i9-

\k;
‘95:“
1 n‘ O. ‘

v

1

NM;-
d’.‘ ‘

 

.HI, IV 1..
7?

5—.

 

J

r
lull

7
Q.

ler

 

a? E. .3. ‘53.“.3

.7 . x‘ K . _ ‘ '1
MNJ‘A-qu-wa-

a

..:~.fih&.
. gram»
3 giant...
LN r. é :.

A2!
=

I. V
JM‘.‘

..l

..
'1

I

Q!“

a. : .1
.\.h7<i
.r

H

115’}: .

 

l,.~;l 5‘"
‘Vl‘l'u
IN“:

‘ \
E

“m .‘
Jiwx

‘ .n. I¢t|hb|vh .!
5; v . A lily... ‘ :2.
I (I... 3.. l. . t
\. 3.1!“. (gut.

 

 

5.4 K... .51..¢- ‘1.

win.

   

v‘tll'r‘l-I
.Q. 1. .99.)...1... .- .

«6.- E...

 
 

,. - awn .. u
nu. :— bl‘h- ..a. I. ”ma-Zn : ‘l to '

tun.

     

I’ll"
i I

 

 

THESlS

33

.74

D ,

lllllllllllllfllHill!!!IllllllHlllllllllllllll

293 02058 2106

This is to certify that the
dissertation entitled

Values and Attitudes Toward Interracial Marriage:
An Examination of the Value Expressive Function

presented by

Victoria 0. Orrego

has been accepted towards fulfillment
of the requirements To‘r‘

PhJ) degree in Communication

 

42-61114! 1“ AQZZA @ L

Major professor

Date 11-24-99

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY

Michigan State
UnIvers'ty

 

 

 

PLACE IN REI‘URN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

DA'lE DUE DUE DATE DUE '

*8
§

0

7 2008
'0310 “8

 

moo muse.“

VALUES AND ATTITUDES TOWARD INTERRACIAL MARRIAGE:

AN EXAMINATION OF THE VALUE EXPRESSIVE FUNCTION

by

Victoria 0. Orrego

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Communication

1999

ABSTRACT

VALUES AND ATTITUDES TOWARD INTERRACIAL MARRIAGE:
AN EXAMINATION OF THE VALUE EXPRESSIVE FUNCTION

by

Victoria 0. Orrego

This study examined the role of the value expressive function in moderating the
relationship between values, attitudes toward interracial marriage and the extent to which values
are manifested in message production. Strongest associations between values, attitudes and use
of value appeals were predicted to occur among high value expressive individuals versus low
value expressives. Results did not support the predicted moderating efiect, rather post hoc
analyses suggest a joint additive effect for value expressiveness and values on attitudes. The
viability of a mediating role that attitudes may have on message production is proposed.
Additionally, the constructs of other-directedness and extroversion, as indirect measures of the

value expressive function were found to be irrelevant in the explication of this process.

ACKNOWLEDGMENTS

I would like to thank my advisor Dr. Sandi Smith for her continuous support and
friendship. Her mentoring was invaluable to my development as a researcher and teacher
in communication. Dr. Smith was always willing to listen and help me sort out all my
ideas. Her feedback made it possible for me to successfully complete this and many other
projects. Dr. Smith is an excellent role model for all her advisees, and we all have
benefited from knowing and working with her.

I would also like to thank Dr. Frank Boster, Dr. Ron Tamborini, and Dr. Stan
Kaplowitz for all of their insight and constructive comments on this project. Their
suggestions and contributions were invaluable to the completion of this dissertation.
Additionally, taking their classes extended my knowledge of and application of
communication research. I am a better critical thinker and evaluator because of them.

I would also like to thank the entire faculty, staff, and graduate students in the
Communication department. I had a great experience at Michigan State and that was
possible because of them. Specifically, I would like to thank Cynthia Zuckerman for the
hundreds of things she did to for me ranging from dog sitting to prelim. and dissertation
cheerleader. She is a very special person, and I am truly thankful to count her as one of
my closest friends. I also want to thank Maria Lapinski, my office-mate for her
fiiendship, sense of humor and most importantly for her mutual love of pizza. Through
the years we shared countless pizzas. I also want to thank Betty LaFrance, Monique

Mitchell, and Craig Hullett for their frienship; they made graduate school a fun place to

be. I could always count on them for feedback on research, advice on life and lots of
laughs.

Finally, I would like to thank the most important people in my life because
without them I never would have accomplished this goal. Jon Dunleavy is truly an
instrumental person in my life. He has unconditionally been there supporting, listening,
encouraging and loving me throughout this whole process. When I really was stressed
out or overwhelmed, Jon could always find a way to make me feel better. I look back at
all our wonderful memories and anticipate an even better future.

Last, but not least are my parents Conrad and Esther Orrego. They have been my
greatest supporters. I am indebted to them for everything they have given me. Even
though we were far apart I could always count on marathon phone chats with my mom
and festive emails from my dad. Their encouragement and confidence in me never

wavered. They are wonderful people who I dearly love.

iv

TABLE OF CONTENTS

List of Tables

List of Figures

CHAPTER 1

CHAPTER 2: Literature Review
CHAPTER 3: Method
CHAPTER 4: Results

CHAPTER 5: Discussion

Author’s Notes

Appendices

91.9.0 S”?

Consent Letter and Questionnaires

Items, Means, SD, and Factor Loadings of Self-Monitoring
Items

Inter-F actor Correlations of the 6 Value Scales

Descriptive Statistics and T-values for the Western and
Midwestern Samples

Means and Standard Deviations for the Regression of
Attitudes on the Benevolent Success for Low
Value Expressives and High Value Expressives
for the Overall Sample

Means and Standard Deviations for the Regression of
Attitudes on the Power-Tradition Scale for Low
Value Expressives and High Value Expressives
for the Overall Sample

Means and Standard Deviations for the Regression of

Attitudes on the Varied Life-Value for Low Value
Expressives and High Value Expressives
Correlations for the Equality Path Model

1%."

vii

viii

34

63

87

92

93

112
116

118

120

122

124
126

1"

@923

References

Correlations for the Benevolent Success Path Model for
the Midwest Sample
Correlations for the Benevolent Success Path Model for
the Western Sample
Correlations for the Power-Tradition Path Model for the
Midwestem Sample
Correlations for the Power-Tradition Path Model for the
Western Sample
Correlations for the Varied Life Model for the Midwest Sample
Correlations for the Varied Life Model for the Western Sample
Correlations for the Honoring Parents Model
Correlations for the Social Order Model

vi

128
130
132
134
136
138

140
142

144

LIST OF TABLES

Schwartz’ definitions of motivational types of values in terms of
the goals, the single values that represent them, and the universal
requirements that represent them

Means and standard deviations for the regression of attitudes on the
equality value for low value expressives and high value expressives

Means and standard deviations for the regression of attitudes on the
benevolent success scale for the low value expressives and high
value expressives

Means and standard deviations for the regression of attitudes on the
power—tradition scale for low value expressives and high value
expressives

Means and standard deviations for the regression of attitudes on the
varied life scale for low value expressives and high value expressives

Means and standard deviations for the regression of attitudes on the
honoring parents scale for low value expressives and high value
expressives

Means and standard deviations for the regression of attitudes on the
social order scale for low value expressives and high value expressives

Means and standard deviations for the regression of equality messages
and attitudes for low value expressives and high value expressives

Means and standard deviations for the regression of benevolent

success messages and attitudes for the low value expressives and
high value expressives

vii

13

52

54

56

57

58

59

60

61

LIST OF FIGURES

The protypical structure of value systems
The equality path model

The benevolent success path model for the midwestem and
western samples

The power-tradition path model for the midwestem and western
samples

The varied life multiple regression model for the midwestem and
western samples

The honoring parents multiple regression model
The social order multiple regression model

The process model

viii

1%“
ind
"‘ a»

75

77

79

80

80

81

85

CHAPTER 1

Despite societal changes towards equality and cultural diversity, many Americans
still find it difficult if not impossible to accept interracial dating and marriage practices
(Mills, Daly, Longrnore, & Kilbride, 1994; Paset, & Taylor, 1991; Simpson, & Yinger,
1985). The number of interracial marriages has been on the increase in the past two
decades (Davidson, 1992; Porterfield, 1978; Solsberry, 1994). This rise in interracial
marriage will subsequently make issues of prejudice and racial tolerance more salient for
individuals, especially for those who find interracial marriage unacceptable or wrong.
Previous research on prejudice has implicated both Value Theory and Functional Theory
as relevant to attitude change and to tolerance (Biemat, Vescio, Theno, & Crandall, 1996;
Herek, 1984; Kristiansen, 1989; Snyder & Miene, 1994; Wyman & Snyder, 1997).
Biemat et al. (1996) found that the values of individualism, self-reliance, work ethic, T I
obedience, and discipline were positively associated with racism towards blacks; while
the values of equality, social justice, and concern of other’s well being were negatively
correlated with racism towards blacks. Additionally, Rokeach (1973) and Schwartz
(1992) argue that attitudes are guided by tradeoffs made among competing values that are
triggered by the attitude target or domain.

Functional Theory (Herek, 1986) can clarify the relationship between values and
attitudes toward the domain of interracial marriage. Numerous Functional Theories have
been posited, and all Functional Theories argue that people maintain their attitudes
because they serve important psychological needs that allow individuals to execute plans
and achieve goals. (Herek, 1986, 1987; Katz, 1960; Shavitt, 1990; Smith Bruner, and

White, 1956). Recently Herek (1986) identified six specific functions, experiential-

schematic, experiential-specific, ego-defensive, social-expressive, anticipatory, and
value-expressive. Herek (1984; 1986) argues that the value-expressive function is
relevant for issues regarding racial intolerance or changes in the racial status quo. The
issue of interracial marriage fits this criteria. The value expressive function delineates
attitudes whose benefit is derived from the expression of general value priorities which '
are consistent with a larger ideology supported by certain referent groups (e. g., racial .-
identity, religious groups or political groups).

If an individual’s attitudes function to express their values as a vehicle for self
identification, then it is possible to examine specific values that are linked with the
domain of interracial marriage. Furthermore, given its emphasis on expression, the value
expressive function can provide insight on the extent to which this function impacts
individuals’ use of value appeals in their message production regarding the issue of
interracial marriage.

Consequently, this study examines the relationship among values, attitude
functions, attitudes toward interracial marriage, and subsequent message production.
The focus of this study will be on individuals whose predominant function is value-
expressive. First, attitude, functions concerning attitudes toward interracial marriage are
identified through using a thought listing measure adapted from Herek (1986).
Additionally, indirect measures of the value expressive function via the other-
directedness and extroversion scales are tested. Second, the relationship between values
and attitudes is hypothesized to be moderated by attitude functions such that individuals
categorized as high value expressives will yield a stronger correlation between their

relevant values and attitudes than will individuals who are low value expressives.

Finally, participants’ persuasive messages, arguing for their opinion about interracial
marriage, will be examined for the extent to which they contain value appeals congruent
with their value relevance ratings. The following literature review will include research
on attitudes toward interracial marriage, value theory, functional theory, and action

assembly theory.

CHAPTER 2

LITERATURE REVIEW

Attitudes Toward Interracial Marriage

Current research provides insight into societal perceptions of interracial
marriages. Historically, interracial romantic relationships have existed since colonial
times but have been perceived as a cultural taboo (Porterfield, 197 8). This was illustrated
by social sanctions placed on marriages and dating among the races making them illegal
and morally wrong. It was not until 1967 that the United States govermnent finally
outlawed any legislation prohibiting racially mixed marriages (Davidson, 1992).

Social attitudes toward interracial marriage have been affected by the increasing
racial integration that has occurred in the workplace, educational institutions, and social
environments (Kouri, & Lasswell, 1993; Solsberry, 1994). The latest census figures
report that during this decade interracial marriage has been on the rise across all minority
groups, with the sharpest (10%) increase being between young black menand white
women (Peterson, 1997). A total of 330,000 black-white interracial marriages exist in the
United States (US Bureau of the Census, 1998). Yet, despite elimination of
discriminatory barriers, many interracial marriages are still not positively evaluated by
many whites or blacks in the United States (Paset & Taylor, 1991; Porterfield, 1978;
Simpson, & Yinger, 1985).

Although interracial marriages occur with all racial combinations, most of the
extant research focuses on the black/white combination, with the majority featuring black

male and white female couples. Historically, the black-white racial combination is

perceived most negatively (Porterfield, 197 8) and occurs more infrequently when
compared to other racial combinations (US Bureau of the Census, 1998). Additionally,
most theories of prejudice that help explain attitudes toward intergroup relationships such
as interracial marriage focus on attitudes toward blacks. This paper will limit its scope by
maintaining a focus on the black/white racial combination in order to extend previous
theoretical work.

In his historical review of black-white marriages, Porterfield (1984) found that
some overt hostility toward racially mixed couples existed moth black andmgltita' f
communities in-the form of housing dlscnnunanda and bad treatment by service
employees (restaurants, shOps, and other public areas). However, most of the evidence
on attitudes of interracial marriage comes directly from reports of famial and friend
support received by the married couples. Porterfield (1984) reported that the black
husband: 5 family tended to accept or just tolerate the marriage between the interracial
couple sooner-than the wife’s white relatives and friends. Porterfield (1978) found that
twenty-six out of forty white families rejected the interracial black-white marriages of the
subjects even after the couples had been married between 1 to 4 years. In contrast, black
families were more receptive to their children’s interracial marriages and most of those
who initially opposed the marriage gradually accepted it.

In a more recent study, of twenty-three black male/white female and six black
female/white male married couples, Kouri and Laswell (1993) found that seventeen of
the black men reported that their families immediately accepted the marriage while two
eventually accepted and one had not accepted. The white women respondents reported

that six of their families accepted the marriage immediately while eleven gradually

accepted it and six had refirsed to accept it. For the black female/white male couples,
black female respondents reported that three of their families accepted immediately and
two accepted later. The white male participants reported that four of their families
accepted immediately while one accepted later.

Evidence of attitudes toward interracial marriage from individuals who are not
family members of the interracial couple is somewhat dated. Bontemps (1975) reported
that 51% of Americans would “be willing to accept and live with” an interracial marriage
involving members of their own families. However; 76% Psald that they would not be “in

-wfavor’t’cf such- a marriage. The wording of the items resulted in differential responses.
More recent evidence (Paset and Taylor, 1991) cites that the level of acceptance varies
across racial lines, such that white female college students have more favorable attitudes

.than black female college students. Black female unfavorability parallels reports that
black females are the least likely to marry out of their race (Kouri & Laswell, 1993;
Tucker & Mitchell-Keman, 1990). Davidson and Schneider (1992), however, found that
black respondents reported being more willing to accept black-white interracial marriage
than white respondents, regardless of respondent sex.

Attitudes toward interracial marriage have been shown to vary depending on the
racial makeup, such that the racial combination of black-white has been documented to
provoke the strongest societal response in terms of discrimination (Davidson, 1992; Paset
& Taylor, 1991; Porterfield, 1978; Simpson & Yinger, 1985; Solsberry, 1994). For
example, Simpson and Yinger (1985) found that as Hispanics attain middle-class status,
ethnicity tends to become blurred. Specifically-this change in socioeconomic status

brings abdut- greater involvement with the dominant cultural group, which results in a

decreased use of the Spanish language and less contact with one’s extended family.
Ultimately, Hispanics are perceived as more similar and less threatening partners in
interracial relationships than Afiican-Americans.

Simpson and Yinger (1985) also reported that marriages between whites and
persons of ethnic backgrounds other than African-American are more frequent than
marriages between whites and African-Americans. Similarly, Tucker and Mitchell-
Keman (1990) reported that interracial marriages among Japanese women and Native
American women is now practically normative (40.6% and 53.7%) whereas such
behavior is still rare for black men and women (12.1%).

F arnilial relationships within interracial families are severely affected by the
prevailing attitudes and beliefs such that racist views negatively impact both. the family’s
Messner (1986) have pointed out that racial Inequalityls a strong predictor of the success
of interracial marriages, and that the breakdownof prejudice and discrimination via 1
education may lead to greater success of (these marriages.

Overall, research on attitudes toward black-white marriages is marked by
opposition and negative affect. Opposition within the white community stems from a
belief in a hierarchyiof races that views intermarriage as a disgrace to white families, a
betrayal of racralpurlty and of Protestarttvrorlt ethic values (Davidson & Schneider,
1992; Kinder & Sears, 1981). Opposition within the black community stems from
betrayal of the black identity and fear of negative repercussions (physical or verbal) from
the white majority (Davidson & Schneider, 1992). A critical element in this opposition is

the differential values that drive both blacks’ and whites’ attitudes toward interracial

marriage. An examination of how values impact attitudes toward interracial black-white

marriage can explain the nature of these attitudes.

\_/_al_ue;_and Valge Theory

Many theories have been posited to explain anti-black affect and prejudice. In
particular, Biemat et a1. (1996) note the role of values as contributors of both white
antipathy and sympathy toward blacks. Specifically, they refer to Symbolic Racism
(Kinder, 1986; Kinder & Sears, 1981; McConahay and Hough, 1976) and the Ambivalent
Racism perspective (Katz & Hass, 1988). McConahay and Hough (1976) defined
sganbolieracism as “the expression in terms of abstract ideological symbols and symbolic
i behaviors ofthe feeling that blacks are violating cherishedlvalues and making illegitimate
demands for changesin the status quo’? (p.38). Symbolic racism is rooted in traditional
religious and value socialization of “secular American civil Protestantism (values of hard
work, discipline, obedience and individualism), socialization to political conservatism
and unacknowledged negative feelings toward blacks. (McConahy & Hough, 1976).

The Racial Ambivalence perspective suggests that white Americans’ response to
blacks is driven by “ambivalence”. According to Katz and Hass (1988) ambivalence .
stemsfiiom the simultaneous holding of both anti-r and pro-black attitudes. This duality
of attitudes is a result 'of conflict between two core American values: humanitarianism/
egalgamgméndthe Protestant work ethic (individualism) respectively held by white
Americans. Both/Symbolic Racism and Racial Ambivalence explain anti- and pro-black
attitudes and as such identify certain values that can be applied to the domain of

interracial marriage, however, they do not posit how values function to impact attitudes.

The importance of values for facilitating attitude formation necessitates a discussion of
value theories by Rokeach (1973) and Schwartz (1992).

Rokeach (1973) defined values as specific types of beliefs that are central in the
system and act as life guides. Rokeach (1973) differentiated between instrumental values
and terrmngl values. Instrumental values are guidelines for living on which daily behavior
is assessed. In contrast, terminal values are the ultimate aims of life. Rokeach (1973)
argued that values have a motivational function in that values guide people’s actions and
attitudes. According to Rokeach (1973) values are determinants of attitudes whereby
favorable attitudes emerge as a function of whether the attitude is congruent with
fimdamental values. For certain individuals (pro interracial marriage). the attitude
domain of interracial marriage may trigger terminal values of equality, freedom and
mature love, while those who oppose interracial marriage may place significantly less
relevance on these values.

Influenced by Rokeach, Schwartz and Bilsky (1987) propose a theory of the
universal content and structure of human values. Schwartz (1992) presents a revised
theory thatdefines values as “desirable, transsituational goals, varying in importance, that
serve as guiding principles in. people’s lives? (p. 4). Additionally, the primary content
aspect of a value is the type of goal or motivational concern it expresses. Schwartz
(1992) derived a typology of ten value types by reasoning that values represent, in the
form of conscious goals, three universal requirements: biological needs, requisites of
coordinated social interaction, and demands of groups survival and functioning (Table 1
contains definitions and a list of value types). “Groups and individuals represent these

requirements cognitively as specific values about which they communicate in order to

explain, coordinate, and rationalize behavior.” (Schwartz, 1996, p. 2) For example, the
conformity value type was derived from both the prerequisite of smooth interaction and
of group survival--that individuals restrain impulses and inhibit actions that might hurt
others.
Additionally, the theory specifies relations among the types of values. Schwartz
(1992) posits that actions taken in pursuit of each type have psychological and social
consequences that may conflict with or may be compatible with the pursuit of other value
types. The total pattern of relations of value conflict and compatibility among value
priorities gives rise to a circular structure of value systems (see Figure l). Competing
value types emanate in opposing directions from the center; complementary types are in
close proximity going around the circle.
Finally, two major value conflicts that structure value systems were found in over

95% of cross cultural studies across 41 countries surveyed. The results produced two-
dimensional representations of values that revealed four higher order domains: self-
enhancement/ self transcendence and openness to change/conservation (see Figure 1). One
dimension opposes Openness to Change (combining the self-direction and stimulation
value types) with Conservation (combining security, conformity, and tradition). This
dimension reflects a conflict between emphases on own independent thought and action
and favoring change versus submissive self—restriction, preservation of traditional
practices, and protection of stability. The second dimension opposes Self-Transcendence
(combining benevolence and universalism) with Self-Enhancement (combining power
and achievement). This dimension reflects a conflict between acceptance of others as

equals and concern for their welfare versus pursuit of one’s own relative success and

10

    
 

s'v Self-Direction

   

Universalism

Benevolence

    

 

Conformity

  
   
 

Tradition

 

Security

  

Figu_r_e 1. The Protypical structure of value systems. This figure illustrates the relations
between Schwartz’s (1992) value domains. Adjacent domains are positively related.
Opposing domains are negatively related. Higher order domains are shown on the
periphery. Examples of the values within each domain are described in Table l. The
dashed lines around the hedonism domain indicates that it shares elements of the self-
enhancement and the openness higher order value domains.

Note. From Schwartz, 8. (1996). Value priorities and behavior: Applying a theory of
integrated value systems. In C. Seligman, J. M. Olson, & M. P. Zanna (Eds), IE

psychology of values: The Ontario smmsium, volume 8 (pp.l-24). Mahwah: NJ:
Erlbaum. Copyright 1996 by the Name of Copyright Holder. Reprinted with permission.

dominance over others. Hedonism shares elements of both Openness and Self-
Enhancement. Table 1 is presented here.

This pattern of relations among values allows one to predict a pattern of relations
between values and other variables like attitudes. Sagiv and Schwartz (1995) utilized this
theory to explain how value priorities of participants (Israeli Jews) helped to explain their
readiness for social contact with members of an outgroup (Israeli Arabs). Results
demonstrated that the Conservation values (tradition, conformity, and security) correlated
negatively—with readiness for out-group contact while Openness to Change (self-direction
and stimulation) and Self-Transcendence values (benevolence and universalism)
correlated positively with readiness for social contact. The Self-Enhancement dimension
yielded near-zero correlations.

A similar set of predictions can be made for attitudes toward black-white
interracial marriage. For example, a person may be favorable toward interracial marriage
because they consider equality and broadnrindedness (self-transcendent values) to be
relevant and important. Thus, self-transcendent values may be positively related to
favorable attitudes toward interracial marriage, while negatively or not at all related to
unfavorable attitudes, which would reflect self-enhancement values. The same
relationship can be posited for openness to change, being positively related to favorable
attitudes toward interracial marriage while being negatively related to unfavorable
attitudes, which reflect a more conservative stance.

In sum, prior research on Symbolic Racism and value theory (Schwartz, 1992) on

intergroup relations has suggested that symbolic attitudes fulfill certain functions in that

they ‘eXpress values tied to a person’s group membership or social identity (Herek, 1986;

12

Bible 1. Schwaltz’s definitions of motivational ms of values in terms of the goals, the single

glues that represent them, and the universal requirements that represent them

Power:

Achievement:

Hedonism:

Stimulation

Self-Direction

Universalism

Benevolence

Tradition

Conformity

Security

Social status and prestige, control or dominance over people and resources.
(social power, authority, wealth). Universal requirement is individual/
organismic need.

Personal success through demonstrating competence according to social
standards. successful, capable, ambitious, influential). Universal requirements
are individual/organismic and need for coordinated social interaction.

Pleasure and sensuous gratification for oneself. (pleasure, enjoying life).
Universal requirement is individual/organismic need.

Excitement, novelty, and challenge in life. (daring, a varied life, an exciting life).
Universal requirement is individual/organismic need.

Independent thought and action-choosing, creating, exploring. (creativity,
freedom, independent, curious, choosing own goals). Universal needs are
individual/organismic and need for coordinated social interaction.

Understanding, appreciation, tolerance, and protection for the welfare of all
people and for nature. (broadminded, wisdom, social justice, equality, world at
peace, world of beauty, unity with nature, protecting the environment).
Universal requirements are need for coordinated social interaction and survival
and welfare of groups.

Preservation and enhancement of the welfare of people with whom one is in
frequent personal contact. (helpfirl, honest, forgiving, loyal, responsible).
Universal requirements are need for coordinated social interaction and survival
and welfare of groups.

Respect, commitment and acceptance of the customs and ideas that traditional
culture or religion provide the self. (humble, accepting my portion in life,
devout, respect for tradition, moderate). Universal requirement is survival and
welfare of groups.

Restraint and actions, inclinations, and impulses likely to upset or harm others
and violate social expectations or norms. (politeness, obedient, self-discipline,
honoring parents and elders). Universal requirements are coordinated social
interaction and survival and welfare of groups.

Safety, harmony, and stability of society of relationships and of self. (family,

security, national security, social order, clean, reciprocation of favors). Universal
requirements are individual/organismic needs and survival and welfare of groups.

13

Kristiansen, 1989; Shavitt, 1989). Interracial marriage is an issue that can evoke
symbolic attitudes that are based on certain important values. Functional theory, then,
can be used to clarify the relations between values, attitudes, and ultimately behavior
regarding interracial marriage. Values have the ability to influence attitudes, especially

when people form attitudes specifically aimed at expressing those values.

Functional Theory

Historical development of fimctional theory. The functional approach to attitude
assessment was first conceptualized by two independent researchers: M. Brewster Snrith
(1959) and Daniel Katz (1960). Both researchers developed the idea that attitudes serve
various psychological needs and thus have variable motivational bases (Eagly & Chaiken,
1993). Attitudes, then, allow individuals to execute certain plans and achieve certain
goals (DeBono, 1987). Central to both researchers was a taxonomy of specific
personality functions that attitudes can serve for an individual. Smith and his colleagues
inductively derived their 3-function taxonomy based on detailed clinical interviews that
probed attitudes toward Russia In comparison, Katz derived his 4—function taxonomy
deductively from his own previous research on ego-defense as a source of prejudice. For
the purpose of this paper Katz’s theory will be briefly reviewed.

Katz’s (1960) theory posits four attitude functions: instrumental/utilitarian, ego-
defensive, value-expressive and knowledge. The instrumental function reinforces the /i,‘
behavioristic perspective that people are motivated to gain rewards and avoid

punishments. The ego-defensive function is used when an individual wishes to protect

hersfiimsclflfiomanknoflnlginathehesiaw about her 0,1“ himself, 01’ Eh? harsh

l4

realities in her or his external world.) The value expressive function is used when an
individual derives satisfaction from attitudes expressing beliefs appropriate to his or her
personal—valuesandto tlre'self-concept. The knowledge function is based upon the need
to give adequate structure to the universe. Katz (1960) assumed that personality
measures could serve as indirect indicators of attitude functions.

The fundamental hypothesis of the functional approach asserts .that changing an
attitude requires understanding its motivational basis. Thus, the functional theories have
important implications for persuasion research in that eliciting attitude change requires
identifying the function that an attitude serves and targeting persuasive appeals to the
function(s) of the attitude (Shavitt, 1989). In the context of the value expressive function,
a message informing an individual that a particular attitude is not expressing an important
value should trigger attitude change only if the motivational basis of the attitude is
predominantly value expressive (DeBono, 1987). For example, a white individual may
Oppose interracial marriage because he believes that black-white marriages challenge the
values of cultural tradition and social order. , If, however, this individual is presented
with a message that explicates how increased race relations will lead to greater awareness
and appreciation of racial differences which will heighten cultural stability and cohesion,
then the message recipient should no longer view opposition to interracial marriage as ‘
threatening his cultural tradition, and his attitude should change:

A major criticism of the early functional theories was the lack of adequate
methods for measuring a priori the functions being served by an individual’s attitudes.
Operationalizing the functions has proven to be quite difficult for subsequent researchers

(DeBono, 1987; Eagly, & Chaiken, 1993; Shavitt, 1989). Consequently, little theoretical

15

development was made and thereafter this theory was overlooked for decades. Recently
a resurgence of interest in the functional approach to attitudes has emerged (DeBono,

1987; Herek, 1987; Shavitt, 1989, 1990).

Herek’s theoretical extension. The recent functional researchers have extended
the original theory by clarifying the overlapping nature of the functions as well as
proposing new methods for measuring the various functions (Eagly, & Chaiken, 1993).
One such extension elaborates on the assumption that any given attitude may serve
multiple functions for an individual by emphasizing the idea that one function is primary.

Herek (1986) outlined six separate functions which fall under two distinct
dimensions: evaluative and expressive First, Herek (1986) defined the evaluative
category as one in which the attitude object is treated as an end in itself because of
perceived rewards or pmrishments. “In general, positive attitudes toward an object tend
to result when it is perceived as a source of benefit, reward or pleasure; negative attitudes
result fromt'pastof anticipated detrimental, unpleasant, or punishing experiences with it” {
(p.105). According to Herek (1986) three evaluative functions exist. First, attitudes may
be exponentral-specrfic, whereby after interacting with a particular instance of the
attitude object category, it is treated as a unique entity, differentiated from its
membership in the larger category, and evaluated in terms of its individuated utility for
that person. For example, a white person may deve10p positive attitudes toward a
specific black co-worker after pleasant interactions with her. Yet, this positive attitude

will not generalize to other black females.

Second, attitudes canheexpcncntial-schctnatic, whereby the individual treats the
attitude object as representative of a larger category perceived as either beneficial or
detrimental to oneself. Past experiences with representatives of the category lead to the
developmentof cognitive schemas that guide subsequent interactions with members of
this category. .. For example, a white person working with a black female may find the
experience beneficial and from then on hold a positive attitude of all black female
coworkers. Finally, evaluative attitudes can be based on anticipated firture utility. This is
not based on direct experience but on other types of information such as vicarious
observation and learning. For example, a white person who expects to work in a diverse
company may rely on attitudinal information about black coworkers obtained through
television or readings.

The second dimension Herek (1986) identified was an expressiveness category.
These attitude functions are manifested when an attitude’s benefit comes from its
expression. The attitude object is a means to an end; it provides a vehicle for securing
social support, for increasing self-esteem or for reducing anxiety. Attitudes that serve
this function can be better understood by analyzing the individual’s group identifications,
self-concept, and intrapsychic dynamics. Symbolic attitudes are best categorized in this
expressive dimension (Herek, 1986). From a communication perspective, the expressive
dimension also lends itself to investigation about how the functions would be manifested
in explicit persuasive messages. There are at least three expressive functions. The soCial
expressive functibn is based on the need to be accepted by others in one’s own immediate
social environment. For examme, a white female: may hold'negative attitudes toward

dating black males because‘her parents disapprove of such a relationship. Thc‘defensive .4

l7

function is based on needs to reduce anxiety caused by intrapsychic conflicts which are
usually unconscious. For example, a black female may have hostile reactions toward
interracial dating because she unable to secure a successful romantic relationship. She
may unconsciously see other women, especially white women, as a barrier or threat to
initiating such a relationship. She then projects her insecurity as hostility whenever she
sees a black man with a white woman. The attitude expressed would be one of aversion
or repulsion

The value expressive function is based on the need to define oneself by
expressing important values and aligning oneself with important reference groups. For
example, a black female who is Muslim-black may hold a negative attitude toward
interracial dating because her attitude enables her to express her Muslim customs
regarding religious tradition. Her opposition to interracial marriage is based on the extent
to which she upholds and verbally identifies with her Muslim-black values. Stronger
relations between value importance and attitudes should occur when attitudes are value-
expressive than when they serve any other function, since value-expressive attitudes are
assumed to be based on underlying values (Maio & Olsen, 1994; 1995). Thus,
individuals with varying value-expressive attitudes (pro vs. con) toward interracial
marriage would be expected to consider different values (Transcendent versus
Conservative) important.

In sum, Herek’s neofunctional model is useful because it distinguishes expressive
functions from evaluative functions. Herek (1986) acknowledged that many people can
hold the same attitude but for different reasons (frmctions). Whereas Katz and Smith

tended to view attitude functions as comparable to personality traits and therefore

18

relatively stable; Herek’s model allowed attitude functions to vary across attitude
domains, personality traits, and situations. Previous research on values and symbolic
attitudes suggests that the domain of interracial marriage, an indicator of intergroup
relations, is useful for examining the value expressive function as it impacts attitudes and
behavior. Aside from Shavitt’s (1990) direct thought listing measure of attitude
functions, most functional research has used indirect measures via a personality scale.

Situation variation is not examined here.

BMW sources of thgyalue expressive function

Herek suggested that value expressive attitudes are likely to be manifested in
peoplemthstrdirgbehefs and emotions associatedwitha particular ideological system.
Persons with value-expressive functions are likely to pay less attention to social cues, and
more attention to internal beliefs and values. Recent studies have employed the self-
monitoring construct to identify attitude functions ( DeBono, 1987; Kristiansen & Zanna,
1988; Maio & Olson, 1994). Snyder (1974) defines self-monitoring as the manipulation
of the image that we present to others in our interpersonal interactions. High self-
monitors carefully adjust their behaviors on the basis of feedback from others so that they
may produce the most desirable impression. In contrast, low self-monitors are not
concerned with the image they present to others. Rather, their interactions are
characterized by an extreme openness in which they communicate their thoughts and
feelings with little attempt to manipulate the impressions they create. Snyder (1974) has

identified five components to the Self-Monitoring construct: concern for appropriateness

of social behavior, attention to social comparison information, ability to control or

19

modify self-presentation, use of this ability in particular situations, and cross-situational
variability.

Snyder and DeBono (1985) demonstrated that high self-monitoring individuals
formed favorable attitudes toward objects that were useful to them as a means for
achieving the goals presented by images appropriate to their social situations: However,
low self-monitors reacted positively to objects and issues that allowed them to express
their underlying values. As a result of this finding additional researchers have employed
the construct of self-monitoring as an indirect measure of identifying the value-
expressive and the social adjustive function.

For instance, DeBono (1987) found that attitudes of high self-monitors were more
likely to serve a social-adjustive function while attitudes of low self-monitors were more
likely to serve a value-expressive function. DeBono’s (1987) explanation for this result
rested on the propensity of low self-monitors to guide their behavioral choices on the
basis of values, feelings and dispositions in contrast to high self-monitors who adapt their
behavior to the social situation. Similarly, Kristiansen and Zanna (1988) found that low
self-monitors, but not high self-monitors justified their attitudes by appealing to values
which they regarded as relevant to the target attitude issue (abortion and nuclear
weapons). Kristiansen and Zanna (1991) also found support for the Self-Monitoring
construct as an indirect indicator of the value-expressive function. They reported that the
attitudes of low Self-Monitors were value-expressive because their attitudes were
correlated only with those values they deemed relevant to the attitude issue (capital
punishment and affirmative action), while there was not such distinction for the high

Self-Monitors. Although Maio and Olson (1994) did not find a correlation between self-

20

monitoring and attitude fimction, they make a call for more research to examine the role
of self-monitoring.

It is important to note that the factor structure of Snyder’s Self-Monitoring scale
does not yield a unidimensional construct of the five underlying components. Rather,
exploratory factor analysis (Briggs, and Check, 1988; Gabreyena and Arkin, 1980) and
confirmatory factor analysis (Dillard and Hunter, 1989) revealed that items tended to load
highly on only one of three or four factors. These factors are labeled
Extroversion/Sociability, Other-Directedeness, and Acting. Gabreyena and Arkin (1980)
yielded a fourth factor which they labeled Speaking Ability. The most recent factor
analysis was conducted by Dillard and Hunter (1989) and they found evidence for a
primary four factor model (Extroversion, Sociability, Other-Directedness, and Acting),
but after tests for parallelism, the Self-Monitoring scale was invalidated and yielded the
primary factor of Other-Directedness and the second order factor of Acting with
Exu'oversion/Sociability. Dillard and Hunter (1989) emphasized the construct invalidity
of this measure and point out that research results utilizing this measure must be
reconsidered. As a result, Lennox and Wolfe (1984) developed a revised scale contains
13 items which measure two dimensions of the construct of Self-Monitoring: sensitivity
to the expressive behavior of others and ability to modify self-presentation. Dillard and
Hunter (1989) report that this scale “exhibits good conceptual fit with a slightly altered
construct as well as strong psychometric properties” (p. 126). Lennox and Wolfe (1984)
did not incorporate the other-directedness and extroversion scales in their construct of
self-monitoring because those items yielded differential associations with social anxiety

as an external comparison scale.

21

Kristiansen and Zanna (1988) note that the extroversion subscale was primarily
responsible for the self-monitoring differences such that value justification was displayed
by individuals low in extroversion as opposed to those high on extroversion.
Conceptually, the claim that Self-monitoring is related to value justification is nullified.
Similarly, Zweigenhafi and Cody (1992) examined the Self-Monitoring of black students
on a predominantly white campus, but they found that the black students scored
significantly lower than White students only on the Other-directedness subscale. Once
again the conceptual nature of Self-Monitoring construct distorts the findings. In this
case, black students were not less extroverted or less able to act. Rather, they were less
likely to be willing to mask their true feelings by conforming to the perceived
expectations of others.

Given the factor analytic results of the Self-Monitoring scale a better conceptual
link to the value expressive function would be to test the viability of the revised Self-
Monitoring scale (Lennox and Wolfe, 1984) as well as the Other-directedness sub-scale
and the Extroversion sub-scale. Omitting the Other-directedness and Extroversion
subscales, suggests that the revised Self-Monitoring scale is not expected to reflect the
same pattern that Debono (1987) found between value—expressives and social-adjustives.
Past research has demonstrated that Other-Directedness contributes to the relationship
between Self-Monitoring and attitude functions such that individuals high in Other-
directedness may hold the social expressive (adjustive) function while individuals who
score low on Other-Directedness may hold the value expressive function. Extroversion’s
role is tentative at best but worthy of examination. Aside from measurement issues,

research on values and the value-expressive function is important to consider.

22

Research on values and functions
Two sets of researchers (Kristiansen and Zanna, 1988; 1991 and Maio and Olson,

1994; 1995) have examined the relationship between values and attitudes for individuals
holding the value expressive function, and they make a distinction between value
importance and value relevance. Both concepts are important when examining the value
expressive function. Value importance captures the general value priorities that
individuals hold in life, while value relevance reflects how instrumental individuals
believe specific values are in promoting a certain attitude. For instance, a person who
favors abortion might perceive the value of freedom more relevant to this issue, while an
individual who opposes abortion might perceive the value of human life as more relevant
to this issue, however, both individuals similarly rate freedom and human life as two very
important values in their lives. Kristiansen and Zanna (1988) demonstrated that
individuals who had different attitudes (pro and con) towards the issues of abortion and
nuclear weapons did in fact appeal to different values. For example, subjects who
opposed nuclear weapons regarded wisdom as the most relevant value, while subjects
who favored nuclear weapons regarded national security as the most relevant value. This
value justification effect was stronger for low self-monitors relative to high self-monitors,
and this was noted as evidence that value justification effects are more likely to occur
when attitudes fulfill a value expressive function.

Kristiansen and Zanna (1988) provided evidence that the value relevance concept
was more informative in illustrating value justification effects than value importance
because it specifically deals with the attitude issue and the attitude valence. Certain

values will be more strongly connected with certain issues. In a subsequent study (1991)

23

assessing attitudes toward capital punishment and affirmative action, Kristiansen and
Zanna found that when attitudes were value expressive, value relevance affected the
value-attitude relationship, whereas when attitudes were not value expressive relevant
and. irrelevant values did not. differentially impact attitudes. Specifically, the relation
between attitudes toward capital punishment and relevant values was greater among low
self-monitors relative to high self-monitors. Kristiansen and Zanna (1991) suggest an
explanation for the correlation between people’s values and their attitudes, such that
attitudes may be tied to values because they fulfill a value expressive fimction. That is,
people, with favorable attitudes, perceive that an attitude issue (capital punishment)
promotes relevant values.

Maio and Olson (1994; 1995) extend research on the value-attitude relation by
examining the moderating role of attitude firnctions in the latter relation. Unlike
Kristiansen and Zanna, Maio and Olson (1994; 1995) argue that value importance is the
key to understanding the role of the value expressive function because value expressive
attitudes are based on underlying important values. Maio and Olson (1994) posit that
value importance ratings indicate more stable general values that may be related to many
attitudes as opposed to relevance rating which are related to only one attitude, thus
importance ratings are overarching.

Maio and Olson (1994) reported that subjects with value expressive attitudes
(measured by a thought-listing task) exhibited significant relations between value
importance and their attitudes towards attending a dance to support construction of a
enclosed on-campus smoking area, whereas subjects with utilitarian attitudes did not

exhibit significant relations. In this study Maio and Olson made the value expressive and

24

the utilitarian functions salient by presenting all subjects with arguments appealing
toward both values and tangible benefits regarding attending a dance to support the
construction of an enclosed smoking area. Therefore, Maio and Olson created a new
attitude topic so that attitudes and attitude functions would not be pre-existing. This
procedure differs from Kristiansen and Zanna in that issues of abortion, nuclear weapons,
capital punishment and affirmative action are preexisting and may have existing primary
functions. Similarly, in this study it is assumed that attitudes toward interracial marriage
are pre-existing.

Additionally, Maio and Olson (1994) limited their analysis of important values to
four, which they selected a priori; freedom, individualism, collective well-being and
health because they all related to the issue of smoking, whereas Kristiansen and Zanna
(1988; 1991) analyzed all eighteen of Rokeach’s terminal values. Maio and Olson (1994)
also focused on favorable attitudes toward the dance, as opposed to negative attitudes,
because all the arguments presented supported the dance. In this study, certain values
will be selected a priori based on Schwartz’s (1992) work, and both positive and negative
attitudes toward interracial marriage will be examined.

Maio and Olson (1995) extended their previous study by experimentally
manipulating attitude function. Attitude function was manipulated by making salient
either utilitarian or value expressive reasons for attitudes toward donating to cancer
research. As before, attitude frmction was measured via Shavitt’s (1990) thought listing
task. In addition, Schwartz’s value survey was used and two lower order values of
altruism and helpfulness and the higher order domains of self-transcendence/self-

enhancement and openness to change/conservatism were selected a priori. Results

25

indicated that the value-attitude correlations were significant in the value expressive
attitude condition but not in the utilitarian condition. Specifically, for subject’s in the
value expressive condition attitudes towards donating to cancer research were negatively
related to self-enhancement values (success, social power) and positively related to their
self-transcendence values (honesty, equality). In contrast, utilitarian condition subjects’
attitudes were not related to any of the higher order value domains, thus no reliable
pattern of value-attitude relations was apparent.

Maio and Olsen (1995) also examined the extent to which attitude function
predicted behavioral intention. Utilizing the Theory of Planned Behavior (Ajen, 1988),
Maio and Olsen (1995) determined that the correlation between altruistic values and
behavioral intentions to donate to cancer research was significantly stronger for value-
expressive individuals than utilitarian individuals. That is, when attitudes served a value-
expressive function, altruistic values uniquely predicted behavioral intentions apart from
the variables of attitudes toward donating, subjective norms and perceived control.
Maio and Olsen (1995) discuss the possibility that values may be linked to a sense of
moral obligation to behave in a specific fashion which can mediate the relationship
between values and behavioral intentions.

In sum, extant research on the functional approach extended the original theory by
integrating and clarifying the functions as well as introducing three sources that identify
how attitude functions can vary. It is argued that similarattitudes. can serve different
functions, depending on (the individual,.domain or situation; The value expressive
fimction has been demonstrated to be linked to the personality variables of Other-

Directedness and Extroversion. Additionally, Herek (1986) suggests that symbolic

26

attitudes fulfill a value expressive function in that they express values tied to a person’s
group membership. Interracial marriage is an issue that evokes symbolic attitudes that are
based on certain important values. Previous research from Maio and Olsen (1994; 1995)
also found evidence for the moderating role of the attitude functions on the relationship
between values, attitudes and behavioral intentions.

Up to this point discussion has been limited to values and attitudes. An
appropriate extension for this analysis requires examining the extent to which the value-
expressive function can impact behavior in the form of message production. Specifically,
the expressive component of the value-expressive function necessitates an examination of
the extent to which value appeals are used in persuasive messages reflecting attitudes
toward interracial marriage. John Greene’s Action Assembly Theory provides an

explanation for the value-expressive function’s impact on construction of message

appeals.

Action Assembly Theory

Action Assembly Theory (AAT) comprises five axioms and seventeen theoretical
propositions Greene, 1984). Together these statements specify the structures and
processes that account for almost all human communicative production whether
routinized, creative, verbal or nonverbal. AAT posits two basic structures (procedural
records and output representation) and two processes (activation and assembly). AAT
makes a distinction between declarative memory and procedural memory. Declarative

memory refers to the long-term store of factual information, while procedural memory

27

refers to information about how to execute cognitive and motor activity. The focus of
AAT is on procedural memory because it deals with behavioral message production.

According to AAT, the procedural memory is composed of a large number of
modular elements termed “procedural records.” Each procedural record comprises a
number of symbolic primitives linked by action-outcome relations along with any
situational features that have proven significant in mediating that relationship.
Procedural records may be defined at numerous levels of complexity ranging from
controlling muscle movements to abstract planning.

In order to affect behavior, a procedural record must exceed some threshold level
of activation. The activation level of any given record is enabled when (a) a goal or
desired outcome occurs that matches the outcome stored in record or (b) any situation
features occur that match the features stored in the record. Additionally, each procedural
record is characterized by a level of strength which is a function of recency and
frequency of activation. Due to the overwhelming number of simultaneously activated
records occurring, AAT requires an assembly process that can organize and integrate the
activated records into coherent representations of an action to be performed.

This assembly process is a structural representation of action-to-be-taken, defined
as the “output representation.” An output representation results from the assembly of
activated procedural records. The assembly process proceeds serially within each level
of output representation. At the highest levels of output representation are abstract plans
specifying general properties of behavior, and at the lowest levels are concrete
specifications of motor commands guiding muscle movements. Additionally, assembly

takes time and places considerable demand on central processing capacity. This demand

28

on capacity can be circumvented by develOping structures that eliminate the need for
assembly, such as practice or prior assembly.

AAT can explain the extent to which value appeals are prevalent in persuasive
messages for those individuals whose attitudes toward interracial marriage serve a value
expressive function. Applying AAT, attitude functions can be described as general
procedural records whose action-outcome link varies, depending on the type of function
activated. The expressive dimension of the value expressive function specifies that
explicit elaboration of important values is necessary for the attitude’s existence. For the
value-expressive function, maintaining an attitude toward interracial marriage (action) is
linked to the elicitation of values (outcome). Once the function (procedural record) is
activated then it is assembled, producing subsequent message elaboration (output
representation). The output representation is the actual expression (either verbal or
written) of values arguing either for or against interracial marriage.

In order to ensure that message production (output representation) stems from the
value expressive function the correct procedural record has to be readily activated.
According to AAT, activation is a function of recency, frequency and goals associated
with the record. As noted above fimctional theory argues that attitudes exist because they
fulfill various goals (i.e. functions). It follows, then, that attitude functions, as procedural
records, are activated whenever attitudinal issues arise. By definition, functional theory
posits that the link between functions (procedural records) and attitude expression is
inherent and frequent, resulting in increased activation. That is, the subsequent message
appeal (output representation) will depend on the specific procedural record which will

vary depending on which attitude function is being activated.

29

Functional theory also argues that functions can vary depending. on the domain,
personality traits or the situation. Thus, for certain individuals, the domain of interracial
marriage will trigger the need to express values associated with their attitudes toward
interracial marriage. Once value-expressive attitudes are identified, then those same
individuals’ written messages will be the assembled representation of the activated
procedural record. In other words, written persuasive messages of value-expressive
individuals will contain value appeals because they will reflect the activated value

expressive procedural record.

Hypotheses
Herek’s (1987) thought listing measure will be utilized to directly identify

participants’ primary attitude functions held by both favorable and unfavorable attitudes
toward interracial marriage. Research by Herek (1984), Kristiansen (1988), and Wyman
and Snyder (1997) suggests that issues pertaining to prejudice or intolerance will
demonstrate symbolic attitudes tied to the expressive functions.

H1: Similarly, within the domain of interracial marriage, it is expected that
participants will use a significantly higher number of thoughts coded as expressive
(value-expressive and social-expressive) than thoughts coded as evaluative.

Additionally, research indirectly measuring the value expressive and social
expressive functions via the Self-Monitoring scale has yielded inconclusive and invalid
results. Therefore the subscale of Other-directedness will be utilized, and it is predicted

that:

30

H2: As participants’ value-expressive scores on the thought-listing task increase,
their scores Other-directedness scale will decrease.

In terms of the Extroversion scale, research findings have been mixed. Thus, the
following research question is posed:

RQl: What is the relationship between individuals’ value-expressive thought-
listing scores and their scores on the Extroversion subscale?

Prior research has demonstrated that low scores on Snyder’s (1974) Self-
Monitoring scale are indicative of individuals holding a value-expressive fimction while
high scores on the Self-Monitoring scale are indicative of individuals holding a social-
expressive function. As such, scores on the Self-Monitoring scale (Snyder, 1974) were
treated as indirect measures of the value-expressive and social-expressive functions. The
Revised Self-Monitoring scale (Lennox and Wolfe, 1984) may not yield the same pattern
because the construct of Self-Monitoring has been narrowly defined. Lennox and Wolfe
(1984) developed a revised Self-Monitoring scale which excluded the factors of Other-
directedness, Extroversion and Acting. Consequently, the construct that Debono (1987)
used is different than the one presented here and a second research question is posed:

RQ2: What is the relationship between individuals’ value-expressive scores on
the thought-listing task and their scores on the Self-Monitoring subscale?

Prior research by Maio and Olson (1994; 1995) and Kristiansen and Zanna
(1988) examined how value relevance and value importance were distinctly related to the
value expressive fimction. This study will replicate and extend the previous research by
providing clarification on the relationship between values and symbolic attitudes (i.e.

interracial marriage). Schwartz’s value model can be utilized to test relationships

31

between self- transcendent as well as openness to change/conservation values with
favorable and unfavorable attitudes toward interracial relationships for value expressive
individuals, where they are expected to be the strongest. Consistent with Kristiansen and
Zanna (1988; 1991), the relation between value relevance, controlling for value
importance, and attitude will demonstrate that attitude function will moderate the
relationship between relevant values and attitudes supporting and opposing black-white
interracial marriage. Additionally, in their study on predicting readiness for outgroup
contact Sagiv and Schwartz (1995) utilize the over-arching value domains of Self-
Trancendence-Self-Enhancment and Openness to Change-Conservation. Similarly, it is
predicted that: .

H3a: As participants’ Self-transcendent values increase their attitudes toward
interracial marriage will become more favorable, and this relationship will increase in
strength for individuals whose thoughts are higher in value-expressiveness.

H3b: As participants’ Self-enhancement values increase their attitudes toward
interracial maniage will become less favorable, and this relationship will increase in
strength for individuals whose thoughts are higher in value-expressiveness.

H30: As participants’ Openness to change values increase their attitudes toward
interracial marriage will become more favorable, and this relationship will increase in
strength for individuals whose thoughts are higher in value-expressiveness.

H3d: As participants’ Conservation values increase their attitudes toward
interracial marriage will become less favorable, and this relationship will increase in

strength for individuals whose thoughts are higher in value-expressiveness.

32

Action Assembly Theory (Greene, 1984) explains how attitude functions serve as
procedural records that when activated and assembled impact the use of function-specific
persuasive appeals regarding interracial marriage. It is predicted that:

H4a: As participants’ attitudes toward interracial marriage increase they will be
more likely to express Self-Transcendent values and this relationship will increase for
individuals whose thoughts are high in value-expressiveness.

H4b: As participants’ attitudes toward interracial marriage increase they will be
more likely to express Openness to change values and this relationship will increase for
individuals whose thoughts are high in value-expressiveness.

H4c: As participants’ attitudes toward interracial marriage decrease they will be
more likely to express Conservation values, and this relationship will increase for
individuals whose thoughts are high in value-expressiveness.

H4d: As participants’ attitudes toward interracial marriage decrease they will be
more likely to express Self-enhancement values and this relationship will increase for

individuals whose droughts are high in value-expressiveness.

33

CHAPTER 3

METHOD

Participants

The sample consisted of 226 college students enrolled at two separate
universities: a large Western university and a large Midwestern university. All students
were enrolled in undergraduate communication courses. Two hundred twenty-six
subjects were utilized for confirmatory factor analyses (CFA). Subsequent to CFA, ten
subjects’ surveys were pulled and used for coder training and practice while another 31
randomly selected surveys were used to establish unitizing and coding reliability and
were not employed in the main analyses.l Demographics will be provided for the original
226 participants followed by demographics for the remaining 185 as a whole, ending with
demographics pertaining to each individual university.

The sample of 226 participants consisted of 135 females and 91 males. The ages
of the participants ranged fiom 17-49 (M: 21.71, §_D= 3.86). Thirteen and seven tenths
percent of the students were freshman, 13.7% sophomores, 27% juniors, and 42.5%
seniors while 3.1% did not indicate. The ethnic background or race of the participants
were as follows: 8.8% Asian, .9% Pacific Islander, 12.8% African American, 5.8%
Latino/a, .4% Native American, 63.3% Caucasian, and 1.3% did not indicate.

The sample of 185 students consisted of 110 females and 75 males. The ages of
the participants ranged from 17 to 49 (M = 22, & 4.1). Nine and seven tenths percent
of the students were fi'eshman, 13.5% were sophomores, 27% were juniors, and 45.9%

were seniors while seven subjects left this item blank. The ethnic background or race of

34

the subjects is as follows: 10.3% Asian, 1.1% pacific islander, 14.1 African American,
5.4% Latino/a, .5% Native American, 61.6% Caucasian, and 6.5% other/multi-racial
while one subject left this item blank.

The age of the participants of the Midwestern sample ranged from 17-26 (_M_ =
20.5, _S_12= 1.6). Sixty-four females and 47 males participated in this study. Fifteen and
three tenths percent of the subjects were freshman, 18.9% sophomores, 27 % juniors, and
37.8% seniors while one subject left this item blank. The ethnic background or race of
the subjects is as follows: 9.0% Asian, 1.8% Pacific Islander, 18.9% African American,
1.8% Latino/a, 64% Caucasian, and 4.5% other/multi-racial.

The age of the participants of the Western sample ranged from 19-49 (M = 24.3,
_S_D = 5.4). Forty-six females and 28 males participated in this study. One and a half
percent of the students were freshman, 5.9% sophomores, 29.4% juniors, 58.1% seniors,
and 8.1% left this item blank. The ethnic background or race of the subject is as follows:
12.2% Asian, 6.8% African American, 10.8% Latino/a, 1.4% Native American, 58.1%

Caucasian, 9.5% other/multi-racial, and 1 individual left this item blank.

Procedure

This investigation entailed collecting data at two time points. First, participants
were given a packet of surveys, and one week later they were asked to come back to
complete another series of surveys. Each time point lasted approximately twenty to thirty
minutes. Upon completion of the second survey, the students were debriefed and
thanked. In the first survey, participants completed a revised Self-Monitoring scale, an

Other-directedness subscale, Extroversion subscale, a thought-listing task asking them to

35

list all their thoughts and feelings relevant to supporting or opposing interracial marriage,
rated their attitudes toward interracial marriage, indicated the relevance of values to their
attitudes toward interracial marriage, and completed a attitude function survey.

At the second administration (1 week later) participants were asked to construct a
persuasive message arguing their opinion about interracial marriage. They were asked to
direct their message to their sister or brother (condition 1) or their best friend (condition
2) who has just informed the participant that they are considering getting engaged to a
black or white individual (depending on ethnicity of fiiend or sibling). The participant
was asked to identify their sibling’s or friend’s race and sex. Next, they completed an
attitude functions questionnaire and indicated the relevance of specific values for their
attitudes toward interracial marriage. Finally, participants were thanked and debriefed.

(See Appendix A for the complete time 1 and time 2 questionnaires.)

Instrumentation/Measurement._All scales discussed in this section were submitted
to confirmatory factor analysis and their relevant statistics are reported in the results

section.

Self-Monitoring. Participants completed an altered version of Snyder’s (1974)
self-monitoring scale developed by Lennox and Wolfe (1984). The revised scale contains
13 items which measure two dimensions of the construct of Self-Monitoring: sensitivity
to the expressive behavior of others and ability to modify self-presentation. Coefficient
alpha for the sensitivity subscale was .77, for the self-presentation subscale .70, and for

the combined scale .75 (Lennox & Wolfe, 1984). Participants’ scores on the Self-

36

Monitoring scale, ranged from 1( low in self-monitoring) to 7 ( high in self-monitoring)
on a Likert scale.

In addition, two more subscales measuring other—directedness and extroversion
were included. The other—directedness subscale contains five original items from
Snyder’s (1974) scale which Dillard and Hunter (1989) and Briggs and Check (1988)
identified . In addition, 5 items from Lennox and Wolfe’s (1984) cross-situational
variability subscale ( a= .82) were incorporated. Participants’ scores ranged from 1 ( low

in other-directedness) to 7 (high in other-directedness) on a Likert scale.

Four of Snyder’s (1974) extroversion items identified by Lennox and Wolfe
(1984) and Briggs and Check (1988) were also be included. Participants’ scores ranged
from 1 (low in extroversion) to 7 (high in extroversion) on a Likert scale. The other-
directedness and extroversion scale were treated as separate scales nor will they be
combined with the revised self-monitoring scale. No reliability data are reported because

these items have not been combined in this fashion before.

Attitude assessment. Attitude items consisted of seven (self-constructed)
semantic differential items reflecting a general attitudinal measure toward black-white
interracial marriage. Participants indicated their attitudes toward interracial marriage on
eight 7-point (1 to 7) semantic differential scales anchored by the following adjectives:
good/bad, rewarding/punishing, acceptable/unacceptable, respectful/disrespectful,
harmfirl/beneficial, intolerable/tolerable, satisfactory/unsatisfactory, and
offensive/pleasant. Once the items were averaged across, higher scores reflected the
positively valenced adjective while lower scores reflected the negatively valenced

adjective.2

37

 

 

 

 

 

 

 

 

 

 

 

 

(IQ

Value Relevance. Students completed a shortened versions of Schwartz’s (1992)
56- item value survey. At time 1 students were asked to rate “how relevant are the values
listed below to your consideration of black-white interracial marriage.” (Kristiansen,
1991) A short definition was provided adjacent to each value. The 35 value items were
in Likert format with response options ranging from I (totally hinders value) to 7 (totally
promotes value).

The 35 values that were chosen for the shortened version were based on
Schwartz’s (1992) nine movtivationally distinct lower order value types (hedonism is
excluded). Additionally, Schwartz’s value theory posits that the nine lower order value
types can be organized into a higher order 2-dirnensional representation (Self-
Transcendence - Self—Enhancement, and Openness to Change - Conservation). Thus, the
hypotheses in this study were written under the assumption that Schwartz’s value theory
would be supported by a lower-order nine-factor solution of value domains which then
can be summed to create these higher order domains. This would be evidenced by second
order factor analysis with a four-factor solution. The selection of values was based on a
study by Sagiv and Schwartz (1995) that examined Israeli Jewish teachers’ (dominant in-
group) readiness for social contact with Israeli Arabs (minority out-group).

The higher order dimension of self-transcendence should combine lower-order
benevolence (helpful, forgiving, and mature love) and universalism (equality, world at
peace, wisdom, social justice, and broad-minded). Self-enhancement should combine
power (social power, wealth, social recognition, authority, and preserving public image)
with achievement (successful, capable, influential, and ambitious). Openness to change

should combine self-direction (fieedom, independent, curious, choosing goals, and

38

creativity) with stimulation (exciting life, varied life, and daring). Conservation should
combine security (sense of belonging, social order, and family security) with conformity
(self-discipline, honoring parents, and obedient) and tradition (respect for tradition,

moderate, accepting portion in life, and devout).

Thought Listing Analysis

Thought listing across both time points was conducted in the same fashion,
utilizing the same coding schemes, in order to ensure coding consistency of functionally
relevant value expressive thoughts. The thought listing task at time 1 was used to directly
measure participants’ level of value expressiveness. The thought listing task at time 2

was used to measure value appeals in the messages constructed by the participants.

Time 1 - Level of Value Expressiveness. Participants were asked to list all their
thoughts that were relevant to their approval or disapproval of interracial marriage in
order to measure (1) the extent to which an individual holds a general expressive, a
general evaluative or neither as well as (2) the degree to which the participant holds the
value expressive function (high value expressive vs. low value expressive). The first
coding scheme measured the broader expressive, evaluative or no function category while
the second coding scheme measured the level of value expressiveness as a function of
number of the 35 specific values taken from Schwartz’s scale. A total number of value
expressive thoughts was calculated for each participant. Higher numbers indicated higher
levels of value expressiveness. The directions to the thought-listing task were as follows:

In the space below, please describe how you feel about interracial marriage
between blacks and whites. Please, describe and explain your attitude towards black-

39

white interracial marriage. Please indicate why you feel the way you do. Write down all
of your thoughts and feelings that are relevant to your attitude and try to describe the
reasons for your feelings.

**Please be as specific as possible***

Respondents received a blank sheet of paper, and they were asked to begin their
thought listing with the statement, “I have generally positive or negative attitudes toward
black-white interracial marriage because...” Herek’s (1987) and Maio and Olsen’s (1994)
content-coding schemes were adapted to score the functions of attitudes toward
interracial marriage.

The unit of analysis was a functionally relevant thought that was defined as “a
statement that contains one complete thought related to the respondents’ attitude toward
interracial marriage” (adapted from Herek, 1986). Each functionally relevant thought
was coded using two coding schemes. The first coding scheme classified the thought
according to whether it reflected a general expressive function (01), a general evaluative
function (02) or neither fimction (03). For example, “I knew an interracial couple back
home” is an example of a thought coded as evaluative while “My religion does not allow
me to marry outside of my faith” is an example of a thought coded as expressive. The

second coding scheme was established in order to identify the specific value held, when

the thought was value-expressive.3

Coding scheme 1- general function categog. Three general categories were

identified. They include expressive, evaluative and none. A thought was categorized as
expressive if it reflected appeals to or content about one’s values, social groups and social

acceptance or ego-defensive thoughts regarding interracial marriage.

40

A thought was categorized as evaluative if it reflected appeals to or content about
one’s personal experiences, anticipated experiences or attributes about interracial
marriage. These thoughts could reflect either negative, neutral or positive associations
towards interracial maniage. A thought was coded “none” if the thought did not include
appeals or content about values, social ties and acceptance, past/anticipated interactions,
or ego-defensive issues. Similarly, these thoughts could reflect either negative, neutral or

positive associations toward interracial marriage.

Coding scheme 2- smcific value categorization. Value expressive thoughts are

not based on personal interactions or experiences. These thoughts mention the
importance of values such as but not limited to equality, freedom, conservatism or
tradition and their expression as the determinant for attitudes toward interracial marriage.
The statements include descriptions of religious, and moral values as well as political
afliliations influencing attitudes. Consequently this value expressive category included
35 value subcategories based on Schwartz’s (1992) shortened value relevant scale (see
Appendix A) as well as an “other” category, meant to capture any additional values. A
total of 36 subcategories were utilized. Additionally, these thoughts were coded as
reflecting either negative, neutral or positive associations towards interracial marriage.

This score was used to identify people as high or low value expressive.
Time 2- Value AMS. Participants in time 2 were asked to construct persuasive

messages to either a best fiiend (condition 1) or a sibling (condition 2) who is getting

engaged to a black or white individual (depending on friend’s or siblings ethnicity) in

41

order to measure the extent to which individuals, who are coded as being high value
expressives as opposed to being low value expressive, mention specific values in their
message. The directions were as follows:

In the space below, please construct a persuasive message, based on your attitude toward
interracial marriage, arguing for or against your best fiiend’s (or sibling’s) engagement.
List specific reasons in support for or in opposition of this marriage.

Respondents received a blank sheet of paper, and they were asked to begin their
thought listing with the statement, “After listening to my Best Friends (or sibling’s) news
I would argue for or against this marriage by saying...”

The unit of analysis and the coding scheme are identical to time 1. (see above for
specifications) In addition, once the values were coded into their lower order
subcategories they were then summed and organized along Schwartz’s higher-order

domains: Self-Transcendence, Openness to change, and Conservation. This was done in

order to test hypotheses 4a-d.4

Coding of Functionally Relevgant Value Expressive Thopghts

 

Two independent coders coded 31 subjects’ thoughts on the time 1 task. The time
1 unitizing and interrater reliabilities were applied to time 2 because the manner in which
time 2 data was coded was identical to time 1 with the exception being that time 2 coding
was shorter because only value expressive thoughts were coded in their respective
subcategories.

The coders agreed 98% of the time when identifying functionally relevant
thoughts and their location, and Guetzkow’s U was .01 (Guetzkow, 1950). They then

resolved any differences in unitizing and proceeded to code the firnctionally relevant

42

thoughts using both coding schemes described above. The coders agreed 93% of the time
on the general category of the fimctionally relevant thought (coding scheme 1) with an
interrater reliability, using Cohen’s Kappa (Cohen, 1960) of .88. The results for the
second coding scheme are as follows: value expressive thoughts yielded 94% agreement
between the coders with a Kappa of .91 for value identification along the 36
subcategories and a Kappa of .78 for thought valence. The coders resolved any
differences in coding, then one coder coded the remaining 185 subjects’ thoughts across

time 1 and time 2.

43

CHAPTER 4

RESULTS

Overview

Results will be presented in order to parallel the analysis. Preliminary analyses
will consist of factor analyses of the scales (self-monitoring, other directedness,
extroversion, and value relevance), followed by the significance tests between the
Midwest and West samples and then the thought listing analysis. The main analyses will
consist of Chi square, correlational and linear regression results presented for time 1 and

then linear regression results presented for time 2.

Preliminary analyses
Confirmatory factor analyses (CF A) (Hunter & Gerbing, 1982) was employed to

test the dimensionality of the measures listed above. An eleven factor solution was
obtained with a reduced set of indicators that passed internal consistency and parallelism
tests (items, factor loadings, means, and standard deviations are listed in Appendix B).
When applicable internal consistency tests showed that the errors calculated between
items measuring the same construct were generally within sampling error of zero. The
parallelism test demonstrated similar results, the errors calculated between items

measuring different constructs were also within sampling error of zero.

Self-Monitoring. CF A confirmed the presence of two subscales: sensitivity to the

expressive behavior of others and ability to modify self-presentation. Sensitivity to

expressive behavior was reduced to 4 items with a standardized item alpha of .67 (_M for
West= 5.22, S_D=.94; M for Midwest= 4.78, SD; 1.07), and subjects from the
Midwestern and Western samples differed on this scale, _1 (183)= 2.98, p < .003. Ability
to modify self-presentation was reduced to 3 items with a standardized item alpha of .65
CM=5 .45, §D_= .91) and no differences between the samples were found. The scales were
not combined, but analyzed independently because the two samples of college
participants significantly differed on the sensitivity to expressive behavior scale.

Otherdirectedness. CFA retained 3 items with a standardized item alpha of .71
M=4.34, fl= 1.38).

Extroversion. CF A retained 2 items with a standardized item alpha of .52 (M=
4.73, §D_=130).

Attitpde toward interracial marrige. CF A retained 5 items with a standardized
item alpha of .89 (M; 5.36, S_D= 1.27).

Value relevance. CF A did not yield the 9-factor solution posited by Schwartz
(1992). Alternatively a 6-factor model was supported, however, it was not theoretically
consistent with Schwartz’s value domain organization, and it could not be combined in
any fashion to reflect the higher-order value domains of self-transcendence- self-
enhancement or Openness to change- Conservation.(see Appendix C for correlations
among the factors) The six factors that emerged will now be described. Factor 1:
gum consisted of just one item measuring the equality value therefore reliability
could not be estimated CM= 6.18, S_D= 1.28).

Factor 2: benevolent success consisted of 7 value items (helpful, forgiving,

mature love, capable, ambitious, independent, and choosing own goals) that combined

45

the lower order domains of benevolence, achievement and self-direction with a
standardized alpha of .92 (M for Wes? 5.1, S_D= 1.07; M for Midwest= 6.02, S_D= 1.01 ).
Participants from the Western university significantly differed from the students of the
Midwestern university on this variable, so all subsequent analyses using benevolent
success will not be collapsed across university location. The benevolent values retained
were helpful, forgiving, and mature love. The achievement values retained were capable
and ambitious. The self-direction values obtained were independent and choosing own
goals. Face validity implies that the domains of achievement and self-direction would be
closely associated because these specific values share qualities that reflect a motivation to
. pursue individual interests. On the other hand, benevolent values reflect a tendency to
promote the welfare of others first. However, interrfactor correlation among the 3
subscales were 1.00 and higher, indicating that essentially subjects were responding to all
3 lower order domains in the same manner. Thus, they were combined into one factor.
Factor 3: Power-tradition consisted of 3 items that combined the lower order
value domains of power (authority value) and tradition (moderate and accepting portion
in life values). Similar to the benevolent success factor subjects from the two university
samples significantly differed on this variable. The power-tradition factor had a
standardized item alpha of .75 (M for West= 4.70, §Q= 1.23; M for Midwest= 5.10, _SQ=
1.40) and subjects across locations signifcantly differed on this variable, t(183)= 2.01, p <
.046. On the face of it power and tradition may be closely associated because these
values reflect upholding dominance over people or resources with little resistance to the

status quo.

46

Factor 4: Social Order consisted of a single item measure of the social order
value therefore reliability could not be estimated (M=5.3 1, SD: 1.78). This factor
reflects a desire for social stability as form of a collective interest (Schwartz, 1992).

Factor 5 : Honoring parents consisted of a single item measure of the honoring
parents value therefore reliability could not be estimated (M=5.36, SD; 1.27). This
factor reflects a restraint on inclinations or social impulses to violate social norms or
expectations (Schwartz, 1992).

Factor 6 : Varied life consists of a single item measure of the value of a varied
life therefore reliability could not be estimated. Students scored significantly different on
this factor (M for West=5.59, S_D= 1.37; M for Midwest=6.08, §Q=1.29) and subjects
significantly differed on this variable, 1(1 83)= 2.45, p_< .02. This factor reflects
preference of stimulation and change that promotes optimal levels of activation and
arousal (Schwartz, 1992).

As a consequence of the new 6-factor value solution the study hypotheses
concerning values were re-worded to parallel the six value factors. So instead of
hypotheses H3a-d, there will be six (H3a-f) subcategories, reflecting the six factors. The
same will be done for hypotheses H4a-f. The actual content of the re-written hypotheses
will be explicated in the main analyses where the results of the linear regression runs will

be presented.
T-test analysis of university samples. Since two independent samples of subjects

participated in this study, t-tests were conducted to determine whether the 27 variables of

interest could be collapsed across samples or whether they would have to remain

47

separate. The t-tests identified five variables that yielded significant differences. Subjects
differed on the sensitivity to the expressive behavior subscale of Self-Monitoring,
t(183)= 2.98, p < .003 (Mfor West= 4.78, S_D= 1.07; M for Midwest= 5.23, S_D= .94).
Subjects also differed on the power/tradition value dimension, 1(1 83)= 2.01, p < .046 (M
for WesF 4.70, S_D= 1.23; M for Midwest= 5.10, S_D= 1.4). Subjects also differed on the
benevolent/success value dimension, 1(1 83)= 2.65, p < .009 (M for WesF 5.61, _S_I_)=
1.07; M for Midwest= 6.02, S_D= 1.01). Participants also differed on the single item
measure of the stimulation value dimension (a varied life), t(183)= 2.45 , p_< .02 (M for
West= 5.60, S_D= 1.37; M for Midwest= 6.08, §Q=1.29). Finally, subjects differed on
age, _t_(183)= -6.89, p < .000 (M for Midwest = 20.5, SD: 1.63; M for West= 24.27, SD:
5.43). Any further analyses involving these variables did not collapse across samples,

instead they were kept separate.

Content analysis. The results yielded two variables: level of value

 

expressiveness (M= 1.36, _S_D: 1.09) and amount of value content in a message
paralleling the six value factors of equality (M=.044, §Q= .26), Benevolent success
(M=.69, S_D= .76), varied life (M=0, SD=0), social order (M=0, SD=0), honor (M=0,
SD=0), and power-tradition (M= .006, S_D= .08).

The heavy basement effect (restriction in range) led to an alternative method
where variance of scores were increased by using a ratio of total number of value
expressive thoughts divided by total number of all thoughts generated and multiplying by
100 to obtain whole numbers. This same procedure was used for computing the value

content in the messages at time 2 (total number of each value type divided by total

48

number of thoughts generated.) These scores provided substantive information on the
proportion of value expressive thoughts and message content that contains specific
values.

The new descriptive statistics are as follows: the variable of level of value
expressiveness ranged from 0-100 (M=36.5 (36.5%), SD=28.18), and the variables of
amount of value content in a message paralleling the six value factors of equality ranging
from 0-50 (M=1.33 (13.3%), §Q=7.58), Benevolent success ranging from 0-100
(M=26.46 (26.5%), S_D= 31.73), varied life with (M=0, SD=0), social order (M=0,
SD=0), honor (M=0, SD=0), and power-tradition ranging from 0-25 (M= .16 (1.6%),
S_D= 1.99). These data suggested that value expressive thoughts did not predominate

students’ attitudes towards interracial marriage or their subsequent message production.

Hypptheses testing
Hyp_othesis 1: Expressives v. evaluatives. To test this hypothesis two new

variables were created: an Evaluative score (EV) and an expressive score (EX). These
variables were constructed in the same manner described in the content analysis section
with the value expressiveness (V E) and value message content variables. Essentially,
variance of scores were increased by using a ratio of total number of expressive and
evaluative thoughts divided by total number of all thoughts coded and multiplying by 100
to obtain whole numbers. It was predicted that a higher number of participants’ thoughts
would be coded as expressive (EX) significantly more than evaluative (EV). The Nest
comparison supported this prediction, t(180)= -4.24, p < .000 (M for .E_X= 22.51, _SQ=

27.03; M for EV=37.78, SD=28J). Overall, participants used 22.5 % evaluative

49

thoughts and 37.8% expressive thoughts when writing about their attitudes toward

interracial marriage.

H2: VE scores and other-directedness. It was predicted that participants’ value
expressive scores (VE) would be negatively related to their other—directedness scores.

Correlation data was inconsistent with this prediction, [(180) = .03, ns.

RQI: VE scores and extroversion. It was asked how value expressive scores were
related to extroversion scores. Correlation analysis showed that there was no linear

association between VE scores and extroversion, {(180), = .04, ns.

RQ2: VE scores and self-monitoring. The self-monitoring subscales were unable
to be combined so separate correlation analyses were done. Correlation analysis showed
that there was no linear association between VE scores and the ability to modify self-
presentation scale, ;(180) = .02, ns. Similarly, no linear association was found between
VE scores and the sensitivity to expressive behavior scale, g(180)= .03, ns. Additionally,
participants from the Western university significantly differed from the Midwestern
university on this scale so separate runs are also presented and they are insignificant,

West {(74) = .14, ns and Midwest [(1 1 1) = -.10, ns.
H3 structure. Given the content invalidity of Schwartz’s 9-factor model an

alternative model yielding 6 value scales was proposed. Consequently, hypothesis 3 and

it’s subcategories have been re-written to reflect the 6 value dimensions. Second, in order

50

to test this hypothesis and all subsequent ones the value expressive variable (V E) was
dichotomized by dividing it into high and low groups at the median (Mal= 37.50). It is
notable that the median is relatively low on a scale that ranged fi'om 0 - 100, indicating
that overall participants were not predominantly referencing their values when they wrote
about their attitudes toward interracial marriage. Finally, participants fiom the Western
sample significantly differed fiom the Midwestern sample on the value scales of
benevolent success, power-tradition and varied life. As a result, regression analyses
concerning these three factors were not collapsed across location of university but were
run separately on each sample, and the results will be presented reflecting this. Overall
analyses, combining the two samples, are presented in Appendix E - G for the purposes

of comparison only.

H3a: VB and @ualifl. It was predicted that as participants’ equality scores
increase their attitudes toward interracial marriage will become more favorable, and that
this relationship will increase in strength for individuals who were higher in value
expressiveness. Linear regression of attitudes on equality yielded mixed results. The
standardized slope for low value expressives was (B= .48), 1= 5.28, p < .000. The
standardized slope for high value expressives was (B = .21), _t = 1.99, _p < .05 . The
difference between the two unstandardized slopes was not significant, t(90)= 1.6, ns.
Table 2 presents the obtained means and standard deviations for equality and attitudes

toward interracial marriage.

51

Table 2. Means and Standard Deviations for the reggession of attitudes
on the egualitv value for low value expressivesjand high value expressives

 

 

Variable Low Value Expressive High Value Expressive
Equality M = 6.01 M = 6.48

SD = 1.39 SD = .91

N= 94 N =85
Attitude M = 5. 09 M =5.73

SD=1.38 SD=1.04

 

The data are consistent with the prediction that the regression of attitudes on the
equality value for low and high value expressives is substantial and positive. The data,
however are not consistent with the prediction that the effect of equality on attitudes was
stronger for high value expressives than for low value expressives. Comparison of the
unstandardized slopes showed that high and low value expressives did not significantly
differ on the extent to which their equality value impacted their attitudes towards

interracial marriage.

H3b: VB and Benevolent success. It was predicted that as participants’
benevolence values (higher order self-transcendence) and self-direction values (openness
to change) would positively impact attitudes while achievement values (self-
enhancement) would negatively impact attitudes, and that these relationships would
increase in strength for high value expressives. CF A yielded a single factor that

combined all three of these factors unidirnensionally. So, the new prediction would be

52

that as participant’s benevolent success value scores increase their attitudes toward
interracial marriage will become more favorable, and this relationship will increase in
strength for individuals who high in value expressiveness.

Linear regression of attitudes on benevolent success form value expressives in
the Midwestern sample yielded a statistically significant standardized slope (B = .56), t =
5.16, p < .000. Additionally, high value expressives yielded a significant standardized
slope @= .27), t= 1.99, p_< .05. The difference between the unstandardized slopes was
not significant, t(56) = 1.45, ns. Thus, the data are partly consistent with the hypothesis
in the sense that the regression SIOpes are substantial, however the direction is positive
reflecting the prediction for the benevolent and self-direction values but not for the
achievement values. Subjects scores on the achievement values did not reflect the
negative association with attitudes instead they were alternate indicators of benevolence
and self-direction. However, the data are inconsistent because the impact of
benevolence values on attitudes did not significantly differ for low and higher value
expressives.

Linear regression of attitudes on benevolent success for _lgfl value expressives in

the Western sample yielded a statistically significant standardized slope (B = .64), t =

 

4.75, p < .000. Additionally, high value expressives yielded a nonsignificant
standardized slope (B = .31), 1= 1.83, p < .ns. The difference between the unstandardized
slopes was not significant, 1(31) = 1.45, ns.

The data were consistent with the hypothesis such that the impact of benevolent
success was positive and substantial. for the lpw value expressives across both samples.

However, in the high expressive group the beta was not substantially different than zero

53

(within sampling error). Additionally, the effect of benevolent success did not
differentially impact attitudes for low and high value expressives. Table 3 presents the
obtained means and standard deviations for benevolent success and attitudes toward

interracial marriage.

Table 3. Means and Standard Deviations for the regpession of attitudes
on the ben_evolent success scale for low value expressives and higl_r value expressives

 

 

Variable Low Value Expressive High Value Expressive
Midwest M = 5.77 M = 6.32
Ben-success SD = 1.14 SD = .72
N= 60 N = 51
Attitude M = 4.96 M = 5.79
SD = 1.40 SD = .97
West M = 5.69 M = 5.69
Ben-success SD = 1.25 SD = .80
N= 35 N= 34
Attitude M = 5.25 M = 5.63
SD=1.37 SD=1.14

 

H4c: Power-tra_dition. It was predicted that the values of power (higher order self-
enhancement) and the values of tradition (conservation) would negatively effect attitudes
toward interracial marriage, and that this relationship would be stronger for high value
expressives. Linear regression of attitudes on benevolent success for_l_ow value
expressives in the Midwestem sample yielded a statistically significant standardized

SIOpe (B = .39), t = 4.11, p < .002. Similarly, high value expressives yielded a

54

significant standardized slope (B = .22), t: .2.10, p < .05. The difference between the
unstandardized slopes was significant, 1(56) = 1.88, p < .05. The data are not
consistent with the prediction because the slopes were not negative and the impact of
power-tradition did differentially impact attitudes in the opposite direction of what was
predicted.

Linear regression of attitudes on benevolent success for _l_o_vy value expressives in

the Western sample yielded a statistically significant standardized slope (B = .40), t_=

 

2.51,_p < .02. Additionally, high value expressives yielded an insignificant standardized
slope (B = .12), 1= .69, ns. The difference between the unstandardized slopes was not
significant, 3(31) = .50, ns. The data are not consistent with the prediction because the
slopes were not negative and the impact of power-tradition did not differentially impact
attitudes. Table 4 presents the obtained means and standard deviations for power-tradition

and attitudes toward interracial marriage.

55

Table 4. Means and Standard Deviations for the regression of attitudes
on the power-tradition scale for low value expressives and high value expressives

 

 

Variable Low Value Expressive High Value Expressive
Midwest M = 4.77 M = 5.49
Power-trad. SD = 1.40 SD = 1.31
N= 60 N = 51
Attitude M = 4.96 M = 5.79
SD = 1.40 SD = .97
West M = 4.75 M = 4.68
Power-trad. SD = 1.40 SD = 1.09
N= 35 N= 34
Attitude M = 5.25 M = 5.63
SD=1.37 SD=1.14

H3d: Varied life. It was predicted that varied life (stimulation) would be
positively associated with attitudes toward interracial marriage, and that this relationship
would be stronger for high value expressives.

Linear regression of attitudes on varied life for_lo_w value expressives in the
Midwestern sample yielded a statistically significant standardized slope (B = .46), t =
3.90, p < .000. Additionally, high value expressives yielded a insignificant standardized
slope (B = .23), t: 1.62, ns. The difference between the unstandardized slopes was
insignificant, t(56) = 1.81, p < .05. The data are partly consistent with the prediction
because the slopes were positive and significantly different than zero. However, the
impact of varied life for low and high value expressives did not differentially impact

attitudes in the opposite direction of what was predicted.

56

Linear regression of attitudes on varied life for low value expressives in the

Western sample yielded a statistically significant standardized slope (B = .48), t_= 3.18, p

< .003. Similarly, high value expressives yielded a significant standardized slope (B =

.43), F .2.67, p < .01. The difference between the unstandardized slopes was not

significant, _t_(30) = .29, ns. The data are partly consistent with the prediction because the

slopes were positive and significantly different than zero. Second, the impact of varied

life for low and high value expressives did not differentially impact attitudes. Table 5

presents the obtained means and standard deviations for varied life and attitudes toward

interracial marriage.

Table 5. Means and Standard Deviations for the regression of attitudes
on the varied life scalp for low value expressives and high value expressives

 

 

 

Variable Low Value Expressive High Value Expressive
Midwest M = 5.98 M = 6.20
Varied life SD = 1.36 SD = 1.20
N= 60 N = 51
Attitude M = 4.96 M = 5.79
SD = 1.40 SD = .97
West M = 5.71 M = 5.59
Varied life SD = 1.49 SD = 1.23
N = 35 N = 34
Attitude M = 5.25 M = 5.63
SD=1.37 SD=1.14

 

57

H3e: Honoring parents. It was predicted that the effect of honoring parents value
would be negatively related to interracial marriage attitudes, and that this relationship
would be stronger for high value expressives. The standardized slope for low value
expressives was (B= .53), F- 6.05 , p < .000. The standardized slope for high value
expressives was (B = .24), t = 2.20, p < .03 . The difference between the two
unstandardized slopes was significant, t(91)= 2.61,p < .05.

The data are not consistent with the prediction that the regression slopes of
honoring parents would be substantial and negative, instead the slopes are significant and
positive. Also, the effect of honoring parents on attitudes was significantly different
between the two groups, but the difference was in the opposite direction of what was
predicted. So, the effect of honoring parents on attitudes was stronger for low value
expressives than for high value expressives. Table 6 presents the obtained means and

standard deviations for honoring parents and attitudes toward interracial marriage.

Table 6. Means and Standard Deviations for the reggession of attitudes
on the honoring parents scale for low value expressives and higl_r value expressives

 

 

Variable Low Value Expressive High Value Expressive
Honoring M = 4.99 M = 5.71
parents SD = 1.91 SD = 1.58
‘ N= 95 N =85
Attitude M = 5. 07 M =5.73
SD=1.39 SD=1.04

 

58

H3f: Social order. It was predicted that social order (conservation) would be
negatively associated with attitudes, and that this relationship would be stronger for value
expressives. Linear regression of attitudes on social order for low value expressives
yielded a statistically significant standardized slope (B = .35), _t_= 3.60, p< .000.
Additionally, high value expressives yielded a insignificant standardized slope (B = .13),
fr— 1.16, ns. The difference between the unstandardized slopes was significant, 1(90) =
1.9, p < .05. The data are not consistent with the prediction because the slopes were
positive, and for higher value expressives the beta was not significantly different than
zero. Additionally, the impact of social order, for low and high value expressives, did
differentially impact attitudes, but the difference was in the opposite direction of what
was predicted. Table 7 presents the obtained means and standard deviations for social

order and attitudes toward interracial marriage.

Table 7. Means and Standard Deviatiflrs for the regression of attitudes

on the social order scale for low value expressives and higl_r value expressives

 

 

Variable Low Value Expressive High Value Expressive
Social order M = 5.21 M = 5.45

SD= 1.70 SD= 1.53

N= 94 N =83
Attitude M = 5. 09 M =5.72

SD= 1.38 SD= 1.05

 

59

H4: Structure. The values coded in order to test for these hypotheses reflect the
6-factor model described above, and they structured in the same manner as hypothesis 3
was.

H4a: Egualig. It was predicted that as attitudes toward interracial marriage
increased so would the amount of equality statements expressed, and this relationship
would increase for high value expressives. Linear regression of equality values on
attitudes for low value expressives yielded a statistically insignificant standardized slope
(B = .34), t_= .77, ns. Additionally, high value expressives yielded a insignificant
standardized slope (B = .07), tr- .63, ns. The difference between the unstandardized
slopes was not significant, :08) = .28, ns. The data are not consistent with the prediction
because the slopes are not substantially greater than zero and there is no differential
impact on equality message production between high and low value expressives. Table 8
presents the obtained means and standard deviations for attitudes toward interracial

marriage and equality messages.

Table 8. Means and Standard Deviations for the reggession of egualig messages and
attitudes for low value expressives and high value expressives

 

 

Variable Low Value Expressive High Value Expressive
Equality M= .61 M=2.14

SD = 5.52 SD = 9.36

N= 82 N = 74
Attitude M = 5. 14 M =5.72

SD=1.38 SD=1.04

 

60

H4b: benevolent success. It was predicted that as attitudes toward interracial
marriage increased so would the amount of benevolent success statements expressed, and
this relationship would increase for high value expressives. Linear regression of
benevolent success values on attitudes for low value expressives yielded a statistically
significant standardized slope (B = .22),; = 2.02, p_< .05. Additionally, high value
expressives yielded a insignificant standardized slope (B = .06), _t= .53, ns. The
difference between the unstandardized slopes was not significant, 1(78) = .63, ns. The
data are not consistent with the prediction because the slopes for value expressives is not
substantially greater than zero and there is no differential impact on benevolent success
message production between high and low value expressives. Table 9 presents the
obtained means and standard deviations for attitudes toward interracial marriage and

benevolent success messages.

Table 9. Means and Standard Deviations for the reggession of benevolent success
mesgges and attitudes for low value expressives and high value expressives

 

 

 

Variable Low Value Expressive High Value Expressive
Ben-success M = 22.39 M = 31.26

SD = 30.07 SD = 33.12

N: 82 N = 74
Attitude M = 5. 14 M =5.72

SD=1.38 SD=1.04

 

61

H4c: Power-tradition. It was predicted that as attitudes toward interracial
marriage decreased the amount of power-tradition statements expressed would increase,
and this relationship would increase for high value expressives. There was insufficient
data to test this hypothesis. Linear regression was only run on the high value expressive

group, (B = .15), t = 1.25, ns. M = .34, SD= 2.91).

_Had: Varied life. It was predicted that as attitudes toward interracial marriage
increased the amount of varied statements expressed would also increase, and this
relationship would increase for high value expressives. There was no data available to
test this hypotheses. Participants did not list any thoughts that were coded as honoring

parents.

H4c: Honoring parents. It was predicted that as attitudes toward interracial
marriage decreased the amount of honoring parents statements expressed would increase,
and this relationship would increase for high value expressives. There was no data
available to test this hypotheses. Participants did not list any thoughts that were coded as

honoring parents.

H4f: Social order. It was predicted that as attitudes toward interracial marriage
decreased the amount of social order statements expressed would increase, and this
relationship would increase for high value expressives. There was no data available to

test this hypotheses. Participants did not list any thoughts that were coded as social order

62

CHAPTER 5

DISCUSSION

Overview

The present study sought to investigate the impact of the value expressive
function of attitudes on the relationship among values, attitudes and subsequent message
production pertaining to the domain of black - white interracial marriage. Firsthas
predicted, subjects’ were coded as having significantly more expressive thoughts than
evaluative thoughts regarding their attitudes toward interracial marriage. Second, the
other-directedness subscale was not negatively correlated with value expressiveness, and
extroversion and the self-monitoring subscales of sensitivity to expressive behavior and
ability to modify self-presentation were not positively correlated with the value-
expressive measure as had been predicted. Third, six new value factors were derived
(equality, benevolent success, power-tradition, social order, honoring parents, and a
varied life), however, they were not consistent with Schwartz’s value theory (1992). In
general, regression analysis demonstrated sizable regression coefficients between each
value factor and attitude toward interracial maniage as predicted. Although statistically
insignificant for the majority of values and samples, the impact of values on attitudes
toward interracial marriage was higher for low value expressives than for high value
expressive across the six value factors, and this was contrary with the predicted
hypotheses. In terms of message production, attitudes toward interracial marriage did not
affect appeals to values of equality and benevolent success nor was there a difference
between low and high value expressive individuals as had been predicted. Finally,

subjects did not use the expected power-tradition, social order, honoring parents or varied

63

life value appeals when constructing persuasive messages designed to reflect their

attitudes toward interracial marriage.

The expressive function and attitudes. It was demonstrated that the majority of

attitudes toward black-white interracial marriage were based, on an expressive function

.’ ' ,
1 ,r‘ i. i
o L, I J
‘ J 9 f 2

i
1 ‘r ’ir‘.‘1 .
' .1 1,; l in", I I"

(social or value) rather than an evaluative functronJOrily orie thloriightwas coded as ego- ‘ .
defensive, suggesting that participants were not extremely opposed to interracial
marriage. Priorresearchonattitudes toward interracialmarriage have. reflected negative
perceptipiisfiontemps, 1975; Paset and Taylor, 1991; Porterfield, 1984, Tucker and
Mitchell-Keman, 1990), and claims that the black-whiteracial. combination has provoked
the strongest responses. The present results do not demonstrate this pattern. Mean
attitude was moderately high (m=5.36) illustrating a positive evaluation for black-white
interracialfimarriage. Additionally, attitudes did not significantly differ for the Midwest
or the Western samples. These results are encouraging because they suggest a more
tolerant view of black-white marriages in both of these-'iiniversity student‘bgdies,‘

The predominance of expressive fimctions underlying this attitude domain is in
line with previous work claiming that attitudes involving marginalized groups. stem. from
Protestantwork ethic values, racial identity factors, and families’ and friends’ perceptions
(Davidson & Schneider, 1992; Herek, 1984; Kinder & Sears, 1981). In addition, Herek
(1986) suggests that symbolic attitudes (such as interracial marriage) best serve the
expressive function. However, the focus of these studies was opposition whereas this

study clearly illustrates a positive evaluation toward interracial marriage. This study

extends prior research because it demonstrates that value expressive and social expressive

fimctions underlie positive perceptions of interracial marriage. Attitudes such as these
can be best understood by examining group identifications and values. The present study
focused on the impact of values and personality variables associated with value

expressiveness.

_S_algrponitoring subsgles. DeBono (1987) has suggested that people high in
value expressiveness are likely to pay less attention to social cues and more attention to
internal beliefs and values, much like individuals who are low self-monitors. Similarly,
other researchers (Kristiansen and Zanna, 1988; Maio and Olsen, 1994) have used the
concept of self-monitoring as an indirect measure of value expressiveness. A review of
the results using Snyder’s (1979) self-monitoring construct revealed that the subscales of
other-directedness and extroversion accounted for the differential correlations between
high and low value expressives and attitudes. Additionally, other—directedness and
extroversion were excluded from the revised self-monitoring scale, which focused on the
subscale sensitivity to self-expression and ability to modify self-presentation.

The predicted negative correlation between other-directedness and value
expressiveness was trivial and insignificant. Similarly, the correlations among value
expressiveness and extroversion, ability to modify self-presentation and sensitivity to
expressive behavior were all trivial and nonsignificant. Therefore, these results fail to
demonstrate that any of Snyder’s self-monitoring subscales are linearly associated with
value expressiveness. In addition, the revised self-monitoring subscales (Lennox and
Wolfe, 1984) are also not linearly associated with value expressiveness. Thus, anyone

interested in examining value expressiveness should not use the self-monitoring variables

65

as indirect indicators of value expressiveness. They are poor indicators, and it would be
misleading to use them to distinguish high and low value expressives. Further research
should concentrate on validating direct measures of value expressiveness, such as content

analysis.

Schwartz’s universal values. Schwartz’s value theory (1992) was utilized
because it proposed to capture 10 motivational dimensions. This study, however, only
examined nine dimensions, excluding hedonism. Confirmatory factor analysis yielded a
6—factor solution which was inconsistent with the 9-factor theory. The six value factors
validated were: equality, benevolent success, power-tradition, varied life, honoring
parents, and social order. The CFA results contradict prior research (Maio and Olsen,
1995; Schwartz, 1992; 1996; Schwartz and Bilsky, 1987; Sagiv and Schwartz, 1995) that
had content validated the theory using Guttman-lingoes smallest space analysis. Aside
from the more stringent CFA method used here there are two possible explanations that
can account for the failure to replicate the nine distinct value types.

First, CFA interfactor correlations (among the 6 value factors) were positive and
substantial in size. They did not indicate any negative (opposing) associations, as was
predicted by Schwartz. For example, the benevolent success factor combines the lower
order benevolence domain with the self-direction and achievement domains.
Theoretically, benevolence and self-direction are compatible with each other because
they reflect an open-minded and altruistic tendency, but achievement was predicted to
conflict because of its motivation for self-interest and competence according to social

standards. It appears that all the value domains were highly interrelated and thus cannot

66

be distinguished into nine distinct value dimensions, thus there is no pattern of conflict
observed here. The individuals in this study reported that they pursue multiple
motivational goals at the same time. Thus, their attitudes toward interracial marriage
were guided by values reflecting six different motivational concerns that did not compete
against each other.

Second, subjects in this study were asked to rate the relevance of each value to
their attitude toward interracial marriage. In contrast, Schwartz’s prior research had
subjects rate the general importance of each value as a guiding principle in their lives.
Schwartz’s method is a broader method because it is not presented within a specific
context. This study measured value orientation in the context of interracial marriage, and
asked subjects to rate how instrumental each value was in promoting that attitude.
Consequently, the value factor structure is domain specific and less likely to reveal the
predicted basic universal value structure. Further research should determine whether this
6-factor solution can be replicated within the domain of interracial interactions.
Additionally, the utility of using the relevance method over the general importance rating
will help determine if there are specific values connected to the domain of interracial
marriage. Aside fiom factor analysis issues, understanding the role of the value
expressive function on the relationship between values and attitudes was one central

focus of this study.

Values and attitudes. Overall, the results of this study are consistent with the

predictions that values impact attitudes substantially. However, the size of the effect

differed depending on the specific value factor as well the degree of value

67

expressiveness. Six separate regression runs were made, one for each value factor. Since
the values were shown to be highly interrelated any analysis combining all six value
factors would have been attenuated by the severe multicollinearity among all six
predictors. Each value result (H3a-f) will be briefly discussed and will be followed by
two alternative explanations for the failure to identify a moderating effect.

The equality factor was just a l-item measure of the equality value. It was found
that the effect of the equality value on the attitude towards interracial marriage for low
and high value expressives was positive and substantial, thus in line with hypothesis 3a.
However, the moderating role of value expressiveness was not supported. Low and high
value expressive individuals did not significantly differ on the extent to which their
equality value promotes their attitude toward interracial marriage. Thus, values and
attitudes are significantly related, however low and high value expressive subjects do not
differ on the extent to which equality differentially impacts their attitudes toward
interracial marriage. Participants’ views of interracial marriage are relevant to their

values of tolerance and “general! acceptance of others. Interracial marriage may be seen as;

4

a vehicle for expressing equality and acceptance of all others. '
Benevolent success value analyses were conducted separately for the Midwestern
sample and the Western sample, and the results differed marginally. It was found that the
effect of benevolent success value on the attitude towards interracial marriage for both
low and high value expressives was positive and substantial, thus in line with hypothesis
3b. However, the moderating role of value expressiveness was not supported. Low and
high value expressive individuals did not differ significantly on the extent to which their

benevolent success values promotes their attitude toward interracial marriage. Thus,

68

values and attitudes are significantly related, however low and high value expressive
subjects do not differ on the extent to which benevolent success values differentially
impact their attitudes toward interracial marriage.

Hypothesis 3b was not consistent with the data from the Western sample in two
ways. First, the impact of values on attitudes was significant for low value expressives
only. Also, the moderating role of value expressiveness was not found. These results
suggest that for the Western sample, the impact of values on attitudes toward interracial
marriage is substantial and positive, but only for low value expressives. It is important to
note that the Beta was .31 for the high value expressives, although insignificant it still
reveals a positive relationship. In general, participants are viewing interracial maniage as
relevant to benevolent success values which include mature love, success and self-
direction. Interracial relationships are perceived as situations which do not hinder but
may heighten one’s personal success, opportunities for individual growth or opportunities
to promote the welfare of close others.

Similar to benevolent success, analyses involving power tradition values were
conducted separately based on sample location. Results were not consistent with the H3c
prediction in that the impact of power tradition values on attitudes was positive and
significant for both Midwest value expressive groups and for the low value expressive
Western group. Additionally, the moderating effect was not found for either sample.
These results do not support Schwartz’s prediction that values with self-interest and
status quo motivations are negatively related to attitudes regarding intergroup contact
(Sagiv & Schwartz, 1995). Instead, various types of values, including conservative ones,

positively impacted attitudes toward interracial marriage for the majority of value

69

expressives in both samples. Additionally, power tradition values did not differentially
impact attitudes based on level of value expressiveness. These results demonstrate that
power-tradition positively effects attitudes for the majority of samples and that the effect
is consistent across low and high value expressives. Participants may not view interracial
marriage as a threat to any traditional customs they hold about love or marriage, or they
may view interracial marriage as a vehicle to assert their interpersonal power over a
partner of mfenor status and thus viewinterracial relationshipsas an. opportunity for
gaining the power that theyvalue (Sagiv & Schwartz, 1995).

Analyses conducted on the stimulation single item value measure yielded results
which was partially consistent with the H3d prediction in that the impact of varied life
value on attitudes was positive and significant for both Western value expressive groups
and for the low value expressive Midwestern group. Additionally, the moderating effect
was not found for either sample. Participants are viewing interracial marriage as a

1

relationship thatcan result innovelty and excitement which is relevant to their individual

l
l
I
‘r

autonomyandpersonal. growth. ..

The honoring parents value was inconsistent with the H3e prediction because the
effect of the conformity value was positive and substantial for both low and high value
expressives. Also, a moderating effect was not demonstrated. Similar to the positive
relationships found in the power tradition results, honoring parents is a conservative
value that was expected to negatively impact attitudes. The results, however,
demonstrate the for the domain of interracial marriage multiple motivational goals of
values positively impact attitudes and that the effect is consistent across levels of value

expressiveness. For example, participants’ parents may endorse interracial marriage and

70

if they do, then, higher conformity with parents would be associated with increased
attitudes toward interracial marriage.

The results utilizing the social order data were inconsistent with the H3f
prediction because the social order value did not negatively impact attitudes and no
evidence of a moderating effect was demonstrated. Once again, results showed that a
conservative security based value type positively impacted attitudes for the low value
expressive group and that this effect was consistent across levels of value expressiveness.
Participants in this study are equating a stable society with a society that tolerates

interracial marriages. In other words, these subjects did not view interracial marriage as a i

.‘l

threat to their social order or to their stability within their social structure. ;
In sum, results across all six value types showed a significantly positive
relationship between value and attitude regardless of motivational basis. There was no
evidence of a moderating effect in the predicted direction specified, instead positive
relationships were relatively consistent across levels of value expressiveness.
Additionally, separate analyses between Midwest and Western samples mirrored this
pattern. This provides evidence that 1) attitude domains may have context-specific
values that underlie them which do not reflect universal motivational concerns, 2) value
expressiveness does not function as a moderating variable between values and attitudes,
and 3) although statistically significant for the social order, honoring parents and the
Midwestern samples of varied life and power-tradition, there was a trend showing that
low value expressives yielded higher regression slopes than high value expressives across

all other values. Alternative models will be presented but before that discussion of post

hoc analyses, hypothesis 4a-f predictions must be described.

71

Attitudes and value sgcific mesgges. Similar to the structure of hypothesis 3,
results for hypothesis 4 are organized by value dimension. Analyses conducted on the
equality messages (H4a) yielded results that were in the predicted direction but
nonsignificant. Also, a moderating effect was not demonstrated. Thus, subjects’ equality
messages were not substantively impacted by their attitudes towards interracial marriage
nor did this effect differ across levels of value expressiveness.

Benevolent success results were partially consistent with the H4b prediction
because the low expressive group demonstrated a positive relationship between attitudes
and subsequent benevolent success messages, however this was not the case for high
value expressives. Also, comparison of the unstandardized slopes revealed that there was
no differential impact on benevolent success message production between low and high
value expressives. It is important to note that specific value messages are being
manifested as a result of participants’ attitudes and this is only apparent with the low
value expressives. So, it appears that level of value expressiveness is not driving
message production.

Finally, the data for power-tradition, stimulation, honoring parents, and social
order was so severely limited that analyses could not be conducted. Power-tradition
messages were only obtained for high value expressives and the regression slope,
although positive was not significantly different than zero. Subjects in this study did not
write messages using stimulation, honoring parents, or social order values.

In sum, results from hypothesis 4 data demonstrate that the impact of attitudes
toward interracial marriage on message production is positive for benevolent success, but

that it is restricted to individuals with low levels of value expressiveness. In addition, the

72

co:

t3

moderating role of value expressiveness was not extended to message production for any
of the value domains. Thus, the communicative nature of the value expressive function
was not established here. Prior functional research by definition predicts that individuals
who hold a value expressive function will express values tied to their attitudes as a
vehicle to align oneself with specific reference groups or to enhance self-concept by
virtue of mentioning specific values that are central to them. The study shows that
although the impact that values have on attitudes is substantially positive the effect for
message production is limited and no difference in message production can be attributed
to varying levels of value expressiveness. Thus, an attitude’s existence was not
dependent on the elaboration of specific values. Action Assembly theory did not
adequately explain the data because subjects holding a high value expressive function did
not significantly reflect values in their messages so no evidence of a procedural record
for value expressiveness was established.

The central interest of this paper was to explore the predicted moderating effect of
the value expressive function. This study demonstrates that value expressiveness does
not moderate the relationship between values and attitudes in the specified predicted
direction. However, a pattern demonstrating the opposite relationship emerged. Results
found a trend, albeit statistically insignificant, such that individuals in the low value
expressive group yielded stronger value-attitude relations than the individuals in the high
value expressive group, and this pattern occurred across all value dimensions. This result
contradicts all prior research investigating the utility and viability of the value expressive
function. Further interpretation of the results is required, however, before any conclusion

can be drawn about the nature of the value expressive fimction.

73

There are two possible explanations for the absence of the predicted moderating
effect. First, the basement effect (restriction in range) for the value expressiveness
variable most likely attenuated the impact of values for high value expressives. The
median for value expressives was m_d= 37.5 with a mean of 36.5 with a distribution that
was skewed to the right. Descriptive statistics of each subsample were also calculated (
low VE group = p_r= 14.23, m_d= 16.67, range 0- 37.5, §c_1= 13.15; while m= 60.43, gap:
50, range 40-100, ad= 18.93 for the high VE group). These descriptive statistics
demonstrate that the scores for the low value expressive group are more centrally
distributed about the mean; while the high value expressive group scores predominantly
lie on the lower end of the range, thus decreasing variability. Therefore, one cannot
conclude that a moderating effect, as predicted, does not exist; rather this data
demonstrates that the individuals in this study generated moderate levels of value
expressiveness that restricted the range of the overall distribution. This restriction‘may
have been masking a true moderating effect. Replicating this predicted relationship, with
normally distributed data, will help resolve the ambiguities.

On the other hand, specificity error points to alternative models that can
accurately portray the relationships among values, the value expressive function attitudes
and subsequent message production. Post hoc analyses, integrating values and message
production were conducted across all six value dimensions. In some cases path analysis
was used because data were available for all variables while in other cases multiple

regression was utilized.

74

Post-hoe analyses. Results of this study did not confirm a moderating effect for
value expressiveness so it was decided to investigate other predictions by transforming
the variable of value expressiveness into a continuous variable instead of having it remain
dichotomized. Prior regression analysis and scatter plots revealed positive linear
associations between the majority of the six values and attitudes. Also value
expressiveness was positively correlated with the attitudes and to the majority of the
values. Hence, the general pattern examined involved a two predictor model where each
specific value factor and value expressiveness simultaneously predict attitudes toward
black-white interracial marriage which then predicts subsequent message production of
that value (when data available). Each value analysis is discussed below. Appendices H
through P contain the correlation matrices of the following models.

The resulting equality path model, with path coefficients is presented in Figure 2.

 

Equality .39
.09

20 Attitude —— Equality messages

VE/

 

Figure 2 The equality path model

The fit of the model was assessed in two ways, first predicted values were

generated fiom both correlations not constrained to equal their obtained value to see if the

predicted values were within sampling error of those obtained. Both residuals between

75

equality and equality messages and value expressiveness (VET) and equality messages
were small, .01 and .05 respectively, and within sampling error of zero. Second, chi
square was employed as a global measure of fit. This model yielded a small and
insignificant chi-square MR2) = .29, p < .05) .Multiple regression analysis regressing
attitudes on equality value and value expressiveness showed that these variables did
significantly impact attitudes toward interracial maniage (R = .47, F(2, 176) = 24.97, p <
.000). The path from the equality value to attitude was significant t(176) = 5.72, p < .00,
and so was the path fi'om value expressiveness t(176) = 2.94, p < .003. Finally, the path
from attitude to equality message was .09 and was not statistically significant. The
independent additive effects of value expresssiveness and the benevolent success values
has been established, but the subsequent mediating impact of attitudes onto message
production cannot be accepted because of the extremely low parameter between attitude
and equality message.

The resulting benevolent success path model, with path coefficients is presented
in Figure 3. The Midwestern coefficients are presented without parentheses and the
Western sample’s coefficients are in parentheses. Both residuals between benevolent
success and benevolent messages and value expressiveness (VET) and benevolent
success messages for the Midwest sample and for the Western sample were small were
small, -.01 , and .18 and .06 and -.09 respectively, and within sampling error of zero.
Second, chi square was employed as a global measure of fit. This model yielded a small
. and insignificant chi-square (12(2) =1 .80, p < .05 for the Midwest and 3(2) =.45, p < .05
for the West) . Multiple regression analysis regressing attitudes on benevolent success

values and value expressiveness showed that these variables did significantly impact

76

attitudes toward interracial marriage (R = .56, F(2, 108) = 25.21, p < .000 for Midwest
and R = .53, F(2, 66) = 12.93, p < .000 for West). The path from the benevolent success
value to attitude was significant (B= .48, t(108) = 5.72, p < .00, and B= .49, t(66)= 4.66,
p < .003 for the West) so was the path from value expressiveness (B= .21, t(108) = 2.53,
p < .01 for the Midwest and B = .13, t(66) = 1.26, p < .01 for the West). Finally, the path
from attitude to benevolent success message was .37 (statistically signifcant, p < .000) for
the Midwest and -.17 (not statistically signficant) for the West. Hence, the model for the
Midwest sample is consistent with the data, but the Western model cannot be accepted
due to the small negative correlation between attitude and benevolent success messages

which is not significantly different than zero.

 

48
Benevolent succes
(fish 37

Attitude TIT)_ Benevolent success messages
.20 '

VE J3)

 

Figure 3 The benevolent success path model for the Midwestem and Western
samples.

The resulting power-tradition path model, with path coefficients is presented in
Figure 4. The Midwestern coefficients are presented without parentheses and the
Western sample’s coefficients are in parentheses. The western sample only utilized

multiple regression because no data was available for power-tradition messages. In the

77

case of the Midwest sample both residuals between power-tradition and power-tradition
messages and value expressiveness (VET) and power-tradition messages were small, -
.02, and .12 respectively, and within sampling error of zero. Second, chi square was
employed as a global measure of fit. This model yielded a small and insignificant chi-
square (33(2) = .78, p < .05 for the Midwest. Multiple regression analysis regressing
attitudes on power-tradition values and value expressiveness showed that these variables
did significantly impact attitudes toward interracial marriage (R = .47, F(2, 108) = 14.96,
p < .000 for Midwest and R = .33, F(2, 66) = 12.93, p < .000 for West). The path from
the power-tradition value to attitude was significant (B= .34, t(108) = 3.81, p < .00, for
the Midwest and B= .26, t(66) = 2.21, p < .003 for the West) so was the path from value
expressiveness (B= .25, t(108) = 2.79, p < .01 for the Midwest and B = .18, t(66) = 1.50,
p < .01 for the West). Finally, the path from attitude to power-tradition message was .13
and not statistically significant for the Midwest. Hence, the model establishes that value
expresssiveness and power-tradition values additively impact attitudes. However, the
mediating role of attitudes onto power-tradition messages cannot be confirmed because
of the low correlation between the two variables. Subjects in the Western sample did not

write any power-tradition messages.

78

 

Power-tradition

 

24 Attitude Power-tradition messages

/ (not available)
VE (-17)

(.17)

 

Figure 4 The power-tradition path model for the Midwestern and Western samples

The resulting varied life multiple regression model with beta coefficients is
presented in Figure 5. The Midwestern coefficients are presented without parentheses and
the Western sample’s coefficients are in parentheses. Multiple regression analysis
regressing attitudes on varied life values and value expressiveness showed that these
variables did significantly impact attitudes toward interracial marriage (R = .49, F (2, 108)
= 17.1, p < .000 for Midwest and R = .48, F(2, 66) = 10.03, p < .000 for West). The path
from the varied life value to attitude was significant (B = 36, t(108) = 4.28, p < .00, and
B= .44, t(66) = 4.02, p < .001 for the West) but the path from value expressiveness to
attitude was only significant for the Midwest sample (B= .31, t(108) = 3.69, p < .004 for

the Midwest and B = .18, t(66) = 1.66, us for the West).

79

.36

Varied Life
M

30 Attitude

 

 

Figure 5 The varied life multiple regression model for the Midwestern and Western
samples

The resulting honoring parents multiple regression model with beta coefficients is
presented in Figure 6. The Midwestern coefficients are presented without parentheses and
the Western sample’s coefficients are in parentheses. Multiple regression analysis
regressing attitudes on the honoring parents value and value expressiveness showed that
these variables did significantly impact attitudes toward interracial marriage (R = .50,
F(2, 177) = 29.18, p < .000. The path from the honoring parents value to attitude was
significant (B = .42, t(177) = 6.25, p < .00). Also the path from value expressiveness to

attitude was significant (B= .21, t(177) =3.12, p < .002).

 

. .45
Honorrng Parents

V
VE

Attitude

 

Figure 6 The honoring parents multiple regression model

80

The resulting social order multiple regression model with beta coefficients is
presented in Figure 7. The Midwestern coefficients are presented without parentheses
and the Western sample’s coefficients are in parentheses. Multiple regression analysis
regressing attitudes on the honoring parents value and value expressiveness showed that
these variables did significantly impact attitudes toward interracial marriage (R = .40,
F(2, 177) = 14.64, p < .000. The path from the social order to attitude was significant (B
= .26, t(177) = 3.75, p < .00). Also the path fi'om value expressiveness to attitude was

significant (B= .26, t(177) =3.73, p < .000).

 

Social Order .41
\ Attitude
'21//
VE

 

Figure 7 The social order multiple regression model

Conclusion

With the exception of the varied life results in the Western sample, the two
predictor model of values and value expressiveness predicting attitudes toward black-
white interracial marriage was consistent with the data Thus, these results demonstrate
that values and value expressiveness have an additive effect on attitudes instead of a
multiplicative effect as was previously hypothesized. Post hoc results yielded positive

value-attitude correlations, and across all six value models the value-attitude relation was

81

stronger than the value expressive (V E)-attitude relationship. This positive association
between all values and attitudes toward interracial marriage is consistent with main
analyses results, indicating that values do significantly impact attitudes. The impact of
value expressiveness is secondary to that of values, but it nevertheless an important
variable that contributes to the prediction of attitudes toward interracial maniage.

In addition, the paucity of message appeal data in addition to the reduced sample
sizes, due to significant differences among the Western and Midwestern samples, limited
the extent to which models extending the process to actual message production could be
empirically supported. Therefore, Action Assembly theory does not adequately explain
the data from these analyses either. The claim that the value expressive function acts as a
procedural record which facilitates use of values in persuasive appeals is not supported
here. In fact, participants in this study did not activate the use of value appeals, and this
was reflected in the minimal data obtained. The resulting correlations between attitudes
and specific value message appeals were predominantly low and statistically insignificant
(equality, power-tradition appeals, and benevolent success appeals for the Western
sample). The one exception was the Midwestern sample (N = 111) for their use of
benevolent success message appeals. This model demonstrated how attitudes mediate the
relationship between value expressiveness, benevolent success values and resultant
message production of benevolent success value appeals.

Prior research has likened the value expressive variable to an individual
difference variable, such as self-monitoring which is often dichotomized. Although
theoretically compelling, this line of thinking does not fit the data presented here. It has

been demonstrated that value expressiveness is positively related to attitudes but it does

82

not differentially impact the relationship between values and attitudes. Value
expressiveness was more appropriately found to be a joint exogenous predictor of
attitudes. The consistency of this pattern is reflected in both Western and Midwestern
samples and across all value domains, thus lending support to the validity of this
relationship.

This research has extended past work on fimctional theory because it has clarified
the role of the value expressive function with respect to values and attitudes toward
interracial maniage. Additionally, this research proposes an alternative mediating role
which attitudes may serve that can explain how the cognitive internal processes of values
and value expressiveness may indirectly affect more overt message behavior. Results
have shown how benevolent success values and the value expressive function have

communicative consequences. By definition the value expressive function requires that -- , ‘,

A , l‘
. a

one’s attitudes exist because they allow individuals to express relevant values, however, in: J .. ,
this definition has not been empirically supported.» This study has introduced one model
that may accurately depict the overall relationship among values, the value expressive
fimction, attitudes and message behavior.

F m research. The value of this study is that it provides direction for future
research about the persuasive properties tied to interracial interactions. Other domains
not pertaining to issues relating to intergroup interactions, whether romantic or platonic,
may not be suited for the value expressive function or for the specific values reported
here. First, results of this study suggest that the moderating role has not been clarifed so it
is suggested that future work in this area clarify the role of the value expressive fimction

by testing two competing models: an moderating efl‘ect (in the predicted direction), and

83

the joint additive effect. Large sample sizes and normally distributed value expressive
data should be carefully obtained.

Second, the values of equality, benevolent success and power-tradition have i “1
been positively linked to interracial interaction attitudes and should thus be utilized agairi
in future work involving race or intergroup relations. Past research has reinforced the
utility of the value expressive function when investigating symbolic attitudes. More
importantly, however, is determining the extent to which our attitudes, if grounded in
value expressiveness, impact our persuasive communicative behavior.

Third, the functional approach to attitudes is useful because it can identify certain
motivational bases for attitudes, and in theory can predict change if the motivational basis
is met. The assumption here is that the attitude formed by the attitude function will
somehow affect behavior, if not why should the attitude be altered at all? There has not
been, however, data explicating any connection between attitude functions and actual
communicative behavior. The value expressive frmction specifies a communicative
component to attitude maintentance and development. Therefore, it is essential to
determine if attitudes, grounded in value expressive function, can be used as
communicative tools to persuade others. The current models are impetus for such work.

Finally, an alternative model that depicts the value expressive function as a
process is suggested. Prior research has measured the value expressive variable (either
indirectly or directly) in order to determine how this variable impacts values and
attitudes. An alternative approach would not treat the value expressive function as a

variable but would show how values, attitudes and expresssion of value messages are

related for individuals who are value expressive. It can be argued that not all attitudes are

84

value driven, but some are. This study has demonstrated that in the case of interracial
marriage values are very important and ar tied to attitudes. If that is the case, then, these
individuals would also be more likely to express their relevant values in messages they
construct. It then follows that the relationship between attitudes toward interracial
marriage and expression of values in a message is driven by important values, and this
would be the case for individuals who are value expressive only. This pattern would not
hold for the other functions. That is, the correlation between attitude and expression of
value messages is spuriously driven by values. Figure 8 presents this process model.

Future research should test this model for it’s accuracy.

 

Attitudes Expression

 

Figure 8 The process model

Methodological issues. There are two methodological issues that have
contributed to isolating the relationship among values, value expressiveness, attitudes and
message behavior. Participants rated how relevant each value was in promoting their
attitudes toward interracial marriage. This was a specific measure unlike past research
which has asked participants to rate the general importance of values as guiding life
principles. The specificity of the measure would, then, be more likely to capture the true

relationships between values, value expressive function, attitudes and message appeals

85

because the likelihood of subjects utilizing value appeals that they have previously
identified as being relevant to their attitude is greater than using generalized value
priorities without any frame of reference (attitude domain).

Second, content analysis provided rich information on individuals’ thought
processes underlying their attitudes toward interracial marriage. Prior research has used
close-ended scale items to measure attitude function. This method assumes that
individuals are consciously aware of the motivational basis for their attitudes. Subjects
are required to interpret their attitudes and evaluate them against a limited number of
alternatives, and this can be a nearly impossible task. Content analysis, in contrast,
provides researchers with specific indicators of the participant’s mental representation
surrounding their attitudes. The open-ended format allows individuals the fi'eedom to
focus on the salient issues triggered by the attitude domain. This information can provide
insight on motivational concerns that contribute to the attitude’s maintenance and
resiliency. Further research on attitude functions should use content analysis as a
measure of attitude functions because it will provide richer, more detailed information,
and it is the only paper-pencil method that can directly measure actual communicative

behavior.

86

Authors Notes

‘ The design of this study specified that a subset of 31 subjects were to be used for
establishing reliability and clarifying coding scheme issues only. It is argued that since
Kappa’s were not all 1.00 these data had to be reviewed by the coders in order to clarify
the remaining discrepancies before one coder could begin coding the main sample. The
subset data, then, was never meant to be included in the main sample because it would
not reflect the same coding scheme accuracy as the subsequent main sample did. Issues
that were problematic in the subset had been resolved for the main sample (n= 185).
Alternatively, the coder did not go back and recode the subset because prior familiarity
with those messages would have tainted the a priori coding scheme measure.

Additionally, many argue that excluding the subset will reduce statistical power
needed to detect significant differences. The more important point, however, is whether
detectable effect size, when power is set at .80, is reduced by the drop in sample size. In
other words, the question to be answered should be, what must the effect size be if power
is set at .80. In this study (N= 185) most analysis were broken down into 4 groups (hi
value expressive VE/MI (midwest) group, n= 60, low VE/ MI group, n= 51, hi VE /CA
(west) group, n= 34, low VE/CA, n= 35). The subset contained 20 midwest subects and
11 west subjects; this would have added 10 to each MI group and 5 to each CA group (on
average). The average effect size for time 1 data is .36 ( range= .13- .64), and for time 2
the average effect size was rather low . 12 (range= .06- .22).

Using Cohen’s (1977) power tables (pp. 92-93), the average effect size of .36 is
not sufficient to detect significant differences at .80 power for any of the 4 groups.

Adding 5 more subjects to each of the CA low & high value expressive group may have

87

brought power up from .56 to .60, not a significant amount, and detectable effect size is
still insufficient. The more serious issue is the low sample size obtained despite adding
five per case. Adding ten more to the MI low and high value expressive groups has the
effect of moving power from .71 and .78 to .79 and .84, respectively. In this case, .36
effect size can sufficiently detect significant differences with only one group (high
VE/MI). Finally, the time 2 average effect size of .12 is not sufficient for detecting
differences even if the subset of 31 had been included. Admittedly, a bigger sample is
always preferable, but in this case excluding the 31 subjects effects power moderately in
the MI sample, but not in the CA sample at time 1, and is not relevant to the samples in
time 2. The researcher feels justified in excluding the subset because utirnately the effect
sizes in this study were sufficient enough to yield post hoc path models that were
consistent with time 1 data. It is felt that these analyses depict interesting and compelling
relationships worthy of more investigation. The validity of this study is not threatened

by the exclusion which was done to preserve coding accuracy.

2 Two additional attitude scales will be incorporated, but their analysis is outside
the scope of this particular study. They are reserved for secondary analysis in a separate
study. The second attitude scale presented is fiom Sedlacek and Brooks’ situational
attitude scale (1970). Two situations are taken; best fiiend’s engagement to a black
person and sibling’s engagement to a black person. In both situations participants will be
asked to indicate their feelings by circling a number from 1 to 7 placed between a series

of ten bipolar adjectives. They are aggressive/passive, happy/sad;

88

tolerable/intolerable; complimented/insulted; angered/overjoyed; secure/fearful;
hopeful/hopeless; excited/unexcited; right/wrong; disgusting/pleasing.

The third set of item are taken from Davidson and Schneider (1992) and are
composed of items that reflect increasing intimacy of social involvement and contact with
a black-white interracial couple: (1) working with someone who is interracially married ,
(2) having your child visit the home of a black-white interracial couple, (3) buying a
house next door, (4) having a close friendship, (5) accepting a sibling who interracially
marries, (6) accepting child who interracially maries, and (7) marrying a person outside

of your race. Response options range from 1 (very unwilling) to 7 (very willing).

3 Additional coding criteria for function categories is reserved for separate analysis
that is beyond the scope of this study. Aside fiom the value expressive function , other
functions were be coded. Social expressive functions are not based on personal
interactions or past experiences. Rather, these statements focus on descriptions of peer
and family influences on attitudes toward interracial marriage. These statements also
reflect a desire to be accepted by individuals who are considered important and
influential in the participants’ social environment. Consequently 5 subcategories were
established: messages about parents, siblings, general family, friends/peers and general
socieity. Additionally, these thoughts were coded as reflecting either negative, neutral or
positive associations towards interracial marriage.

Ego-defensive functions involve projection of unacceptable motives onto
interracial couples and expression of hostility toward them. Attitudes that are categorized

under this function are assumed to be generally unfavorable. Statements expressing fear,

89

disgust, discomfort or personal threat are indicative of this function. Ego defensive
thoughts are relatively one-sided in that they do not reflect positive associations.
Consequently, coders marked whether the thought was ego-defensive or not (e. g “yes” or
“no”)

Evaluatives are based on past interactions that involve reinforcement or
punishments with specific interracial couples. Attitudes are part of a knowledge structure
that organizes past experiences and provides guidelines for future interactions. This
includes prior stereotypical beliefs where the focus is on positive or negative attributes
about interracial couples. Three subcategories were established: past experiences (which
were either seen as punishing or rewarding) , anticipated rewards or punishments and

attributes. Additionally, these thoughts were coded as reflecting either negative, neutral

or positive associations towards interracial maniage.

Additional coding criteria for message appeals is reserved for separate analysis
that is beyond the scope of this study. Aside from the appeals to values, other types of
appeals reflecting the other attitude functions will be coded. Experiential appeals will be
statements that reflect past experiences that were either rewarding or punishing. They
will also reflect beliefs and attributes of interracial couples. Ego-defensive appeals will
reflect hostility, anxiety, threat or discomfort regarding marrying someone who is black
(or white). Social-adjustive appeals will reflect statements referencing how parents,
peers, and friends will react to the marriage. Coders agreed 100% of the time on social
expressive thoughts with a Kappa of 1.00 for content and a Kappa of .79 for thought

valence. Coders also agreed 100% of the time on evaluative thoughts with a Kappa of

90

1.00 for content and a Kappa 1.00 for thought valence. There was only one subject that
was coded as ego-defensive, and both coders identified the same thought as ego-

defensive.

91

APPENDICES

92

APPENDIX A

93

APPENDIX A

CONSENT LETTER and QUESTIONNAIRES for Time 1 and Time 2

The following survey asks you questions regarding your thoughts and opinions
about interracial marriage. The survey also assesses self-presentation variables and
values. The survey will take approximately 20—30 minutes.

The data you provide will be completely confidential and anonymous. Because
this consent form is not attached to the questionnaire your name will not be associated
with the responses you provide. By writing and signing your name on this consent form
you indicate your voluntary agreement to participate in this study.

Print your name

 

Signature

 

Com. Class T. A.

 

If you would like to obtain copies of the results or if you have any questions regarding
this study, please contact:

Victoria Orrego

468 Communication Arts & Sciences Bldg.
Dept. of Communication

Michigan State University

East Lansing, MI 48824

(517) 347-8616

Orregovi@pilot.msu.edu

Thank you for your participation. I appreciate it.

Victoria 0. Orrego

94

QUESTIONNAIRE - TIME 1

LAST FOUR NUMBERS OF PID# DATE OF BIRTH
(MM/DD/YR)

 

 

[ Please respond to the following questions. I

 

What year are you in college?

Freshman Sophomore Junior Senior
What is your age? years
Are you (please check one)? Female Male
What is your ethnicity?
Asian African- American/Black Native American
Caucasian
Pacific Islander Hispanic/Latino(a) Other: Please specify

 

 

Please read the following questions and answer them honestly about how you
feel. There are no “right” or “wrong” answers. Use the following scale to
place the number that best represents your position.

 

 

 

Strongly Disagree 1 2 3 4 5 6 7 Strongly
Agree

In social situations, I have the ability to alter my behavior if I feel that something
else is called for.

I have the ability to control the way I come across to people, depending on the
impression I wish to give them.

When I feel that the image I am portraying isn’t working, I can readily change it
to something that does.

I have trouble changing my behavior to suit different people and different
situations.

95

I have found that I can adjust my behavior to meet the requirements of any
situation I find myself in.

Even when it might be to my advantage, I have difficulty putting up a good front.

Once I know what the situation calls for it’s easy for me to regulate my actions
accordingly.
Strongly Disagree 1 2 3 4 5 6 7 Strongly
Agree

I am often able to read people’s true emotions correctly through their eyes.

In conversations, I am sensitive to even the slightest change in the facial
expression of the person I’m conversing with.

My powers of intuition are quite good when it comes to understanding others’
emotions an motives.

 

I can usually tell when I’ve said something inappropriate by reading it in the
listener’s eyes.

If someone is lying to me, I usually know it at once from that person’s manner of
expression.

In different situations and with different people, I often act like very different
persons.

I’m not always the person I appear to be.

I would not change my opinions (or the way I do things) in order to please
someone else or win their favor.

In order to get along and be liked, I tend to be what people expect me to be rather
than anything else.

I may deceive people by being fiiendly when I really dislike them.
In a group of people I am rarely the center of attention
I am not particularly good at making other people like me.

At a party, I let others keep the jokes and stories going

96

I feel a bit awkward in company and do not show up quite so well as I should.
__ I tend to show different sides of myself to different people.
__ Although I know myself, I find that others do not know me.

Different situations can make me behave like very different peOple.
__ Different people tend to have different impressions about the type of person I am.

I sometimes have the feeling that people don’t know who I really am.

97

THOUGHT LISTING TASK

 

 

In the space below, please describe and explain your attitude towards interracial
marriage between blacks and whites. Please indicate WHY you feel the way you
do. Write down all your thoughts and feelings that are relevant to your attitude
and try to describe phe ream for your feelings.

 

 

** please be as specific as possible“
I generally have a positive or negative attitude toward black-white interracial

marriage because...

98

 

 

 

Please circle the best number that corresponds with your attitudes toward
black-white interracial marriage. (the husband is black and the wife is white, in
this case)

 

Black-white interracial marriages are:

good 1 2 3 4 5 6 7 bad
rewarding 1 2 3 4 5 6 7 punishing
acceptable 1 2 3 4 5 6 7 unacceptable
respectful 1 2 3 4 5 6 7 disrespectful
harmful 1 2 3 4 5 6 7 beneficial
intolerable 1 2 3 4 5 6 7 tolerable
satisfactory 1 2 3 4 5 6 7 unsatisfactory

offensive 1 2 3 4 5 6 7 pleasant

 

 

Each situation is followed by 10 descriptive word scales. Your task is to
select for each descriptive scale, the rating which best describes your feelings
toward the item. Indicate your choice by circling the number that best
matches your feelings.

 

 

o If your best friend is White indicate how you would feel if your fiiend has just

become engaged to a black person
OR

99

 

o If your best friend is Black, Hispanic or Asian-American indicate how you would
feel if the your best fiiend became engaged to a white person.

Your best fiiend is: male female
aggressive 1 2 3 4 5 6 7 passive
happy 1 2 3 4 5 6 7 sad

tolerable 1 2 3 4 5 6 7 intolerable

insulted 1 2 3 4 5 6 7 complimented
angered 1 2 3 4 5 6 7 overjoyed
secure 1 2 3 4 5 6 7 fearful

o If your best friend is White indicate how you would feel if your friend has just
become engaged to a black person
OR
o If your best friend is 312%. Hispanic or Asian-American indicate how you would
feel if the your best fiiend became engaged to a white person.

hopeful 1 2 3 4 5 6 7 hopeless
excited I 2 3 4 5 6 7 unexcited
right 1 2 3 4 5 6 7 wrong

disgusting 1 2 3 4 5 6 7 pleasing

o If your brother or sister is White indicate how you would feel if he or she just

became engaged to a black person
OR

100

o If your brother or sister is Blfik, Hisparpric or Asian-American indicate how you
would feel if he or she became engaged to a white person.

Your sibling is: male female
aggressive 1 2 3 4 5 6 7 passive
happy 1 2 3 4 5 6 7 sad

tolerable 1 2 3 4 5 6 7 intolerable

insulted l 2 3 4 5 6 7 complimented
angered 1 2 3 4 5 6 7 overjoyed
secure 1 2 3 4 5 6 7 fearful
hopeful 1 2 3 4 5 6 7 hopeless
excited I 2 3 4 5 6 7 unexcited
right 1 2 3 4 5 6 7 wrong

disgusting 1 2 3 4 5 6 7 pleasing

101

 

Please indicate your willingness to participate in the following interactions. Place
the number that best represents your position.

 

 

Unwilling 1 2 3 4 5 6 7 Willing
Describe your willingness to work with a white person who is married to a
black person

If you have, or had, a child describe your willingness for your child to visit the
home of a black-white married couple

Describe your willingness to buy a house next door to a black-white interracial
couple

Describe your willingness to have a close fiiendship with an black-white
interracial couple

Describe your willingness to accept one of your close relatives, for example, a
brother or sister, marrying a black individual.

Describe your willingness to accept your son or daughter choosing to marry a
black individual. 1

If you are, or if you were single, describe your willingness to marry a black

individual.

102

 

 

Please rate how relevant the following values listed below are to your
consideration of black-white interracial marriage. Use the following scale to
place the number that best represents your position.

 

 

Totally
Totally
Hinders Values 1 2 3 4 5 6 7 Promotes

Values

Equality (equal opportunity for all)

Social Power (control over others, dominance)
Freedom (freedom of action and thought)

Sense of Belonging (feeling that others care about me)
Social Order (stability of society)

An Exciting Life (Stimulating experiences)

103

 

 

 

Please rate how relevant the following values listed below are to your
consideration of black-white interracial marriage. Use the following scale to
place the number that best represents your position.

 

Totally Promotes
Hinders Values 1 2 3 4 5 6 7 Values

Wealth (material possessions, money)

Creativity (uniqueness, imagination)

World at Peace (free of war and conflict)

Respect for Tradition (Preservation of time-honored customs)
Mature Love (Deep emotional and spiritual intimacy)
Self-Discipline (Self-restraint, resistance to temptation)
Family Security (safety for loved ones)

Social Recognition (respect, approval by others)

A Varied Life (filled with challenge, novelty, and change)
Wisdom (a mature understanding of life)

Authority (the right to lead or command)

Social Justice (Correcting injustice, care for the weak)
Independent (self-reliant, self-sufficient)

Moderate (Avoiding extremes of feeling and action)
Ambitious (hardworking, aspiring)

Broad-Minded (Tolerant of different ideas and beliefs)
Daring (Seeking adventure, risk)

Influential (having an impact on people and events)

104

 

 

 

Please rate how relevant the following values listed below are to your
consideration of black-white interracial marriage. Use the following scale to
place the number that best represents your position.

 

Totally Promotes
Hinders Values 1 2 3 4 5 6 7 Values

Honoring Parents and Elders (Showing respect)

Choosing Own Goals (Selecting own purposes)

Capable (competent, effective, efficient)

Accepting My Portion in Life (submitting to life’s circumstances)
Preserving My Public Image (Protecting my “face”)

Obedient (Dutiful, meeting obligations)

Helpful (Working for the welfare of others)

Devout (Holding religious faith and belief)

Curious (Interested in everything, exploring)

Forgiving (willing to pardon others)

Successful (achieving goals)

105

 

 

 

Use the following scale and place the number that best represents your
agreement or disagreement with the statements below.

 

 

Strongly Disagree 1 2 3 4 5 6 StronglyAgree

My opinions about black-white interracial marriage mainly are based on my
concern that we safeguard equality of interaction for all people in our society

My opinions about black-white interracial marriage mainly are based on my
concern that we safeguard the civil liberties of all people in our society

My opinions about black-white interracial marriage mainly are based on my
moral beliefs about how things should be.

My attitudes about black-white interracial marriage generally tell others a great
deal about my values

A person’s values should be one of the most important factors in determining
their attitudes toward black-white interracial marriage.

When my opinions about black-white interracial marriage are challenged , I
often accept other’s contrary opinions rather than sticking to my own attitudes.

Although I sometimes consider others’ views when I am formulating my
opinion about black-white interracial marriage, I ultimately adhere to my own personal
values or beliefs about what is right and wrong.

My view of black-white interracial marriage are simply a reflection of my peers

I would be extremely uncomfortable holding attitudes toward black-white
interracial marriage that are different than my peers.

I would be extremely uncomfortable holding attitudes toward black-white
interracial marriage that are different than my family.

I usually look to the opinions of others and not inside myself in order to decide
what my attitude on black-white interracial maniage is.

Rather than being a sole reflection of my values, my attitudes toward black-
white interracial maniage are often influenced by the opinions of my close friends.

To some degree, my parent’s opinions about black-white interracial marriage
issues influence my attitudes.

106

QUESTIONNAIRE - TIME 2

LAST FOUR NUMBERS OF PID#

DATE OF BIRTH

 

(MM/DD/Y R)

 

 

[ Please respond to the following questions.

 

What year are you in college?

Freshman Sophomore Junior
What is your age? __ years
Are you (please check one)? Female Male
What is your ethnicity?
Asian African-American/Black
Caucasian
Pacific Islander Hispanic/Latino(a)

 

107

Senior

Native American

Other: Please specify

MESSAGE CONSTRUCTION

 

 

Imagine that your best friend has just informed your that he or she is
considering getting engaged to a black individual. Note: Depending on M
best friend’s ethnicity the race of his or her fiancee will vary. If your best
friend is: Asian, Hispanic or Black the fiancee will be white. The couple
should be black-white or other-ethnicity-white.

 

 

What is the ethnicity of your best friend?

Asian (fiancee is white)
African-American/Black (fiancee is white)
Native American (fiancee is white)
Caucasian (fiancee is black)
Pacific Islander (fiancee is white)
Hispanic/Latino(a) (fiancee is white)

Other: Please specify

 

Your best friend is: (please check one)? Female Male

 

In the space below please construct a persuasive message, based on your attitude
toward interracial marriage, arguing for or against your best friend’s
engagement. List specific reasons in support for or in opposition against this
marriage.

 

 

AFTER LISTENING TO MY BEST FRIEND’S NEWS I WOULD ARGUE FOR
DR AGAINST THIS MARRIAGE BY SAYING...

108

 

 

 

Please rate how important the following values are as general guiding principles
in your life. Use the following scale to place the number that best represents your

position.

 

Not at all Very
important 1 2 3 4 5 6 7 Important

Equality (equal opportunity for all)
Social Power (control over others, dominance)
Freedom (freedom of action and thought)
__ Sense of Belonging (feeling that others care about me)
Social Order (stability of society)
An Exciting Life (Stimulating experiences)
Wealth (material possessions, money)
Creativity (uniqueness, imagination)
World at Peace (free of war and conflict)
Respect for Tradition (Preservation of time-honored customs)
Mature Love (Deep emotional and spiritual intimacy)
Self-Discipline (Self-restraint, resistance to temptation)
Family Security (safety for loved ones)
Social Recognition (respect, approval by others)
__ A Varied Life (filled with challenge, novelty, and change)
__ Wisdom (a mature understanding of life)

Authority (the right to lead or command)

Social Justice (Correcting injustice, care for the weak)

109

 

 

 

Please rate how important the following values are as general guiding
principles in your life. Use the following scale to place the number that
best represents your position.

 

Not at all Very
important 1 2 3 4 5 6 7 Important

Independent (self-reliant, self-sufficient)

Moderate (Avoiding extremes of feeling and action)
Ambitious (hardworking, aspiring)

Broad-Minded (Tolerant of different ideas and beliefs)
Daring (Seeking adventure, risk)

Influential (having an impact on pe0ple and events)
Honoring Parents and Elders (Showing respect)
Choosing Own Goals (Selecting own purposes)
Capable (competent, effective, efficient)

Accepting My Portion in Life (submitting to life’s circumstances)
Preserving My Public Image (Protecting my “face”)
Obedient (Dutiful, meeting obligations)

Helpful (Working for the welfare of others)

Devout (Holding religious faith and belief)

Curious (Interested in everything, exploring)
Forgiving (willing to pardon others)

Successful (achieving goals)

110

 

 

 

Use the following scale and place the number that best represents your
agreement or disagreement with the statements below.

 

 

Strongly Disagree 1 2 3 4 5 6 7 Strongly
Agree

My opinions about black-white interracial marriage mainly are based on my
concern that we safeguard equality of interaction for all people in our society

My opinions about black-white interracial marriage mainly are based on my
concern that we safeguard the civil liberties of all people in our society

My opinions about black-white interracial marriage mainly are based on my
moral beliefs about how things should be.

My attitudes about black-white interracial maniage generally tell others a great
deal about my values

A person’s values should be one of the most important factors in determining
their attitudes toward black-white interracial marriage.

When my opinions about black-white interracial marriage are challenged , I
often accept other’s contrary opinions rather than sticking to my own attitudes.

Although I sometimes consider others’ views when I am formulating my
opinion about black- white interracial marriage, I ultimately adhere to my own personal
values or beliefs about what is right and wrong.

My view of black-white interracial marriage are simply a reflection of my peers

I would be extremely uncomfortable holding attitudes toward black-white
interracial maniage that are different than my peers.

I would be extremely uncomfortable holding attitudes toward black-white
interracial maniage that are different than my family.

I usually look to the opinions of others and not inside myself in order to decide
what my attitude On black-white interracial marriage is.

Rather than being a sole reflection of my values, my attitudes toward black-
white interracial maniage are often influenced by the opinions of my close friends.

To some degree, my parent’s opinions about black-white interracial marriage
issues influence my attitudes.

 

lll

APPENDIX B

112

Appendix B

Items, means, SD, and factor loadings of the self-monitoring items

 

 

Item M SD F
Abiliitv to modiy self-presentation
1. In social situations, I have the ability to alter my 5.83 1.13 .56

behavior if I feel that somethng else is called for
2. I have found that I can adjust my behavior to meet

the requirements of any situation I frnd myself in 5.20 1.32 .49
3. Even when it might be to my advantage, I have

difficulty putting up a good front 5 .25 1.57 .48
4. Once I know what the situation calls for it’s easy

for me to regulate my action accordingly 5.57 1.11 .73
Sensitivity to expressive behavior
1. I am often able to read people’s true emotions

correctly throught their eyes 4.91 1.43 .73
2. My powers of intuition are quite good when it

comes to understanding other’s emotions and motives 5.43 1.14 .63
3. If someone is lying to me I usuallly know it at once

from that person’s manner of expression 4.80 1.32 .54

fl=226

113

Items, means, SD, and factor loadings of the otherdirectedness, extroversion, attitude

 

 

items

Item M SD F
Otherdirectedness scale
1. I’m not always the person I appear to be 3.80 1.96 .64
2. I tend to show different sides of myself to

different people. 4.96 1.59 .71
3. Different situation can make me behave like

very different people 4.27 1.68 .67
Extroversion scale
1. In a group of people I am rarely the center

of attention 4.38 1.61 .61
2. I feel a bit awkward in company and do not

show up quite so well as I should 5.16 1.54 .61

Attitude toward interracial marraiage scale

Black and White interracial marriage are:

1. Rewarding --- Punishing (R) 5.12 1.60 .80
2. Respectul -- Disrespectful (R) 5.60 1.58 .80
3. Harmful --- Beneficial (R) 5.10 1.63 .66
4. Satisfactory --- Unsatisfactory 5.49 1.63 .85
5. Offensive --- Pleasant 5.48 1.48 .84

( R) Item was reflected for analysis

lj= 226

114

Items, means, SD, and factor loadings of the value relevant items

 

 

 

Item M SD F
Rate how relevant the following values are to

your attitude of black and white interracial marriage

Baum 6.12 1.34 1.00
Sucesflrl Benevolence

Helpful 5.65 1.43 .83
Forgiving 5.84 l .37 .84
Mature Love 6.14 1.31 .71
Capable 5.86 1.31 .84
Ambitious 5.67 1 .40 .82
Independent 5 .76 1 .36 .67
Choosing own goals 6.06 1.24 .76
Power-Tradition scale

Authority 4.76 1 .75 .79
Moderate 4.69 1 .68 .62
Accepting portion in life 5.46 1.51 .71
Social Order 5.32 1.66 1.00
Honoring Parents 5.26 1.83 1.00
Varied Life 5.89 1.35 1.00

N=226

115

APPENDIX C

116

2 61

3 42

4 56

5 42

6 44
F1 = Equality

Appendix C

Inter— Factor Correlations of the 6 value scales

78
49 49
68 56 42

68 50 34 44

F2 = Benovolent Sucess

F3 = Power-Tradition
F4 = Social Order

F5 = Honoring Parents

F6 = Varied Life

117

APPENDIX D

118

Appendix D

Descriptive Statistics and T-values for the Western and Midwestern Samples

 

Margble M for W SD M for MW SD T-
Lillie.

Race 5.04 1.84 4.92 1.76 -.45
Sex .62 .49 .58 .50 -.61
SelfP 5.43 .94 5.46 .89 .20
Oth'erd 4.33 1.52 4.35 1.3 .08
Ext 4.70 1.33 4.76 1.29 .31
Att 5.38 1.25 5.34 1.28 -.20
Equal 6.12 1.4 6.23 1.19 .50
SocOrd 5.07 1 .63 5.43 1 .63 1 .46
Honor 5.23 1.75 5.37 1.81 .52
Total 1 4.35 1.91 3.96 2.11 -1.26
Total 2 2.82 1.19 2.79 1.55 -.15
Express 1.49 1.1 1 1.37 1.09 -.74
Vethought 1.49 1.1 1 1.27 1.08 -l .32
VE 38.85 29.80 34.32 27.1 -1.05
E .833 6.46 1.63 8.21 .64
Ex 38.83 29.82 37.13 27.08 -.40
Ev 20.65 25.32 23.67 28.10 .73
Bas 23.17 29.45 28.45 32.89 1.01
Pt .00 .00 .26 2.59 79
So .00 .00 .00 .00 --
Hon .00 .00 .00 .00 --
Var .00 .00 .00 .00 --

APPENDIX E

120

Appendix E

Means and Standard Deviations for the regression of Lttitudes
on the benenvolent success scale for low yalue expressives

and high value expressives for the overall sample

 

 

 

 

Variable LowValue Expressive High Value Expressive
Ben-success M = 5.74 M = 6.06

SD=1.18 SD=.81

N= 95 N =85
Attitude M = 5. 07 M =5.73

SD=1.39 SD=1.04

 

121

APPENDIX F

122

Appendix F

Means and Standard Deviations for the regression of attitudes
on the power-tragtion scale for low value expressives and high value expressives for

the overall sample

 

 

Variable LowValue Expressive High Value Expressive
Power-trad. M = 4.76 M = 5.17

SD = 1.40 SD = 1.28

N= 95 N =85
Attitude M = 5. 07 M = 5.73

SD=1.39 SD=1.04

 

123

APPENDIX G

124

Appendix G

Means and Standard Deviations for the regpession of attitudes
on the varied life value for low value expressiveL and high vfie expressives

 

 

 

 

Variable LowValue Expressive High Value Expressive
Varied Life M = 5.88 M = 5.95

SD= 1.41 SD=1.24

N= 95 N =85
Attitude M = 5. 07 M =5.73

SD=1.39 SD=1.04

 

125

APPENDIX H

126

Appendix H

Correlations for the equality path model

F1 = Equality values

F2 = Value expressiveness
F3 = Attitude

F4 = Equality messages

127

APPENDIX I

128

Appendix I

Correlations for the benevolent success path model for the Midwest sample

1
2 27
3 53 33

F1 = Benevolent success values
F2 = Value expressiveness

F3 = Attitude

F4 = Benevolent success messages

129

APPENDIX J

130

Appendix J

Correlations for the benevolent success path model for the Western sample

2 l6
3 52 21
4 -O3 -.31 -17

F 1 = Benevolent success values
F2 = Value expressiveness

F3 = Attitude

F4 =Benevolent success messages

131

APPENDIX K

132

Appendix K

Correlations for the Power-tradition path model for the Midwest sample

F 1 = power tradition values
F2 = Value expressiveness

F3 = Attitude

F4 = power-tradition messages

133

APPENDIX L

134

Appendix L

Correlations for the power-tradition model for the Western sample

1 2 3
1
2 l4
3 3O 21

F 1 = Power-tradition values
F2 = Value expressiveness
F3 = Attitude

135

APPENDIX M

136

Appendix M

Correlations for the Varied life model for the Midwest sample

1 2 3
l
2 .07
3 .38 33

F 1 = Varied life values
F2 = Value expressiveness
F3 = Attitude

137

APPENDIX N

138

Appendix N

Correlations for the Varied life model for the Western sample

1 2 3
1
2 O7
3 46 21

F 1 = Varied life values
F2 = Value expressiveness
F3 = Attitude

139

APPENDIX 0

140

Appendix 0

Correlations for the Honoring parents model

1 2 3
1
2 l9
3 45 29

F 1 = Honoring parents values
F2 = Value expressiveness
F3 = Attitude

141

APPENDIX P

142

Appendix P

Correlations for the Social order model

1 2 3
l
2 05
3 28 27

F1 = Social order values
F2 = Value expressiveness
F3 = Attitude

143

REFERENCES

144

References

Abelson, R. P. (1982). Three modes of attitude-consistency. In M. P. Zanna, E.
T. Higgins, & C. P. Herman (eds.) Consistency in social behavior: The Ontarip

syr_nmsium, (V 01 2, pp. 131-147).

Ajen,1. (1988). Attitudes, p_ersonality and belpaviora, Homewood, IL: Dorsey
Press.

Anderson, D. S., & Kristiansen, C. M. (1990). Measuring attitude functions. TM
Journal of Social Psychology, 130(3 1, 419-421.

Biemat, M., Crandall, C. S. (1994). Stereotyping and contact with social groups:
Measurement and conceptual issues. European Journal of Social Psychology, 24, 659-
677.

Biemat, M., Vescio, T., Theno, S. A., & Crandall, C. S. (1996). Values and
prejudice: Toward understanding the impact of American values on outgroup attitudes.
In C. Seligrnan, J. M. Olson, & M. P. Zanna (Eds), The psychology of values: The
Ontario symmsium, volume 8 (pp.l-24). Mahwah: NJ: Erlbaum.

Briggs, S. R., Cheek, J. M., & Buss, A. H. (1980). An analysis of the self-
monitoring scale. Journal of Personalin and Social Psychology, 38, 679-686.

Bymes, D. A. & Kiger, G. (1988). Contemporary measures of attitudes toward
blacks. Educational and Psychological Measurement, 48, 107-118.

Carithers, M. N. (1970). School desegregation and racial cleavage, 1954-1970:
A review of the literature. Journal of Social Issues 26 25-47.

 

Cohen, J. (1977). Statistical mwer analysis for the behavioral sciences. New
York: Academic Press.

Cohen, J. (1960). A coefficient of agreement of nominal scales. Educational and
Psychological Measurement, 20, 37-46.

Davidson, J. R (1992). Theories about black-white interracial marriage: A

clinical perspective. Journal of Multicultmal Counselmg' and Development, 20, 150-
157.

Davidson, J. R, & Schneider, L. J. (1992). Acceptance of black-white interracial
marriage. The Journal of Interggoup Relations, 19(3 1, 47-52.

145

DeBono, K. G. (1987). Investigating the social-adjustive and value-expressive
functions of attitudes: Implications for persuasion processes. Jogal of Personalig and
Social Psvchology,_5_(;)_, 279-287.

Devine, P. G. (1989). Automatic and controlled processes in prejudice: The role
of stereotypes and personal beliefs. In A. R. Pratkanis, S. J. Breckler, & A. G.
Greenwald (Eds), Attitude; structure and function (pp.l82-211). Hillsdale, NJ:

Erlbaum.

Dillard, J. P., & Hunter, J. E. (1989). On the use and interpretation of the
emotional empathy scale, the self-consciousness scales, and the self-monitoring scale.

Communication Research, 16(11, 104-129.

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth,
TX: Harcourt Brace Jovanovich.

Gadberry, J. H., & Dodder, R A. (1993). Educational homogamy in interracial
marriages: An update. Journal of Social Behavior and Personalig, 8161, 155-163.

Garcia, S. D., & Rivera, S. M. (1999). Journal of Social and Personal
Relationships, 161 1 1, 65-86.

Greene, J. 0., Geddes, D. (1993). An action assembly perspective on social skill.
Communication Theopy, 3111, 26-49.

Greene, J. O. (1984). A cognitive approach to human communication: an action
assembly theory. Communication Monographs, 51, 289-306.

Guetzkow, H. (1950). Unitizing and categorizing problems in coding qualitative
data. Journal of Clinical Psychology, 6, 47—58.

Herek, G. M. (1984a). Beyond “homophobia”: A social psychological
perspective on attitudes toward lesbians and gay men. Journal of Homosexualig,

1011/21, 1-21.

Herek, G. M. (1984b). Attitudes toward lesbians and gay men: A factor-analytic
study. Journal of Homosexualig, 10,11/21, 39-51.

Herek, G. M. (1986). The instrumentality of attitudes: Toward a neofunctional
theory. Journal of Social Issues, 42(21, 99-114.

Herek, G. M. (1987). Can functions be measured? A new perspective on the
firnctional approach to attitudes. Social Psychology Quarterly. 50(4). 285-303.

Hegar R. L. (1994). Parental abduction of children from interracial and cross-
cultural marriages. Journal pf Commtive Family Studies, 25(11, 135-142.

146

Johnson, B. T., & Eagly, A. H. (1989). Effects of involvement of persuasion: A
meta-analysis. Psychological Bulletip, 106121, p.290-313.

Katz, D. (1960). The functional approach to the study of attitudes. Public
Opinion Merly, 136, 163-204.

Katz, 1. (1991). Gordon Allport’s the nature of prejudice. Political Psychology,
1_2_(1_)_,_125-157.

Katz, 1., & Hass, R. G. (1988). Racial ambivalence and American value conflict:
correlational and priming studies of dual cognitive structures. Jourh_al of Personaligy and

Social Psychology, 55161, 893-905.

Kelman, H. C. (1958). Compliance, identification, and internalization: Three
processes of attitude change. Journal of Conflict Resolutiop, 2, 51-60.

Kehnan, H. C. (1961). Processes of opinion change. Public Opinion Quarterly,
2_5_, 57-78.

Kinder, D. R. (1986). The continuing American dilemma: White resistance to
racial change 40 year after myrdal. Journal of Social Issues, 42121, 151-171.

Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics: Symbolic racism
versus racial threats to the good life. Journal of PerscMality and Social Psychology, 40,
414-430.

Kitano, H. L., Yeung, W., Chai, L., & Hatanka, H. (1984). Asian-American
interracial marriage. Journal of Marriage and The Family, F ebrpfl, 179-190.

Korolewicz, M. & Korolewicz, A. (1985). Effects of sex and race on interracial
dating preferences. Psychological Remus, 57, 1291-1296.

Kouri, K. M. (1993). Black-white marriages: Social change and
intergenerational mobility. Marriage and Farrrilv Review,19(3-4). 241-253.

Kristiansen, C. M. (1989). The symbolic/value-expressive function of outgroup
attitudes among homosexuals. The Journal of Social Psychology, 130111, 61-69.

Kristiansen, C. M., Zanna, M. P. (1988). Justfiying attitudes by appealing to
values: A firnctional perspective. British Journal of Social Psychology, 27, 247-256.

Lampe, P. E. (1982). Interethnic dating: Reasons for and against. International
Journal of Intercultural Relations 6 115-126.

 

Lampe, P. E. (1981). Towards amalgamation: Interethnic dating among blacks,
Mexican Americans, and Anglos. Ethnic Groups, 3, 97-109.

147

Lennox, R. D., & Wolfe, R. N. (1984). Revision of the self-monitoring scale.
Journal of Personalifl and Social Psychology, 46, 161, 1349-1364.

McConahay, J. B., & Hough, J. C. (1976). Symbolic racism. Journal of Social
Issues 32 2 23-45.

Maio, G. R, & Olson, J. M. (1994). Value-attitude-behaviour relations: The
moderating role of attitude functions. British Journal of Social Psychology, 33, 301-312.

Maio, G. R, & Olson, J. M. (1995). Relations between values, attitudes, and
behavioural intentions: The moderating role of attitude function. Journal of
Experimental Social Psychology, 31, 266-285.

Mills, J. K., Daly, J., Longmore, A. & Kilbride, G. (1994). A note on family
acceptance involving interracial friendships and romantic relationships. The Journal of

Psychology, 129131, 349-351.

Moore, J. W., Hauck, W. E., & Denne, T. C. (1984). Racial prejudice, interracial
contact, and personality variables. Journal of Experimental Psychology, 52, 168-173.

Murray, S. L., Haddock, G. L., Zanna, M. P. (1996). On creating value-
expressive attitudes: An experimental approach. In C. Seligman, J. M. Olson, & M. P.

Zanna (Eds), The psychology of values: The Ontario sylpppsium, volume 8 (pp.lO7-
131). Mahwah: NJ: Erlbaum.

Niemann, Y. F. , Secord, P. F. (1995). Social ecology of stereotyping. Journal
for the Theopy of Social Behavior, 25, 1-13.

Paset, P. S., & Taylor, R. D. (1991). Black and white women’s attitudes toward
interracial maniage. Pachological Rgpgrts, 69131, 753-754.

Patterson, A. M., Sedlacek, & Perry, F. W. (1984). Perceptions of blacks and
Hispanics in two campus environments. Journal of College Student Personnel,
November 513-518.

 

Peterson, K. S. (1997, April 7). Racial intermarriage is on the rise in USA. USA
Today, pp. OD. 1.

Porterfield, E. (1978). Black and white marriages. Chicago: Nelson-Hall.
Relaxing an old taboo. (1996, July 15). Newsweeh p. 51.

Rokeach, M. (1973). The nature of human values. New York: Free Press.

148

Ryan, C. S., Judd, C. M., & Park, B. (1996). Effects of racial stereotypes on
judgments of individuals: The moderating role of perceived group variability. Journal of

Exmrimental Social Psychology, 32, 71-103.

Sagiv, L., & Schwartz, S. H. (1995). Value priorities and readiness for out-group
social contact. Joumnal of Personalig and Social Psychology, 69131, 43 7-448.

Schwartz, S. (1996). Value priorities and behavior: Applying a theory of
integrated value systems. In C. Seligman, J. M. Olson, & M. P. Zanna (Eds), The
psychology of values: The Ontario smpgsium, volume 8 (pp.l-24). Mahwah: NJ:

Erlbaum.

 

Schwartz, S. (1992). Universals in the content and structure of values:
Theoretical advances and empirical tests in 20 countries. In M. P. Zanna (Ed), Advances

in eapprimental social psychology, volume 25 (1-64). San Diego: CA: Academic Press.

Schwartz, S., & Bilsky, W. (1990). Toward a theory of the universal content and
structure of values: Extensions and cross-cultural replications. Journal of Personaligy

and Social Psychology, 58151, 878-891.

Sedlacek, W. E., & Brooks, C. G. (1970). Measuring racial attitudes in a
situational context. Pachological Ramrts, 27, 971-980.

Simpson, G., & Yinger, J. (1985). Racial and cultural minorities: An analysis of
prejudice and discrimination (5th ed.). New York: Plenum.

Shavitt, S. (1990). The role of attitude objects in attitude functions. Journal of
Exparimental Soci_al Psvcholgv. 26. 124-148.

Snyder, M. (1974). The self-monitoring of expressive behavior. Journal of
Personality and Social Psychology, 30, 526-537.

Snyder, M. & DeBono, K. G. (1985). Appeals to image and claims about
quality: Understanding the psychology of advertising. Journal of Personalig and Social

Psychology, 49, 586-597.

Solsberry, P. W. (1994). Interracial couples in the united states of America:

Implications for mental health counseling. Journal of Mental Health Counseling, 16131,
304-317.

South, S. J ., & Messner, S. F. (1986). Structural determinants of intergroup

association: Interracial marriage and interracial crime. American Journal of Sociology,
_6_, 1409-1430.

149

Stephan, W. G., Ageyev, V., Stephan, C. W., Abalakina, M., Stefanenko, T., &
Coates-Shrider, L. (1993). Measuring Stereotypes: A comparison of methods using
Russian and American samples. Social Psychology Quarterly, 56, 54-64.

The state of our unions. (1996, July 8). Time p. 13.

 

Todd, J., McKinney, J. L., Harris, R, Chadderton, R, & Small, L. (1992).
Attitudes toward interracial dating: Effects of age, sex, and race. Journal of
Multicultural Counseling and Development. 20. 202-208.

Tucker, M. B., & Mitchell-Keman, C. (1990). New trends in black American

interracial marriage: The social structural context. Journal of Marriage and the Family,
_5__2_,_ 209-218.

U. S. Bureau of the Census, Current population reports. (1998, March). Marital
status and living arrangements: series p20-514.

Wyman, M., & Snyder, M. (1997). Attitudes toward “gays in the military”: A

functional perspecitve. Journal of Applied Social Psychology, 24141, 306-329.

Zweigenhafi, R L., Cody, M. L. (1993). The self-monitoring of black students on
a predominantly white campus. The Journal of Social Psychology, 133111, 5-10.

150

HICHIGAN smrE UNIV. LIBRRRIES
11111111 111 111111111 1111111111111 111 111 1111111
31293020582106