132-147.
Jacobs, J. B. & Potter, K. (1998). Hate crimes: Criminal law and
identity politics. New York: Oxford University Press.
Jain, S., Dimsdale, J. E., Roesch, S. C., & Mills, P. J. (2004).
Ethnicity, social class and hostility: Effects on in vivo B-adrenergic receptor
responsiveness. Biological Psychology, 65, 89-100.
Johnson, E. H. (1990). The deadly emotions: The role o f anger,
hostility, and aggression in health and emotional well-being. New York, NY:
Praeger Publishers.
Johnson, R. L. (1998). The differences within: The relationships
between racial identity, sociodemographic factors, health beliefs, and health-
promoting behaviors in African-Americans. Dissertation Abstracts
International, 59 (3-B), 1047.
Kaiser, C. R. & Miller, C. T. (2001). Reacting to impending
discrimination: Compensation for prejudice and attributions to discrimination.
Personality and Social Psychology Bulletin, 27, 1357-1367.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
201
Katz, L. & Woodin, E. M. (2002). Hostility, hostile detachment, and
conflict engagement in marriages: Effects on child and family functioning. Child
Development, 73, 636-651.
Keltikangas-Jarvinen, L. & Heinonen, K. (2003). Childhood roots of
adult hostility: Family factors as predictors of cognitive and affective hostility.
Child Development, 74, 1751-1768.
Keltikangas-Jarvinen, L., Raikkonen, K., & Adlercreutz, H. (1997).
Response of the pituitary-adrenal axis in terms of type a behaviour, hostility and
vital exhaustion in healthy middle-aged men. Psychology and Health, 12, 533-
542.
Kim, I. J. & Zane, N. W. S. (2004). Ethnic and cultural variations in
anger regulation and attachment patterns among Korean American and European
American male batterers. Cultural Diversity & Ethnic Minority Psychology, 10,
151-168.
Kim, J., Bramlett, M. H., Wright, L. K., & Poon, L. W. (1998). Racial
differences in health status and health behaviors of older adults. Nursing
Research, 47, 243-250.
Kino, K. (2000). Japanese anger expression styles and their interpersonal
influence. Japanese Journal of Psychology, 70, 494-502.
Klonoff, E. A., Landrine, H., & Ullman, J. B. (1999). Racial
discrimination and psychiatric symptoms among Blacks. Cultural Diversity and
Ethnic Minority Psychology, 5, 329-339.
Knowles, E. D., Morris, M. W., Chiu, C. Y., & Hong, Y. Y. (2001).
Culture and the process of person perception: Evidence for automaticity among
East Asians in correcting for situational influences on behavior. Personality and
Social Psychology Bulletin, 27, 1344-1356.
Koeveses, Z. (2000). The concept of anger: Universal or culture
specific? Psychopathology, 33, 159-170.
Kohn, P. M., Lafreniere, K., & Gurevich, M. (1990). The inventory of
college students’ recent life experiences: A decontaminated hassles scale for a
special population. Journal of Behavioral Medicine, 13, 619-630.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
202
Korte, J. E. (2000). Psychosocial stress and its relationship to weight for
gestational age and gestational age at delivery. Dissertation Abstracts
International, 60 (12-B), 6052.
Rrieger, N. (1990). Racial and gender discrimination: Risk factors for
high blood pressure? Social Science Medicine, 30, 1273-1281.
Krieger, N. & Sidney, S. (1996). Racial discrimination and blood
pressure: The CARDIA study of young Black and White adults. American
Journal o f Public Health, 86, 1370-1378.
Kwate, N. O. A., Valdimarsdottir, H. B., Guevarra, J. S., & Bovbjerg,
D. H. (2003). Experiences of racist events are associated with negative health
consequences for African American women. Journal of the National Medical
Association, 95, 450-460.
Laforge, R. G., Rossi, J. S., Prochaska, J. O., Velicer, W. F. et al.
(1999). Stage of regular exercise and health-related quality of life. Preventive
Medicine: An International Journal Devoted to Practice & Theory, 28, 349-360.
Lafromboise, T., Coleman, H. L., & Gerton, J. (1993). Psychological
impact of biculturalism: Evidence and theory. Psychological Bulletin, 114, 395-
412.
Landrine, H., & Klonoff, E. A. (1996). The schedule of racist events: A
measure of racial discrimination and a study of its negative physical and mental
health consequences. Journal of Black Psychology, 22, 144-167.
Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal, and coping.
New York: Springer.
Lee, R. E. & Cubbin, C. (2002). Neighborhood context and youth
cardiovascular health behaviors. American Journal o f Public Health, 92, 428-
436.
Lee, D. J. & Markides, K. S. (1989). Irritability and physical symptoms
in a three-generation study of Mexican Americans. Stress Medicine, 5, 253-258.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
203
Lee, F., Hallahan, M., & Herzog, T. (1996). Explaining real-life
events: How culture and domain shape attributions. Personality and Social
Psychology Bulletin, 22, 732-741.
Lee, Y. T. & Seligman, M. E. P. (1997). Are Americans more
optimistic than the Chinese? Personality and Social Psychology Bulletin, 23, 32-
40.
Lees, N. B. & Tinsley, B. J. (1998). Patterns of parental socialization of
the preventive health behavior of young Mexican origin children. Journal of
Applied Developmental Psychology, 19, 503-525.
Lees, N. B. & Tinsley, B. J. (2000). Maternal socialization of children’s
preventive health behavior: The role of maternal affect and teaching strategies.
Merrill-Palmer Quarterly, 46, 632-652.
Levin, J. & McDevitt, J. (1993). Hate crimes: The rising tide of bigotry
and bloodshed. New York: Plenum Press.
Levy, N. E. (2003). Examining the relationship between health-related
behaviors and subjective well being among college students. Dissertation
Abstracts International, 63 (9-B), 4376.
Littleford, L. N. & Wright, M. O. (1998). Health consequences of
positive and negative interracial interactions at a predominantly Caucasian
public university. Journal of Gender, Culture, and Health, 3, 129-145.
Loo, C. M., Fairbank, J. A., Scurfield, R. M., Ruch, L. O., King, D.
W., Adams, L. J., & Chemtob, C. M. (2001). Measuring exposure to racism:
Development and validation of a race-related stressor scale (RRSS) for Asian
American Vietnam Veterans. Psychological Assessment, 13, 503-520.
Louw-Potgieter, J. (1989). Covert racism: An application of Essed’s
analysis in a South African context. Journal o f Language and Social
Psychology, 8, 307-319.
Lyubomirsky, S., Caldwell, N. D., & Nolen-Hoeksema, S. (1998).
Effects of ruminative and distracting responses to depressed mood on retrieval
of autobiographical memories. Journal of Personality and Social Psychology,
75, 166-177.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
204
Mahon, N. E., Yarcheski, A., & Yarcheski, T. J. (2000). Positive and
negative outcome of anger in early adolescents. Research in Nursing & Health,
23, 17-24.
Major, B. (1995). Academic performance, self-esteem, and race: The
role o f disidentification. Paper presented at the annual meeting of the American
Psychological Association, New York.
Major, B., Kaiser, C. R., & McCoy, S. K. (2003). It’s not my fault:
When and why attributions to prejudice protect self-esteem. Personality and
Social Psychology Bulletin, 29, 772-781.
Major, B., Spencer, S., Schmader, T., Wolfe, C., & Crocker, J.
(1998). Coping with negative stereotypes about intellectual performance: The
role of psychological disengagement. Personality and Social Psychology
Bulletin, 24, 34-50.
Malarkey, W. B., Kiecolt-Glaser, J. K., Pearl, D., & Glaser, R. (1994).
Hostile behavior during marital conflict alters pituitary and adrenal hormones.
Psychosomatic Medicine, 56, 41-51.
Matthews, B. A. & Norris, F. H. (2002). When is believing “seeing”?
Hostile attribution bias as a function of self-reported aggression. Journal of
Applied Social Psychology, 32, 1-32.
McNeilly, M. D., Robinson, E. L., Anderson, N. B., Pieper, C. F.,
Shah, A., Toth, P. S., Martin, P. et al. (1995). Effects of racist provocation
and social support on cardiovascular reactivity in African American women.
International Journal o f Behavioral Medicine, 2, 321-338.
Miller, C. T. & Kaiser, C. R. (2001). A theoretical perspective on
coping with stigma. Journal o f Social Issues, 57, 73-92.
Miller, C. T. & Major, B. (2000). Coping with stigma and prejudice. In
T. F. Heatherton, R. E. Kleck & M. R. Hebl (Eds.). The social psychology of
stigma, (pp. 243-272). New York, NY: The Guilford Press.
Miller, D. M. (2002). Reframing hostility. Qualitative Studies in
Education, 15, 309-315.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
205
Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M., & Hallet,
A. J. (1996). A meta-analytic review of research on hostility and physical
health. Psychological Bulletin, 119, 322-348.
Millner, D., Burt-Murray, A., & Miller, M. (2004). The angry Black
woman’ s guide to life. New York, NY: Plume.
Moghaddam, F. M., Taylor, D. M., Ditto, B., Jacobs, K., & Bianchi,
E. (2002). Psychological distress and perceived discrimination: A study of
women from India. International Journal o f Intercultural Relations, 26, 381-
390.
Mok, T. A. (1998). Getting the message: Media images and stereotypes
and their effect on Asian Americans. Cultural Diversity and Mental Health, 4,
185-202.
Nasby, W., Hayden, B., & De Paulo, B. M. (1979). Attributional bias
among aggressive boys to interpret unambiguous social stimuli as displays of
hostility. Journal o f Abnormal Psychology, 89, 459-468.
Neighbors, H. W., Braithwaite, R. L., & Thompson, E. (1995). Health
promotion and African-Americans: From personal empowerment to community
action. American Journal of Health Promotion, 9, 281-287.
Nelson, D. A. & Crick, N. R. (1999). Rose-colored glasses: Examining
the social information-processing of prosocial young adolescents. Journal of
Early Adolescence, 19, 17-38.
Nelson, K., Geiger, A. M., & Mangione, C. M. (2002). Effect of health
beliefs on delays in care for abnormal cervical cytology in a multiethnic
population. Journal of General Internal Medicine, 17, 709-716.
Ng, D. & Jeffery, R. W. (2003). Relationships between perceived stress
and health behaviors in a sample of working adults. Health Psychology, 22,
638-642.
Nolen-Hoeksema, S., Larson, J., & Grayson, C. (1999). Explaining the
gender difference in depressive symptoms. Journal o f Personality and Social
Psychology, 77, 1061-1072.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
206
Ogden, J. & Mitandabari, T. (1997). Examination stress and changes in
mood and health related behaviours. Psychology & Health, 12, 288-299.
Ohbuchi, K., Kumagai, T., & Atsumi, E. (2002). Motives of and
responses to anger in conflict situations: A cross-cultural analysis. Tohoku
Psycholgica Folia, 61, 11-21.
Operario, D. & Fiske, S. T. (2001). Ethnic identity moderates
perceptions of prejudice: Judgments of personal versus group discrimination an
subtle versus blatant bias. Personality and Social Psychology Bulletin, 27, 550-
561.
Pak, S. K., Olsen, L. K., & Mahoney, B. S. (1999). The relationships
of health behaviors to perceived stress, job satisfaction, and role modeling
among health professionals in South Korea. International Quarterly of
Community Health Education, 19, 65-76.
Park, Y. S. & Kim. U. (1998). Locus of control, attributional style, and
academic achievement: Comparative analysis of Korean, Korean-Chinese, and
Chinese students. Asian Journal of Social Psychology, 1, 191-208.
Pate, R. R., Heath, G. W., Dowda, M., & Trost, S. G. (1996).
Associations between physical activity and other health behaviors in a
representative sample of US adolescents. American Journal o f Public Health,
86, 1577-1581.
Perez-Rivera, K. (2002). Relation of racial identity attitudes to the
experience of and coping with anger in African-Americans and Caucasians.
Dissertation Abstracts International, 62 (9-B), 4231.
Peters, M. F. (1985). Racial socialization of young Black children. In:
H. P. McAdoo & J. L. McAdoo (Eds.), Black children: Social, educational,
and parental environments, (pp. 159-173). Newbury Park, CA: Sage.
Phinney, J. S., DuPont, S., Espinosa, C., Revill, C. et al. (1994).
Ethnic identity and American identification among ethnic minority youths. In:
A. Bouvy, J. R. Fons van de Vijver et al. (Eds.), Journeys into cross-cultural
psychology (pp. 167-183). Lisse, Netherlands: Swets & Zeitlinger Publishers.
Pinel, E. C. (1999). Stigma consciousness: The psychological legacy of
social stereotypes. Journal o f Personality and Social Psychology, 76, 114-128.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
207
Pope, M. K. & Smith, T. W. (1991). Cortisol excretion in high and low
cynically hostile men. Psychosomatic Medicine, 53, 386-392.
Price, J. M. & Glad, K. (2003). Hostile attributional tendencies in
maltreated children. Journal of Abnormal Child Psychology, 31, 329-343.
Ramirez, M. J., Fujihara, T., van Goozen, S., & Satisteban, C. (2001).
Anger proneness in Japanese and Spanish students. In: M. J. Ramirez & D. S.
Richardson (Eds.), Cross-cultural approaches to research on aggression and
reconciliation (pp. 87-97). Hauppauge, NY: Nova Science Publishers, Inc.
Ray, G. E. & Cohen, R. (1997). Children’s evaluations of provocation
between peers. Aggressive Behavior, 23, 417-431.
Reece, M. (2003). HIV-related mental health care: Factors influencing
dropout among low-income, HIV-positive individuals. AIDS Care, 15, 707-716.
Roger, D. & Najarian, B. (1998). The relationship between emotional
rumination and cortisol secretion under stress. Personality and Individual
Differences, 24, 531-538.
Ruggerio, K. M. (1999). The personal/group discrimination discrepancy:
Extending Allport’s analysis of targets. Journal o f Social Issues, 55, 519-536.
Ruggerio, K. M. & Taylor, D. M. (1995). Coping with discrimination:
How disadvantaged group members perceive the discrimination that confronts
them. Journal o f Personality and Social Psychology, 68, 826-838.
Ruggerio, K. M. & Taylor, D. M. (1997). Why minority group
members perceive or do not perceive the discrimination that confronts them:
The role of self-esteem and perceived control. Journal o f Personality and Social
Psychology, 72, 373-389.
Russell, J. A. (1989). Culture, scripts, and children’s understanding of
emotion. In: C. Saarni & P. L. Harris (Eds.), Children’ s understanding of
emotion. Cambridge studies in social and emotional development (pp. 293-318).
New York, NY: Cambridge University Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
208
Rusting, C. L. & Nolen-Hoeksema, S. (1998). Regulating responses to
anger: Effects of rumination and distraction on angry mood. Journal of
Personality and Social Psychology, 74, 790-803.
Saab, P. G., Llabre, M. M., Schneiderman, N., Hurwitz, B. et al.
(1997). Influence of ethnicity and gender on cardiovascular responses to active
coping and inhibitory-passive coping challenges. Psychosomatic Medicine, 59,
434-446.
Samples, F. L. (1998). Cognitions, behaviors, and psychological
symptomatology: Relationships and pathways among African American and
Latino children. Journal o f Negro Education, 66, 172-188.
Sanders Thompson, V. L. (1991). Perception of race and race relations
which affect African American identification. Journal o f Applied Social
Psychology, 21, 1502-1516.
Sanders Thompson, V. L. (1996). Perceived experiences of racism as
stressful life events. Community Mental Health Journal, 32, 223-233.
Schmitt, M. T., Branscombe, N. R., Kobrynowicz, D., & Owen, S.
(2002). Perceiving discrimination against one’s gender group has different
implications for well-being in women and men. Personality and Social
Psychology Bulletin, 28, 197-210.
Schmitt, M. T., Spears, R., & Branscombe, N. R. (2003). Constructing
a minority group identity out of shared rejection: The case of international
students. European Journal o f Social Psychology, 33, 1-12.
Scherwitz, L. W., Perkins, L. L., Chesney, M. A., & Hughes, G.
(1991). Cook Medley Hostility Scale and subsets: Relationships to demographic
and psychosocial characteristics in young adults in the CARDIA study.
Psychosomatic Medicine, 53, 36-49.
Sebej, F. (1989). Psychological risk factors of coronary heart diseases as
culture-related phenomena. Studia Psychologica, 31, 259-269.
Segerstrom, S. C. (2000). Personality and the immune system: Models,
methods, and mechanisms. Annals o f Behavioral Medicine, 22, 180-190.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
209
Shorey, H. S., Cowan, G., & Sullivan, M. P. (2002). Predicting
perceptions of discrimination among Hispanics and Anglos. Hispanic Journal of
Behavioral Sciences, 24, 3-22.
Siegler, I. C., Costa, P. T., Brummett, B. H., Helms, M. J. et al.
(2003). Patterns of change in hostility from college to midlife in the UNC
Alumni Heart Study predict high-risk status. Psychosomatic Medicine, 65, 738-
745.
Siegler, I. C., Peterson, B. L., Barefoot, J. C., & Williams, R. B.
(1992). Hostility during late adolescence predicts coronary risk factors at mid
life. American Journal of Epidemiology, 136, 146-154.
Smith, T. W. (1992). Hostility and health: Current status of a
psychosomatic hypothesis. Health Psychology, 11, 139-150.
Smith, T. W. (1994). Concepts and methods in the study of anger,
hostility and health. In A. W. Siegman & T. W. Smith (Eds.). Anger, hostility
and the heart (pp. 23-42). Hillsdale, NJ: Erlbaum.
Snyder, C. R., Crowson, Jr., J., Kent Houston, B., Kurylo, M., &
Poirier, J. (1997). Assessing hostile automatic thoughts: Development and
validation of the HAT scale. Cognitive Therapy and Research, 21, 477-492.
Steele, C. M. (1997). A threat in the air: How stereotypes shape
intellectual identity and performance. American Psychologist, 52, 613-629.
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the
intellectual test performance of African Americans. Journal o f Personality and
Social Psychology, 69, 797-811.
Steele, R. G., Elliot, V., & Phipps, S. (2003). Race and health status as
determinants of anger expression and adaptive style in children. Journal of
Social & Clinical Psychology, 22, 40-58.
Steinberg, M. S. & Dodge, K. A. (1983). Attributional bias in
aggressive adolescent boys and girls. Journal of Social and Clinical Psychology,
1, 312-321.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
210
Steptoe, A., Wardle, J., Pollard, T. M., & Canaan, L. (1996). Stress,
social support and health-related behavior: A study of smoking, alcohol
consumption and physical exercise. Journal o f Psychosomatic Research, 41,
171-180.
Suarez, E. C. & Harralson, T. L. (1999). Hostility-related differences in
the associations between stress-induced physiological reactivity and lipid
concentrations in young healthy women. International Journal o f Behavioral
Medicine, 6, 190-203.
Suarez, E. C., Kuhn, C. M., Schanberg, S. M., Williams, R. B., Jr., &
Zimmerman, E. A. (1998). Neuroendocrine, cardiovascular, and emotional
responses of hostile men: The role of interpersonal challenge. Psychosomatic
Medicine, 60, 78-88.
Sukhodolsky, D. G., Golub, A., & Cromwell, E. N. (2001).
Development and validation of the anger rumination scale. Personality and
Individual Differences, 31, 689-700.
Surwit, R. S., Williams, R. B., Siegler, I. C., Lane, J. D. et al. (2002).
Hostility, race, and glucose metabolism in nondiabetic individuals. Diabetes
Care, 25, 835-839.
Swim, J. K., Cohen, L. L., & Hyers, L. L. (1998). Experiencing
everyday prejudice and discrimination. In J. K. Swim & C. Stangor (Eds.).
Prejudice: The target’ s perspective (pp. 37-60). San Diego, CA: The Academic
Press.
Tanaka-Matsumi, J. (1995). Cross-cultural perspectives on anger. In: H.
Kassinove (Ed.), Anger disorders: Definition, diagnosis, and treatment. Series
in clinical and community psychology (pp. 81-90). Philadelphia, PA: Taylor &
Francis.
Thomas, K. S., Nelesen, R. A., & Dimsdale, J. E. (2004).
Relationships between hostility, anger expression, and blood pressure dipping in
an ethnically diverse sample. Psychosomatic Medicine, 66, 298-304.
Thomas, L. R., Fox, S. A., Leake, B. G., & Roetzheim, R. G. (1996).
The effects of health beliefs on screening mammography utilization among a
diverse sample of older women. Women & Health, 24, 77-94.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
211
Thompson, P. C., Anderson, L. P., & Bakeman, R. A. (2000). Effects
of racial socialization and racial identity on acculturative stress in African
American college students. Cultural Diversity & Ethnic Minority Psychology, 6,
196-210.
Tinsley, B. J. (1992). Multiple influences on the acquisition and
socialization of children’s health attitudes and behavior: An integrative review.
Child Development, 63, 1043-1069.
Topalli, V. & O’Neal, E. C. (2003). Retaliatory motivation enhances
attributions of hostility when people process ambiguous social stimuli.
Aggressive Behavior, 29, 155-172.
Turkat, I. D., Keane, S. P., & Thompson-Pope, S. K. (1990). Social
processing errors among paranoid personalities. Journal o f Psychopathology and
Behavioral Assessment, 12, 263-269.
Utsey, S. O., & Ponterotto, J. G. (1996). Development and validation of
the index of race-related stress (IRRS). Journal o f Counseling Psychology, 43,
1 - 12.
Vaez, M. & Laflamme, L. (2003). Health behaviors, self-rated health,
and quality of life: A study among first-year Swedish university students.
Journal o f American College Health, 51, 156-162.
Vontress, C. E. & Epp, L. R. (1997). Historical hostility in the African
American client: Implications for counseling. Journal o f Multicultural
Counseling & Development, 25, 170-184.
Waas, G. A. (1988). Social attributional biases of peer-rejected and
aggressive children. Child Development, 59, 969-975.
Wardle, J. & Gibson, E. L. (2002). Impact of stress on diet: Processes
and implications. In: S. A. Stansfield & M. G. Marmot (Eds.), Stress and the
heart: Psychosocial pathways to coronary heart disease (pp. 124-149).
Weiner, B. (1986). An attributional theory o f motivation and emotion.
New York: Springer-Verlag.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
212
Williams, D. R., Spencer, M. S., & Jackson, J. S. (1999). Race, stress
and physical health: The role of group identity. In R. J. Contrada & R. D.
Ashmore (Eds.). Self, social identity, and physical health (pp. 71-100). New
York: Oxford Press.
Williams, L. (2000). It’ s the little things: Everyday interactions that
anger, annoy, and divide the races. New York, NY: Harcourt, Inc.
Wise, E. A., Carmichael, S. K., Belar, C. D., Jordan, C. B., &
Berlant, N. E. (2001). Women’s health: A cultural perspective. In: S. S.
Kazarian & D. R. Evans (Eds.), Handbook o f cultural health psychology (pp.
445-473). San Diego, CA: Academic Press.
Wong, C. A., Eccles, J. S., & Sameroff, A. (2003). The influence of
ethnic discrimination and ethnic identification on African American adolescents’
school and socioemotional adjustment. Journal o f Personality, 71, 1197-1232.
Woodruff, S. I. & Conway, T. L. (1992). Impact of health and fitness-
related behavior on quality of life. Social Indicators Research, 26, 391-405.
Wright, M. O. & Littleford, L. N. (2002). Experiences and beliefs as
predictors of ethnic identity and intergroup relations. Journal o f multicultural
counseling and development, 30, 2-20.
Yang, L. H. (2003). Causal attributions, expressed emotion, and patient
relapse: Recent findings and application to Chinese societies. Hong Kong
Journal o f Psychiatry, 13, 16-25.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
213
APPENDIX
Imagine that you are celebrating a special occasion with some friends at a restaurant you
have never eaten in before. The waitress serving you has a frown on her face. She is abrupt
with you when you order and ends up bringing you the wrong meal.
How LIKELY would your FIRST thought in reaction to this situation be something
like?:
Not at all Somewhat Almost
Likely Likely Definitely
1 2 3 4 5 6 7
1) “This just cracks me up!” ___________________ (Humor/Sarcasm Appraisal)
2) “What a low life idiot!” ___________________ (Hostile-Verbal Appraisal)
3) “This kind of thing happens.” ___________________ (Acceptance Appraisal)
H ow LIKELY w ould you have each o f the follow ing thoughts in reaction to this
situation?:
Not at all Somewhat Almost
Likely Likely Definitely
1 2 3 4 5 6 7
4) T his situation happened because o f m y race o r e th n ic ity .______________ (Racism
Attributions)
5) T his situation happened because this w aitress w as having a bad day. __________
(Situational/Constructive Attributions)
H ow LIKELY w ould you have each o f the follow ing thoughts after this event has
happened?:
Not at all Somewhat Almost
Likely Likely Definitely
1 2 3 4 5 6 7
6) T his w aitress really d id n ’t w ant to serve m e . ________________ (Devaluation by
Other)
7) By h er actions, this w aitress w as trying to m ake m e not w ant to eat h ere again.
____________(Hostile-Aggressive Motives)
8) If I decide to eat in this restaurant again, I w ill be treated the sam e w ay.
_____________(Generalized Negative Expectations)
9) “I w ant to sm ash som ething! ” _________________(Hostile-Physical Aggression
Thoughts)
10) “ I really w ant to get back at h er for th is!” ___________________(Hostile-Revenge
Thoughts)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
214
Not at all Somewhat
Extremely
1 2 3 4 5 6 7
11) How angry would you feel because of what happened?
Reactivity)
(Anger
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
215
HBS
I. Please use the following scale to rate how much you AGREE with each item:
1 = Not at all
2 = Slightly
3 = Somewhat
4 = Moderately
5 = Strongly
1. It is important for me to feel physically fit.__________
2. I find that it is important for my health to read nutrition labels.___________
3. It is important for me to be aware of my daily caloric intake._________
4. I find exercise enjoyable.____________
5. It is important for my health that I get 8 hours of sleep each night.__________
6. It is important for my health that I eat three balanced meals a d a y .__________
7. I feel bad if I don’t exercise.___________
8. It is important for my health that my diet not be primarily composed of food with high
salt and high fat content._________
9. I think it is important to exercise.__________
10. I have mainly had positive experiences with doctors and other
medical personnel.__________
11. Exercise is important for my health._________ __
12. It is important for me to go to the doctor regularly and have regular
dental check-ups._________
13. I tend to trust the health information that I learn about in the
m edia.__________
14. It is important for my health that I eat fresh fruits and vegetables each d ay._________
15. It is important for my health that I drink several ounces of water each d ay.__________
16. I have access to appropriate medical care and illness/disease prevention information
when needed.__________
17. I tend to trust the recommendations my doctor suggests to me when I am ill.
18. I have easy access to foods that m ake up a healthy, w ell-balanced d i e t .____________
19. It is im portant for m y health that I use sunscreen d a il y .___________
20. W hen I am ill it is m ore im portant for m e to self-m edicate than to go to the doctor.
21. It is im portant fo r m y health that I not consum e large am ounts o f caffeine each day.
22. W hen I am not feeling w ell it is im portant fo r m e try to gather as m uch inform ation
about w hat ails m e as p o s s ib le .____________
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
216
1 = Not at all
2 = Slightly
3 = Somewhat
4 = Moderately
5 = Strongly
23. It is important for me to keep well-informed about health and fitness on a regular basis.
24. It is important for me to keep well-informed about diet and nutrition on a regular basis.
25. It is important for my health that me or my partner use protection (e.g. condoms) if I
engage in sexual intercourse.__________
II. Please use the following scale to rate the FREQUENCY with which you have done
or experienced each of the following within THE LAST MONTH:
1 — Never
2 = Rarely (once per month)
3 = Sometimes (1-3 times per week)
4 = Often (4-6 times per week)
5 = Daily
1. Taken non-prescription pain killers__________
2. Used an illegal drug__________
3. Suffered from indigestion__________
4. Taken pills to help you sleep__________
5. Were ill from drinking too much__________
6. Had alcohol with lunch___________
7. Felt overfull after eating___________
8. Smoked cigarettes___________
9. Consumed foods with high salt content__________
10. Consumed foods with high fat content__________
11. Consumed caffeine__________
12. Taken prescription medication for pain or illness.____________
13. Experienced a headache___________
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
217
III. Please use the following scale to rate the FREQUENCY with which
you have done or experienced each of the following within THE LAST MONTH:
1 = Never
2 = Rarely (once per month)
3 = Sometimes (1-3 times per week)
4 = Often (4-6 times per week)
5 = Daily
1. Played a sport
2. Participated in aerobic, cardiovascular exerc
3. Drank 64 ounces of water
4. Had fresh vegetables and salads
5. Had fresh fruit or fruit juice
6. Gotten complimented on how well you look
7. Ate three balanced meals a day
8. Gotten 8 hours of sleep per night
9. Taken vitamin supplements
10. Read nutrition labels
11. Were conscious of your daily, caloric intake
12. Used sunscreen
13. Read a health or fitness magazine
14. Read a diet or nutrition magazine
15. Practiced safe sex (e.g. used a condom)
16. Practiced yoga, pilates or meditation
IV. Please use the following scale to respond to these items:
1 = Yes
2 = No
1. Do you go to the doctor at the first sign of illness?__________
2. Do you have yearly medical and dental check-ups?__________
3. Do you subscribe to or read health-related or fitness-related books or magazines?
4. Do you subscribe to or read diet and nutrition-related books or magazines?
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
218
V. Please use the following scale to rate these items:
Poor
1 2 3 4
Excellent
5
1. My general health_________
2. My physical fitness_________
3. My attitude toward health_____
4. My knowledge of health matters
5. My eating habits__________
6. My general lifestyle_________
VI. What do you perceive your chances are of developing the following illnesses within
your lifetime? (Please use the following scale):
1 = very low (0-10%)
2 = low (11-30%)
3 = moderate (31-50%)
4 = high (51-75%)
5 = very high (76-100%)
1. cancer____________
2. HIV/AIDS___________
3. STD (sexually-transmitted disease)
4. multiple sclerosis___________
5. Alzheimer’s disease__________
6. Parkinson’s disease__________
7. diabetes___________
8. epilepsy____________
9. heart disease__________
10. stroke____________
11. arthritis__________
12. chronic pain_____________
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Emotion regulation as a mediator between family-of-origin aggression and marital aggression

PDF

The relationship of hardiness and racial identity attitudes to the articulated thoughts and psychological impact of simulated racism encounters

PDF

An evaluation of a brief HIV /AIDS prevention intervention using normative feedback to promote risk reduction among sexually active college students

PDF

Anger expression in men with elevated blood pressure: Relationships to assertiveness

PDF

Hostile family environments and children's perceptions of social support

PDF

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

PDF

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

PDF

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

PDF

Content analysis of articulated thoughts of chronic worriers

PDF

Affection and conflict in family relationships

PDF

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

PDF

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

PDF

Assimilation and ethnicity: Adaptation patterns and ethnic identity of Armenian -Americans in central California

PDF

Associations and mechanisms among attention deficit hyperactivity disorder symptoms, cognitive functioning, and drinking habits

PDF

Gender differences in symptom presentation of depression in primary care settings

PDF

Ethnic identity, acculturation, self-esteem and perceived discrimination: A comparison study of Asian American adolescents

PDF

Attachment style, interpersonal guilt, parental alcoholism, parental divorce and eating disordered symptomatology in college women

PDF

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

PDF

Depression and suicidality in Latino adolescents: A study of acculturation and gender role beliefs

PDF

A typology of maritally violent men and correlates of violence

Asset Metadata

Creator Best, Jennifer Lauren (author)

Core Title An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

School Graduate School

Degree Doctor of Philosophy

Degree Program Psychology

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

Tag OAI-PMH Harvest,Psychology, clinical,psychology, social,sociology, ethnic and racial studies

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Davison, Gerald C. (committee chair), Daley, Shannon (committee member), Goodyear, Rodney (committee member), John, Richard S. (committee member), Meyerowitz, Beth E. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-394599

Unique identifier UC11340961

Identifier 3196777.pdf (filename),usctheses-c16-394599 (legacy record id)

Legacy Identifier 3196777.pdf

Dmrecord 394599

Document Type Dissertation

Rights Best, Jennifer Lauren

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