"

“HQ

“gun"

.1
Yr 3.
(Law
H: 2.... .
:1

z .
.quIrtx
[Eh

.
9.51... :1:
7.9.3.. .

2
5

m

at. s.
V 5...:

an...
:1...

3.3:: . .r :
E: :3. .3.
3.1.2:... V I.

Jana.
is. 35...}
..:..,..“..x..
. ..
.f. as. 6.:
.

11.... z
:1 11
.. x: :01

uhkn!..:s
v3

wry... ‘
a! .

 

 

'Hlfins‘

.1. .,

 

’ ’ F‘ nu a. .- f-n’. V N
1' _ v

:‘oliI-I
£51, ;¥fi :1 '2 5'1 ., (at '- _' {F3 .
“frail-t algal” camps

University

 

 

 

This is to certify that the
thesis entitled
SEIF-CCNSTRUAL AS A PREDICI'OR OF COMJNAL
AND EXCHANGE ORIENI‘ATI(1\1:TAIWANESE AND AMERICANS

presented by

Hui CHIH Chiu

has been accepted towards fulfillment
of the requirements for

MA degree in WION

 

M {1%,
Major professor

 

Date .'0W ()1 400;

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN 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.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 CVCIRC/Datmpflfi-piS

SELF-CONSTRUAL AS A PREDICTOR OF COMMUNAL AND EXCHANGE
ORIENTATION: TAIWANESE AND AMERICANS

By

Hui Chih Chiu

A THESIS

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

MASTER OF ARTS

Department of Communication

2001

ABSTRACT

SELF -CONSTRUAL AS A PREDICTOR Of COMMUNAL AND EXCHANGE
ORIENTATION: TAIWANESE AND AMERICANS

By
Hui Chih Chiu

This study investigates cultural and gender differences in independent,
interdependent and relational~interdependent self-construals and these three
dimensions of self-construal serve as predictors of exchange and communal
orientations. The sample consisted of 357 students from Taiwan (178) and the United
States (179). The major findings of this study include: (a) cultural and gender
differences were unlikely to be characterized by the same dimensions of
self-construals (b) interdependent and relational-interdependent self-construals
predicted a preference for communal orientation, but manifested different effects in
two cultures; (c) independent self-construal showed a preference for communal
orientation, rather than exchange orientation; (d) Americans scored relatively higher
on exchange and communal orientations compared to Taiwanese. These findings
suggest that the concept of self-construal is more complex than previously thought.
Future research needs to expand the self-construal construct commonly discussed in

the literature.

Copy Right

Hui Chih Chiu

2001

I dedicate this thesis to my loving family who gave
me their full support and encouragement.

iv

ACKNOWLEDGEMENTS

At Michigan State University, I have had the opportunity to work with admirable
and dedicated scholars. From the very beginning, my major advisor Dr. Mary Bresnahan
provided nothing but ceaseless support and encouragement. She patiently dealt with my
drafty-drafts, and helped me to improve my work. Dr. Timothy Levine and Dr. Stan
Kaplowitz also provided me with challenges and insight that I may have missed
otherwise.

Dr. Mary Bresnahan’s Cross-Cultural Communication and
Dr. Timothy Levine’s Communication in Close Relationships were two instrumental
courses that helped me produce the invaluable first draft of ideas. Dr. Frank Boster and
Dr. Gwen Wittenbaum introduced me to the quantitative research methodology that
always proves useful. I would like to thank all these professors for their intellectual
inspiration.

Finally, I must thank my family and friends without whom I would not be the

person I am today.

TABLE OF CONTENTS

List of Tables

List of Figures

Chapter 1: Introduction

Chapter 2: Self-Construal
Goal of the Study
Definition of Self-Construal
Challenges to Self-Construal
Discussion of Reasons for Inconsistent Findings
The Three Dimensions of Self-Construal
Discussion of Reasons for the Three Dimensions of Self-Construal
Self-construal in the United States and Taiwan
Gender Differences in Self-Construal

Chapter 3: Relationship Orientations
Self-Construal and Relationship Orientations
Communal and Exchange Orientations
Linking Communal and Exchange Orientations to Self-Construal
Culture and Relationship Orientations

Chapter 4: Method
Overview of the Study
Participants
Procedures
Translation
Measures
Preliminary Analysis

vi

viii

\Owaw

11
l4
16

18
18
19
20
21

23
23
23
24
24
24
25

Chapter 5: Results
Additional Analysis

Chapter 6: Discussion
Conclusion
Limitations

Chapter 7: Footnotes
References
Appendices

Tables and Figures

Questionnaire
Research Consent Form

vii

28
33

34
38
38

40

41

48

49

59
64

List of Tables
Table 1: T-Test for Means of Self-Construals between the United States and
Taiwan
Table 2: T-Test for Means between Types of Self-Construal in the United States
Table 3: T-Test for Means between Types of Self-Construal in Taiwan.
Table 4: T-Test for Means of Self-Construals between Men and Women

Table 5: T-Test for Means of Self-Construals between Men and Women in the
United States

Table 6: T-Test for Means of Self-Construals between Men and Women in Taiwan
Table 7: Two-way ANOVA Results for Independent Self-Construal

Table 8: Two-way ANOVA Results for Interdependent Self—Construal

Table 9: Two-way ANOVA Results for Relational-lnterdependent Self-Construal

Table 10: Standardized Multiple Regression Coefficients Predicting Exchange
Orientation—Overall

Table 11: Standardized Multiple Regression Coefficients Predicting Exchange
Orientation—US.

Table 12: Standardized Multiple Regression Coefficients Predicting Exchange
Orientation—Taiwan

Table 13: Standardized Multiple Regression Coefficients Predicting Communal
Orientation—Overall

viii

Table 14: Standardized Multiple Regression Coefficients Predicting Communal
Orientation—US.

Table 15: Standardized Multiple Regression Coefficients Predicting Communal
Orientation—Taiwan

Table 16: T-Test for Means of Exchange and Communal Orientations between the
United States and Taiwan

List of Figures

Figure 1: Means for Self-Construals across Countries

Figure 2: Means for Self-Construals within Countries

Figure 3: Means for Self-Construals of Men and Women

Figure 4: Means for Self-Construals of Men and Women in Taiwan

Figure 5: Means for Self-Construals of Men and Women in the United States

Figure 6: Means for Exchange and Communal Orientations in the United States
and Taiwan

List of Abbreviations

COM: Communal Orientation Scale

EX: Exchange Orientation Scale

IND: Independent Self-Construal Scale

INTER: Interdependent Self-Construal Scale

RISC: Relational-Interdependent Self-Construal Scale

xi

Chapter 1

Introduction

The notion of independent and interdependent self-construals, originally
introduced by Markus and Kitayama (1991), serves as an individual-level cultural
orientation and is theorized to mediate and explain the effects of culture on cognition,
emotion and motivation. For years, much has written about cultural differences based on
the self-construal theory.

Recently, researchers have argued that the dichotomous view of the self
oversimplifies the complexity of self-concept. Masumoto (1999), for example, claims
that “contemporary views of self-culture relations suggest that this relationship is much
more complex than previously thought, and certainly more complex than a generalized
view of self that pits individual and group needs in opposition to each other.” (p.304)
Similarly, other scholars speculate that the theoretical separation of independent and
interdependent self—construals at the individual level is a Western bias (Park, & Levine,
1999). Assuming that independent and interdependent self-construals work in the same
way in different cultures is fundamentally problematic (Park et al., 1999; Levine,
Bresnahan, Park, Lapinski, & Wittenbaum, 2001).

These recent, complex views of the self have received support in the literature as
well. For example, a group of researchers distinguish three, rather than two, dimensions
of self-construal (Triandis, 1989; Madson & Trafimow, 2001; Kashima et al., 1995; Wang,
Bristol, Mowen, & Chakraborty, 2000). In attempting to expand the original construct of

self-construal, the current study focuses on three dimensions of self-construal

(independent, collective-interdependent and relational-interdependent self-construals).
This study begins with a review of self-construal theory (Markus & Kitayama,
1991) and research. The inconsistent findings in self-construal research give reasons for
distinguishing three dimensions of self-construal. Following this, cultural and gender
effects on self-construals are reviewed and examined. Finally, self-construal is argued as

a predictor for exchange and communal orientations.

Chapter 2

Self-Construal

Goal of the Study

This study seeks a fuller understanding of the self-construal construct. There are
two goals in the current study. The first goal is to examine whether there are cultural and
gender differences in the three dimensions that have been proposed for self-construal--
independence, interdependence, and relational interdependence. The second goal is to
investigate whether self-construal is a useful construct for predicting when people are
likely to prefer exchange and communal orientations.
Definition of Self-Construal

What exactly is suggested by the concept of self-construal? In the original
conceptualization, Markus and Kitayama (1991) defined self-construal as two aspects of
self-concept which affect what people “believe about the relationship between the self
and others, especially, the degree to which they see themselves as separate from others or
as connected with others”(p.226). Markus and Kitayama claim that these two divergent
aspects of the self influence an individual’s psychological processes (e.g., cognition,
emotion, and motivation). In other words, self-construal is “a constellation of thoughts,
feelings and actions concerning the relationships of the self to others and the self as
distinct from others” (Singelis, 1994, p.581).

An independent self-construal is defined as a bounded, unitary, stable self that is
separate from social context (Markus & Kitayama, 1991; Singelis, 1994). The behavior of

people who emphasize their independent self is organized and made meaningful primarily

by reference to one’s own abilities, attributes, characteristics, or goals, rather than by
reference to the thoughts, feelings, and actions of others. The social situation in general
and other people are important, but mainly as standards of reflected appraisal, or as
sources that can verify and affirm the inner core of the self (Markus & Kitayama, 1991;
Sharkey & Singelis, 1995). The constellation of elements composing an independent
self-construal includes an emphasis on (a) internal abilities, thoughts, and feelings; (b)
being unique and expressing the self; (c) realizing internal attributes and promoting one’s
own goals; and (d) being direct in communication (Singelis, 1994, p.581).

An interdependent self-construal is defined as a flexible, variable self that is
connected with social context (Markus & Kitayama, 1991; Singelis, 1994). The behavior
of people who emphasize their interdependent self “is determined by what the actor
perceives to be the thoughts, feelings, and actions of others in the relationship”(Markus &
Kitayama, 1991 , p.227), and “is marked by sensitivity to situations and social contexts”
(Kanagawa, Cross, & Markus, 2001, p.91). Within such a construal, the self and others
are intertwined. Status, roles, relationships and belongingness are central in the
understanding of self (Markus & Kitayama, 1991; Sharkey et al., 1995). The constellation
of elements composing an interdependent self-construal includes (a) external and public
features; (b) belonging and fitting in; (c) occupying one’s proper place and engaging in
appropriate action; and ((1) being indirect in communication and reading others’ minds”
(Singelis, 1994. p.581).

Focusing on the differences in self-construals, Markus and Kitayama (1991)
theoretically link independent and interdependent construal of the self to the

cultural-level dimensions of individualism and collectivism. The independent

self-construal is argued to be dominant in individuals from the United States and Western
Europe, while the interdependent self-construal is thought to dominate in collectivist
non-Western cultures (Markus & Kitayama, 1991). Specifically, socialization differences
imposed by different cultures are thought to give rise to different self-concepts (Levine et
al., 2001). Members of individualistic cultures are socialized into their culture. They learn
the major values of their culture (e. g., independence, achievement) and acquire preferred
ways for how members of the culture are expected to view themselves (e. g., as unique
persons). Members of collectivistic cultures, however, learn different major values (e.g.,
harmony, solidarity) and acquire different preferred ways to conceive of themselves (e. g.,
as interconnected with others) (Gudykunst et al., 1996).

In this view, Markus and Kitayama (1991) proposed a mediating model of
self-construals. In this model, culture-level individualism and collectivism are
systematically related to individual-level cultural orientations (independent and
interdependent self-construals), which, in turn, systematically influence a variety of
outcome variables. Consequently, “the validity of self-construal theory and research rests,
at a minimum, on the existence of systematic cultural differences in relative strengths of
the two types of self-construals” (Levine et al., 2001, p.6). Stated differently, “for
self-construals to mediate and explain cultural differences, self-construals must reflect the
intended cultural differences” (Levine et al., 2001, p.6).

Markus and Kitayama’s idea has generated much scholarly interest and has
resulted in at least 101 studies which have been published to date using the self-construal
concept.1 In addition, a large number of research papers based on self-construal have

been presented at professional conferences in communication, psychology, anthropology,

and other disciplines. Some research findings are consistent with Markus and Kitayama’s
(1991) claims (e.g., Singelis, Bond, Sharkey & Lai, 1999; Singelis & Brown, 1995;
Singelis & Sharkey, 1995). For instance, a series of studies conducted by Singelis and his
colleagues support the theoretical framework of self-construals. First, Singelis and Brown
(1995) examined a path model linking cultural collectivism with interdependent
self-construal and, ultimately, high-context communication. The results of the path
analysis allowed them to conclude that self-construal is an ideal candidate to perform the
role of linking culture to behavior. Furthermore, in a study on embarrassment, Singelis
and Sharkey (1995) found that Asian Americans were more interdependent and less
independent than Euro-Americans and, subsequently, were more susceptible to
embarrassment. Based on this finding, these researchers claimed that individual
differences in self-construals accounted for cultural differences in embarrassability. More
recently, in another article about self-esteem and embarrassability, Singelis et al. (1999)
draw similar conclusions.
Eggs to Self-Construal

Other studies, however, have shown inconsistent results with Markus and
Kitayama’s (1991) claims (e. g., Cross, 1995; Gudykunst et al., 1996; Gardner, Gabriel, &
Lee, 1999; Kim, Hunter, Miyahara, Horvath, Bresnahan, & Yoon, 1996; Oetzel, 1998a;
Oyserman, Coon, & Kemmelmeier, 2001; Park & Levine, 1999). First, at the
cultural-level dimensions, a recent meta-analysis of seventy-one
individualism-collectivism studies conducted by Oyserman, Coon, and Kemmelmeier
(2001) reported that European Americans are not more individualistic than African

Americans or Latinos, and not less collectivistic than Japanese or Koreans. Among

Asians, only Chinese show the tendencies of being both less individualistic and more
collectivistic.

Second, at the individual-level dimensions, the data across cultures indicate that
self-construals are not constantly associated with culture in the predicted directions. For
example, in Kim et al.’s (1996) study, the mainland United States participants showed the
highest independent orientations, followed by participants from Hawaii, Japan, and Korea.
This was in the direction predicted by these researchers and by Markus and Kitayama
(1991). The scores for the interdependent self-construal, however, were not in the
expected order. The results indicated no significant differences among participants from
Korea, Japan, and the mainland United States. The scores for Hawaii were significantly
higher than for other three groups. Mixed results were also reported by other researchers
Cross (1995) and Gudykunst et al. (1996) found that although Asian participants scored
higher on interdependent self-construal than did American participants, the groups’
average scores did not differ significantly on independent self-construal, and scores on
independent self-construal were higher than scores on interdependent self-construal
among Asian participants. In addition, Sato and Cameron (1999) compared Canadians
(71% European ancestry; 13% East Asian ethnic background) with Japanese. It was
found that Canadians had significantly higher scores on the interdependence scale than
did Japanese and that the two samples did not differ on the independence scale.

Research findings within a culture also challenge Markus and Kitayama’s (1991)
claims. In particular, Asian people seem unlikely to emphasize their interdependent
self-construal over independent self-construal. Perhaps the most significant evidence is

provided by Levine et al.’s (2001) meta-analysis. Across nine studies and around 4000

subjects, the data clearly contradict the claim that Asians are more interdependent than
independent. More specifically, of nine studies that included participants from Japan,
Korea, and Hong Kong, six studies found that the effects of culture on self-construals
were not in the predicted directions.2

Moreover, little support has been found for the mediating role of self-construals.
For example, in Park and Levine’s (1999) study, the overall results were consistent with
the possibility that interdependent self-construal mediated the effect of cultures on
attitude toward behavior and subjective norm. On the other hand, the data were not
consistent with the mediation model specifying independent self-construal as mediator of
the effect of cultures on attitude toward behavior and subjective norm. Furthermore, these
researchers suggested viewing self-construals more as moderators of cultural effects than
as mediators. Similarly, in Kim et al.’s (1996) study, interdependent self-construal
mediated the effect of culture-level collectivism on other-oriented conversational
constraints. Culture-level individualism, however, had a stronger direct effect on
outcome-oriented conversational constraints than did independent self-construal, the
mediator. Both of these studies provide evidence that self construal is not a mediating
variable.

Although, research results, across cultures or within a culture, are mixed and
inadequate to support Markus and Kitayama’s (1991) claims, only few researchers have
criticized the validity of the self-construal construct. One such study was conducted by
Matsumoto (1999). Based on a careful review of self-construal studies, Matsumoto (1999)
concluded that “the evidence available to date severely challenges the validity of their

theoretical framework for explaining observed national differences in psychological

phenomena”(p. 301). Levine et al. (2001) and Park et al. (1999) draw similar conclusions.
In addition, Cross, Bacon, and Morris (2000) argued that the interdependent
self-construal defined by Markus and Kitayama (1991) does not adequately describe the
nature of the Western version of the interdependent self-view.

Discussion of Reasons for Inconsistent Findings

In general, there are two explanations for the inconsistent findings in
self-construal research. First, self-construals are dynamic. Researchers have argued that
although it is possible to describe cultures as being individualistic or collectivistic, people
are guided both by independent and interdependent self-construals, which are activated
by different contexts, values, and social constraints (Gudykunst et al., 1996; Singelis &
Brown, 1995; Bresnahan, Liu, Nebashi, & Liao, 1999). Similarly, Trafimow, Triandis,
and Gato (1991) propose that private self (independent self-construal) and collective self
(interdependent self-construal) are stored in a separate location in memory. The retrieval
of a particular type of the self depends on the accessibility of the factors with which it is
associated (e. g., culture, gender). Thus, self-construals may not work consistently as
predicted.

Asecond explanation for the inconsistent findings in self-construal research is the
multidimensionality of self-construals. Researchers have argued that Markus and
Kitayama’s (1991) definition of self-construals presents an overly simplistic view of the
self and comprises a variety of constructs (Cross et al., 2000; Levine et al., 2001). Over
the past decade, researchers have overly focused on the investigation of two dimensions
of self-construals: independent and interdependent self-construals. A group of researchers,

however, distinguish three, rather than two, dimensions of the self. For example, Triandis

(1989) conceptualized the self as having three aspects: private self, public self, and
collective self. Likewise, these three dimensions were defined as private, collective, and
allocentric selves by Madson et al. (2001), as individualistic, collective, and relational
selves by Kashima et al. (1995) and as self-other orientation, dependence and
self-orientation by Wang et al. (2000). Recently, Cross et al. (2000) have articulated a
different conceptualization of interdependent self-construal, relational-interdependent
self-construal, which is defined as the tendency to think of oneself in terms of
relationships with close others. Cross et al. redefine dimensions of self- construal as
independent, group-oriented interdependent, and relational-interdependent self-construals.
This study adopts the terminology of collective interdependent self-construal that will be
used to distinguish group-oriented interdependence from relational interdependence.
The Three Dimensions of Self-Constru_a_l

The differences among independent, collective-interdependent, and
relational-interdependent self-construals are the levels of inclusiveness with others (Cross
et al., 2000; Brewer, & Gardner, 1996; Wang et al., 2000). Specifically, independent
self-construal defines the self as independent of relationships and as separated from
others. Collective-interdependent self-construal is the self-concept derived from
connections and role relationships with group members. Relational-interdependent
self-construal emphasizes including close others into the self. The main distinction
between collective-interdependence and relational interdependence is that
collective-interdependence focuses on the connection with one’s group, while relational
interdependence emphasizes personal connection with significant others. Furthermore,

these different self-construals coexist within the same individual, available to be activated

at different times or in different contexts (Brewer et al., 1996). Consequently, a person
shifts the three levels of identity from “I” to “We as a Group” to “We in a Relationship”
depending on social contexts.

Discussion of Reasons for the Three Dimensions of Self-Construal

There are three reasons for distinguishing the three dimensions of
self-construal—definition of who are the others, cultural differences, and gender
differences. First, the three dimensions of self-construal more clearly define who
comprises “others” in self-other relationships, relative to the two dimensions. Levine et al.
(2001) argued that the definitions of interdependent and independent self-construals fail
to describe who constitutes those ambiguous “others” that are constantly referenced in
self-construal theory (Markus & Kitayama, 1991). In the three dimensions of
self-construal, collective-interdependent self-construal defines “we” as a large group
context, while relational-interdependent self-construal refers “we” to a small group of
significant others (Brewer et al., 1996). This further distinction in two levels of
interdependent self-construal partially resolves the ambiguities in the self-construal
definitions proposed by Markus and Kitayama (1991).

Second, the three dimensions of self-construal more accurately explain cultural
differences in self-construals compared to the two dimensions. Cross et al. (2000) argued
that the group-oriented notion of interdependence does not adequately describe the
relationship-centered conception of interdependence that characterizes North Americans.
“North Americans and Western Europeans are more likely to include representations of

close relationships rather than group membership into the self” (p. 791).

Theoretically, because of cultural differences between collectivism and
individualism, members of collectivistic and individualistic cultures will develop
different forms of interdependent self-construal (Cross et al., 2000). In collectivism-based
interdependence, the individual’s position in the group or situation dictates behavior.
Therefore, knowing one‘s place, behaving according to one’s role, and putting the needs
of the group before one’s own needs are central characteristics that shape the
self-construal (Cross et al., 2000). People in individualism-based interdependence,
however, are more likely to include personal relationships (e.g., a spouse, mother, best
friend or colleague) in their self-representations (Cross et al., 2000). Moreover, Triandis
(1989) claimed that people from individualistic cultures treat group memberships rather
casually and have relatively little loyalty to in-groups. Their in-groups are larger, place
relatively fewer demands on members, and are voluntary. On the other hand, people from
collectivistic cultures concern about the results of their actions on members of their
in-groups and feel involved in the lives of in-group members. Their in-groups are small,
and make demands in many areas. These different views of self-other relationships
suggest that it is reasonable to differentiate Markus and Kitayama’s (1991) construct of
collective-interdependent self-construal from the relational-interdependent self-construal
conceptualized by Cross et al. (2000).

While the Cross et al.’s (2000) Relational Interdependence Scale has not been
tested cross-culturally before the current study, there is related empirical research that
supports relational-interdependent self-construal as dominant in members from the
United States and Western Europe and col[ective-interdependent self-construal as

dominant in collectivist non-Western cultures. For example, Kanagawa et al. (2001)

administered the Twenty Statement Test (Hartley, 1970; McPartland, Cummming, &
Garreston, 1961) assessing cultural variation in the self-concept. The findings indicated
that the self-descriptions of American students had more references to close relationships
(e.g., families and friends), while the Japanese students made more references to distant
relationships (e.g., social memberships). They explained that the American students view
themselves as independent from many types of social influence, but their relationships
with their families and friends appear highly salient in their thoughts about themselves.

Taken together, the theoretical and empirical evidence suggests that the three
dimensions of the self more accurately describe cultural differences in self-construals
than do the two dimensions proposed by Markus and Kitayama (1991). That is because
individualistic and collectivistic cultures tend to have different versions of
interdependence. In individualistic cultures, interdependence is better described by
relational-interdependent self-construal, while interdependence in collectivistic cultures is
more represented by collective-interdependent self-construal that measures perception of
group norms.

Thethirdreason for defining three dimensions of the self is to provide a fuller
understanding of gender differences in self-construals. It is believed that females are
socialized to be interdependent and attuned to relationships, whereas males are socialized
to be autonomous and independent. Given their different socialization, one would expect
that females tend to define themselves in terms of their relationships, while males define
themselves in terms of their uniqueness and separateness from others (Madson et al.,
2001). Several studies (e.g. Cross & Madson, 1997; Madson et al., 2001; Kashima et al.,

1995; Wang et al., 2000) have found that females are more connected, collective,

13

allocentric, or interdependent than males, and males are more separated, independent or
individualistic compared to females.

Other researchers, however, claimed that both men and women care about their
connections with other people, but they pursue it in different spheres (Baumeister &
Sommer, 1997; Gabriel & Gardner, 1999; Brewer et al., 1996). More precisely, “women’s
sociality is oriented toward dyadic close relationships, whereas men’s sociality is oriented
toward a large group” (Baumeister et al., 1997, p38) Thus, men and women tend to have
different forms of interdependence and the three dimensions of self-construals can clarify
the dissimilarity.

Considering these three reasons together, the distinction of independent,
collective-interdependent, and relational-interdependent self-construals provides an
alternative way to define individual differences in the extent to which the self is separated
or connected to others. Compared to the two dimensions of self-construal commonly used
in the literature, the three dimensions of self-construal articulate two different forms of
interdependence and represent three levels of inclusiveness. However, whether
collective—interdependent self-construal characterizes collectivistic and men’s cultures,
while relational-interdependent self-construal characterizes individualistic and women’s
cultures remains debatable.

Self-Construal in the United States and Tgiwan

A careful review of the literature reveals that no research has investigated the
differentiations of the three dimensions of self-construal in the United States and Taiwan.
Some researchers, however, do compare Americans with Taiwanese in terms of

independent and co“ective-interdependent self-construals. Partially consistent with

14

Markus and Kitayama’s (1991) claims, Cross (1995) found that Taiwanese were more
collectiveeinterdependent and less independent than were Americans. When analyzed
within the culture, Taiwanese showed greater tendencies toward independent
self-construal than collective-interdependent self-construal. Likewise, Krull et al. (1999)
reported that Taiwanese were significantly higher in collective-interdependent
self-construal than Americans, but no differentiation was found in independent
self-construal. More unexpectedly, another study revealed that American students hold a
more allocentric orientation (collective-interdependent self), while Taiwanese students
hold a more idiocentric orientation (independent self) (Witte, Murray, Liu, & Hubbell,
1993). Thus, none of the previous research completely supports Markus and Kitayama’s
(1991) claims.3

On the other hand, cultural differences in relational-interdependent self-construal
of Americans and Taiwanese have not been studied. It is unclear whether relational-
interdependent self-construal has the possibility of predicting cultural differences. The
only related data were found in Cross et al.’s (2000) study. It was reported that
Americans tended to score higher on the relational-interdependent self-construal scale
relative to noncitizens.4

In sum, because there is conflicting evidence about independence and
collective-interdependence for both Taiwanese and Americans, and because no previous
published research specifically compares relational-interdependent self-construal of
Americans and Taiwanese, it is not possible to predict any tendencies toward the three
dimensions of self-construal for Americans and Taiwanese. Thus, one research question

is advanced:

15

RQl: Are there any significant differences in independent,
collective-interdependent, and relational-interdependent self-construals between
participants in the United States and Taiwan?

Gender Differences in Self-Construal

Considering gender differences in the three dimensions of self-construal, research
in the United States has shown that women are more likely than men to construct an
interdependent or relational self-view (Cross et al., 1997; Madson et al., 2001; Kashima
et al., 1995). Some researchers take one step further to distinguish men from women in
terms of interdependent self-view. For example, Gabriel et al. (1999) claimed that women
are more focused on the close relationships they are part of, whereas men are more
focused on the groups to which they belong. Similarly, Brewer et al. (1996) argued that a
collectivist or group-oriented interdependence is characteristic of men in American
culture. Moreover, Cross et al. (2000) examined the tendencies of men and women to
relational-interdependent self-construal and found that women were more likely to define
themselves in terms of their close relationships than were men. Based on these empirical
findings, three hypotheses are advanced:

H1: Males will score significantly higher on independent self-construal than will

females.

H2: Males will score significantly higher on collective-interdependent

self-construal than will females.

H3: Females will score significantly higher on relational-interdependent

self—construal than will males.

16

Comparing cultural and gender differences in the three dimensions of
self-construal, Kashima et al. (1995) argued that cultural and gender effects do not have a

great deal of overlap. Cultures are more likely to be differentiated on independent and

 

uuiicuiivv ' ‘ J r J ‘ self-construal. Men and women are best differentiated on the
relational dimension. Likewise, Wang et al. (2000) reported that variations in the
self-orientation dimension (independent self) tended to uniquely represent cultural
differences, while the dependence dimension (relational-interdependent self) uniquely
characterized the genders. Moreover, dissimilarities in the dimension of self-other
association (collective-interdependent self) characterized both the cultural and gender
groups. In attempting to examine whether cultural and gender differences in the three
dimensions of the self have little overlap as previously argued, the following research
question is advanced:

RQ2: What are the effects of culture, gender, and the culture by gender interaction

on independent, collective-interdependent, and relational-interdependent

self-construals?

Chapter 3

Relationship Orientations
Self-Construal and Relationship Orientations

As noted in the previous chapter, self-construals are dynamic and activated as
appropriate to different situations. Some researchers take one step further, arguing that
people act differently depending on with whom they are interacting and the situation in
which the interaction is occurring (Park et al., 1999; Triandis, Bontempo, Villareal, Asai
& Lucca, 1988; Kanagawa et al., 2001). In other words, these researchers’ concern is how
individualsshift their self-construals in different types of relationships.

There have been several attempts to investigate the associations between
relationship types and self-concepts. In friendships, for instance, Gardner et al. (1999)
found that shifting perceptions of obligation to friends in American participants could
increase the importance of obligation to others in need. In close relationships, Aron, Aron,
Tudor, and Nelson (1991) reported that Americans tended to see themselves highly
interdependent with their partner. Moreover, in the study by Triandis et a1. (1988), the
results revealed that there were no differences between the Japanese and Illinois samples
in their reactions to parents, while most of the cultural differences were found for
acquaintances, coworkers, and friends. These findings support that self-construals are
quite sensitive to relationship types. Thus, there is reason to specify the relationship types

in self-construal research.

Communal and Exchgng Orientations

Based on the previous argument, the current study focuses on two relationship
orientations, communal relationships and exchange relationships, which are proposed by
Clark and Mills (1979). Clark and Mills (1979) distinguish communal relationships from
exchange relationships based on the rules governing the giving and receiving of benefits.
In communal relationships, members feel a special obligation for the other’s welfare. In
exchange relationships, a benefit is given with the expectation of receiving a comparable
benefit in return or in repayment for a benefit received previously (Clark, Dubash, &
Mills, 1998). According to this definition, the major difference between these two
relationship orientations is that exchange orientation emphasizes self-interests more,
while communal orientation is more concerned about the welfare of others.

To date, many studies have supported this distinction between communal
orientation and exchange orientation. For instance, when a communal relationship is
desired, as opposed to when an exchange relationship is desired, people have been found
to monitor the other’s needs more (Clark, Mills, & Powell, 1986; Clark, Mills, &
Corcoran, 1989), to respond more positively to the other’s expressions of emotions (Clark
& Taraban, 1991; Clark, Ouellette, Powell, & Milberg, 1987), and to help others more
(Clark et al., 1987). In contrast, when an exchange relationship is desired, as opposed to
when a communal relationship is desired, peOple were more likely to monitor inputs into
a joint task for which there was a joint reward (Clark et al., 1989), to react more
positively to being paid back for a favor (Clark & Mills, 1979) and to feel more exploited

if a repayment did not follow a favor (Clark & Waddell, 1985).

I9

LLan Communal and Exchange Orientations to Self-Construal

Communal and exchange orientations can be theoretically linked to the three
dimensions of self-construal. The canal link between self-construals and relationship
orientations is that both of them are concerned with what people believe about the
relationships between the self and others. By definition, a person with an independent
self-construal focuses on the needs, thoughts and feelings of self, while a person with an
collective-interdependent self-construal attempts to meet the needs of others and to
promote the other’s goals (Markus & Kitayama, 1991). Similar to
collective-interdependent self-construal, a person with a relational-interdependent
self-construal tends to take into account the needs and wishes of others (Cross et al.,
2000). Compared to self-construals, Clark and Mills (1979) define communal relationship
as one in which members benefit one another on the basis of concern for the other’s
welfare, and exchange relationship as members benefiting one another in response to
specific benefits received in the past or expected in the future. Thus,
collective-interdependent self-construal, relational- interdependent self-construal and
communal orientation are all characterized by mutual concern for the interests and
outcomes of the other, while both independent self-construal and exchange orientation
focus on self-interests.

To date, although there is no strong empirical evidence for the links between
self-construals and relationship orientations, some cross-cultural research on
self-construals may be able to support the theoretical connections argued above. In a
study of conflict styles, Oetzel (1998b) reported that an independent self-construal was

positively associated with dominating conflict styles, which were defined as a high

20

concern for self and a low concern for others. On the other hand,
collective-interdependent self-construal was positively associated with obliging styles,
which were defined as a combination of low self-concem and high other-concem. Kitani
and Kim (1996) indicated similar results when investigating conflict styles in romantic
relationships. Furthermore, it was found that people with a relational-interdependent
self-construal tend to take into account the needs and wishes of close others when making
a decision (Cross et al., 2000).

Given the theoretical links and the empirical findings, it seems plausible to
assume that an independent self-construal would predict one’s preference for exchange
orientation, while a collective-interdependent self-construal and a
relational-interdependent self-construal would predict one’s preference for communal
orientation. Thus, three hypotheses are advanced:

H4: Independent self-construal is positively related to exchange orientation.

H5: Collective-Interdependent self-construal is positively related to communal

orientation.

H6: Relational-interdependent self-construal is positively related to communal

orientation.

Culture and Relationship Orientations

Regardless self-construals, cross-cultural research on relationship orientations has
shown mixed results. For example, Vanyperen and Buunk (1991) found that American
subjects were more exchange oriented than Dutch with respect to the evaluation of
potential contributions to an intimate relationship. On the other hand, Yum and Canary

(2000) reported that individualistic European Americans were significantly less exchange

21

oriented than were collectivistic Koreans in terms of problem-solving behavior in close
relationships. Furthermore, it was found that Taiwanese were as cognizant of their
relational exchanges as were Americans in their dating relationships or cross-sex
friendships (Lin & Rusbult, 1995). In other words, no difference was found between
Taiwanese and Americans for exchange orientation. Given these mixed results, one
research question is advanced:

RQ3: Are there any significant differences in exchange and communal

orientations among participants in the United States and Taiwan?

22

Chapter 4

Method

Overview of the Study

This study examines whether there is any connection between the three types of
self-construal with communal and exchange orientations. Participants from the United
States and Taiwan were included in this study because they have been described as
representing differences in individualism and collectivism. Taiwan has been described as
a collectivistic culture while the United States is considered to be an individualistic
culture (Hofstede, 1980). Given these reported differences in individualism and
collectivism, participants from these countries should provide a fair test of the predictions
being made in the current study.
Participants

The sample consisted of 357 undergraduate students from Taiwan and the United
States. One hundred and seventy-eight Taiwanese students (68 males and 110 females)
from Yuan Ze University located in Taoyuan, Taiwan and one hundred and seventy—nine
American students (60 males and 119 females) enrolled in Communication classes at
Michigan State University participated in the current study. The overall mean age for
participants in this study was 20.19 (Taiwan, M = 20.33, SD = 1.44; US, M = 20.05, SD
= 1.28).

All of the participants from Taiwan were Taiwanese born in Taiwan. Among the
participants from the United States, there were 152 Caucasians (84.9%), 14 African

Americans (7.8%), 8 Asian/Pacific Islanders (4.5%), 4 Hispanics (2.2%), and 1

23

unclassifiable (.6%).
Procedures

The participants were asked to consent for their participation in the study and then
filled out a questionnaire that included the Communal Orientation Scale (Clark, Ouellette,
Powell, & Milberg, 1987), the Exchange Orientation Scale (Clark, Mills, & Corcoran,
1989), the Relational-Interdependent Self-Construal Scale (Cross, Bacon, & Morris,
2000), and the Self-Construal Scale (Leung & Kim, 1997) with some demographic
questions.
Translation

The questionnaire was translated from English to Chinese and back translated
from Chinese to English by two bilinguals (Brislin, 1986). Discrepancies between the
original and English back-translation were reviewed and corrected. The revised
questionnaires were then piloted with a small sample of students from both countries.
Measures

Communal orientation was assessed by the Communal Orientation Scale (Clark,
Ouellette, Powell, & Milberg, 1987), which consists of 14 items. Items were rated on a
five point Likert scale with the larger mean scores indicating greater communal
orientation. Clark et al. (1987) reported .78 reliability for the original scale.

Exchange orientation was assessed by the Exchange Orientation Scale (Clark,
Mills, & Corcoran, 1989), which consists of 9 items. Items were rated on a five point
Likert scale with the larger means indicating greater exchange orientation. Although the

measure of exchange orientation was an unpublished scale, Vanyperen and Buunk (1991)

24

reported .62 (the American version) and .68 (the Dutch version) reliability for the original
scale.

Relational-interdependent self-construal was assessed by the Relational-
Interdependent Self-Construal Scale (Cross, Bacon, & Morris, 2000), which consists 11
items. Items were rated on a five point Likert scale with the larger means of scores
indicating greater relational-interdependent self-construal. Cross et al. (2000) reported .88
reliability for the original scale.

Independent and collective-interdependent self-construals were assessed by the
Self-Construal Scale (Leung & Kim, 1997). This scale consists 29 items, including 15
items for independent self-construal and 14 items for interdependent self-construal. Items
were rated on a five point Likert scale. The ratings for each scale were averaged as
measures of their respective constructs. Leung and Kim (1997) reported that the
independent self-construal scale had .87 reliability and the interdependent self-construal
scale had .73 reliability.

Levine et al. (2001) have argued that the Self-Construal Scale (Leung & Kim,
1997) is, in fact, multidimensional and measures more than two constructss Therefore,
this study examined the unidimensionality of the Self-Construal Scale to ensure its
validity. Furthermore, the Communal Orientation, Exchange Orientation, and
Relational-Interdependent Self-Construal scales were also tested for their
unidimensionality.

Prelimingy Analysis
Unidimensionality means that all the items in a scale measure the same trait.

According to Hunter and Gerbing (1982), there are two statistical tests for

25

unidimensionality: internal consistency and parallelism. The criterion of internal
consistency specifies that the correlation between two items in the same scale should be
the product of their correlations with the underlying trait (to within sampling error). The
criterion of parallelism specifies that the items in a unidimensional scale have similar
patterns of correlations with items in other scales or other traits (to within sampling
error).

Thus, responses to each scale were first tested for internal consistency and
parallelism by confirmatory factor analysis (CFA) (Hunter & Gerbing, 1982). The
procedure of CF A allowed determining which items of each scale measure the same trait.
Since there are two cultural groups in the current study, Taiwanese, American and overall
responses were analyzed by CF A separately.

Based on the results of internal consistency and parallelism, items that did not
meet the criteria for acceptance were deleted from each scale and this CFA procedure
was repeated until acceptable unidimensional solution was reached. CFA analyses
resulted in deleting several items from each scale.

After performing the CFA procedure, only four out of the original fourteen items
in the Communal Orientation Scale were found to be unidimensional. These four items
were used to compute the Communal Orientation score. The standard score coefficient
alphas of this scale were .73 overall, .66 in the United States and .60 in Taiwan.

CFA results revealed that four out of the original nine items remained to comprise
the Exchange Orientation Scale. The standard score coefficient alphas of this scale
were .63 overall, .61 in the United States and .69 in Taiwan.

For the Relational-Interdependent Self-Construal Scale, six of the original eleven

26

items remained to calculate the Relational-Interdependent Self-Construal score. The
standard score coefficient alphas of this scale were .72 overall, .65 in the United States
and .76 in Taiwan.

The Self-Construal Scale has been considered as an orthogonal two-factor
solution with all independent self-construal items loading one factor and all
interdependent items loading on the other factor. Thus, this study tested Independent
Self-Construal Scale and Interdependent Self-Construal Scale respectively.

For the Independent Self-Construal Scale, five of the original fifteen items
remained to calculate Independent Self-Construal score. The standard score coefficient
alphas of this scale were .81 overall, .70 in the United States and .72 in Taiwan.

For the Interdependent Self-Construal Scale, five of the original fifteen items
remained to calculate the Interdependent Self-Construal score. The standard score
coefficient alphas of this scale were .81 overall, .70 in the United States and .72 in
Taiwan.6

In sum, after deleting problematic items from each scale using the CFA procedure,
a total of twenty-five out of the original sixty-three items remained to continue testing
hypotheses. Furthermore, of the five scales, only the Communal Orientation Scale passed
both internal consistency and parallelism tests. The other four scales failed in the

parallelism test. These data replicate Levine et al.’s (2001) findings that the factor

structures of the self-construal scales are problematic.

27

Chapter 5

Results
R1: Are there any significant differences in independent, collective-interdependent, and
relation_al-interdependent self-construals among participants in the United States and
M2

Independent sample, two-tailed t-tests were used to examine if Americans and
Taiwanese are different from each other in terms of independent,
co“ective-interdependent, and relational-interdependent self—construals. As shown in
Table 1, Americans’ independent self-construal scores (M = 4.04, S_D = .45) were
significantly higher than Taiwanese (M = 3.18, SD = .49 with; [355] = -17.26, p < .001).
On the other hand, Taiwanese scored higher on collective-interdependent self-construal
(M = 3.71, E = .43) than did Americans (M =2.94, S_D = .55 with t [355] = 14.78, p
< .001). For relational-interdependent self-construal, there were no significant differences
between Taiwanese (M =3.84, S2 = .45) and Americans (M = 3.89, S_D = .50 with t [355]
= -.96, p = .34, us.)

[Table 1 and Figure 1 about here]

Paired samples, two-tailed t-tests were used to examine if the mean scores for the
three dimensions of self-construal are significantly different within cultures. As shown in
Table 2, in the United States, the mean score of independent self-construal (M = 4.04, S_D
= .45) was significantly higher than that of collective-interdependent self-construal (M
=2.94, _S_D = .55 with t [178] = 18.23, p < .001) and relational-interdependent

self-construal (M = 3.89, S_D = .50 with t [178] = 3.22, p = .002). The mean score of

28

collective-interdependent self-construal (M = 2.94, M) = .55) was significantly lower
than that of relational-interdependent self-construal (M =3.89, SD = .50 with t [178] =
-l8.94, p < .001). In addition, the most significant mean differences were found between
independent and collective-interdependent self-construals.

As shown in Table 3, in Taiwan, the mean score of independent self-construal (M
= 3.18, M2 = .49) was significantly lower than that of collective-interdependent
self-construal (M =3.71, S_D = .43 with t [177] = -10.42, p < .001) and
relational-interdependent self-construal (M = 3.84, S_D = .45 with t [177] = -14.20, p
< .001). The mean score of collective-interdependent self-construal (M =3.71, S_D = .43)
was also significantly lower than that of relational-interdependent self-construal (M
=3.84, SD = .45 with t [177] = -3.53, p = .001). Moreover, the most evident mean
differences were between independent and relational-interdependent self-construals.

[Table 2,3 and Figure 2 about here]

H1: Males will score significantly higher on independent self-construal than will females.

 

H2: Males will score sigpificantly higher on collective-interdependent self-construal than

will females.

H3: Females will score significantly higher on relational-interdependent self-construal

 

than will males.

Hypothesis 1 to 3 assumed that men and women would emphasize different
dimensions of self-construals. Hypothesis I predicted that men would score significantly
higher on independent self-construal than would women. As shown in Table 4, the result
was not consistent with the hypothesis. Independent self-construal scores were not

significantly different between men (M = 3.52, S_D = .64) and women (M = 3.66, SQ

29

= .63 with t [355] = -1.96, p = .05, us.) Hypothesis 2 posited that men would score
higher on collective—interdependent self-construal than would women. Although the
result showed that men’s scores (M = 3.39, S_D = .62) were higher than women’s (M =
3.29, SD = .63), the differentiation was not significant (t [355] = 1.48, p = .14, us.)
Hypothesis 3 predicted that women would score significantly higher on
relational-interdependent self-construal than would men. As expected, women
emphasized relational-interdependent self-construal (M = 3.91, S_D = .45) more than did
men (M = 3.79, S_D = .50 with; [355] = -2.37, p < .05).

[Table 4 and Figure 3 about here]

Within each culture, gender differences were not evident. As shown in Table 5, in
the United States, men and women did not differ from each other in their tendencies
toward independent and relational-interdependent self-construal. On the other hand,
consistent with hypothesis 3, American men scored higher on collective-interdependent

self-construal (M = 3.07, S_D = .56) than did women (M = 2.88, SD = .53 with t [177] =

 

2.21, p < .05). In addition, table 6 revealed that no gender differences were found in any
dimension of self-construal for Taiwanese.
[Table 5,6 and Figure 4,5 about here]
R2: What are the effects of culture. gender. and the culture by gender interaction on
independent, collective-interdefindenh and relatjppal-interdependent self-construals?
To further compare culture with gender effects on independent, interdependent,
and relational-interdependent self-construals, three two-way ANOVAS were conducted

separately. In Table 7, the first ANOVA with independent self-construal as the dependent

30

variable presented a significant main effect for culture (E [1, 353] = 295.86, p < .001 , 77 2

= .45), but not gender (E [1, 353] = 3.28, p = .07, n.s.). There was no interaction effect for
culture and gender (E [1, 353] = .62, p = .43, n.s.).
[Table 7 about here]
In Table 8, the second ANOVA with collective-interdependent self-construal as
the dependent variable showed a significant main effect for culture (E [1, 353] = 218.9,

p < .001, 7} 2 = .38) and a significant interaction effect between culture and gender
(E [1, 353] = 5.15, p < .05, 77 2 = .01). There was no main effect for gender (E [1, 353]

=1.36, p = .25, n.s.).
[Table 8 about here]
In Table 9, the results for relational-interdependent self-construal indicated a

significant main effect for gender (E [1, 353]=5.38, p < .05, 7} 2 = .01). There was neither

a main effect for culture (E [1, 353] = .73, p = .39, n.s) nor an interaction effect between
culture and gender (E [1, 353] = .09, p = .77, n.s.).

[Table 9 about here]
H4: Independent self-construal is positively related to exchange orientation.

H5: Collective-interdependent self-construal is positively related to communal

orientation.
H6: Relational-interdependent self-construal is positively related to communal
orientation.

Hypothesis 4 to 6 assumed that three dimensions of self-construal would predict
individuals’ exchange and communal orientations. In addition, the above mentioned

analyses indicated that participants of different cultures and genders exhibited more or

31

less differentiation among independent, collective-interdependent, and
relational-interdependent self-construals. Thus, in order to control for culture and gender,
this study included these two factors as independent variables in a multiple regression
model.

As shown in Table 10, the first multiple regression analysis for the entire sample,
showed that, when entered simultaneously, independent self-construal,
collective-interdependent self-construal, relational-interdependent self-construal, culture
and gender contributed significantly in predicting exchange orientation. The model
revealed an overall R2 = .04, E (5, 351) = 3.02, p < .05. It indicated that the five variables
accounted for 4% (adjusted R2 = .03) of the variance in predicting exchange orientation
(R2 = .02, E [4,174] = .83, p = .51, n.s., in the United States; a2 = .11, E [4, 173] = 5.41, p
< .001, in Taiwan) (see Table 11 and 12).

[Table 10, 11 and 12 about here]

Hypothesis 4 anticipated that independent self-construal would be positively
related to exchange orientation. Inconsistent with the hypothesis, the strength of
independent self-construal was not related to exchange orientation across cultures or
within each culture. Overall, the beta weight of independent self-construal was not

significant (,8 = .001, t [351] = .01, p = .99, n.s.). As shown in Table 11 and 12, this was

true for both the United States and Taiwan. Independent self-construal did not carry

significant weight in predicting exchange orientation (6 = -. 14, t [174] = -1.70,
p = .09, n.s., in the United States; ,8 = .13, t [173] = 1.73, p = .09, n.s., in Taiwan).

The second multiple regression analysis, with communal orientation as the

dependent variable and independent self-construal, collective-interdependent

32

self-construal, relational-interdependent self-construal, culture and gender as independent
variables, yielded a R2 of .41, g (5, 351) = 43.57, p < .001. It indicated that the five
variables accounted for 41% (adjusted R2 = .40) of the variance in predicting communal
orientation (R2 = .27, E [4,174] = 16.36, p < .001, in the United States; R2 = .14, 15 [4, 173]
= 7.25, p < .001, in Taiwan). Results were presented in Table 13, 14 and 15

[Table 13, 14 and 15 about here]

Hypothesis 5 anticipated that collective-interdependent self-construal would be
positively related to communal orientation. For both cultures, the beta weight of
collective-interdependent self-construal was significant as expected ([3 = .15, t [351] =
2.71, p < .01). When the beta weight of collective-interdependent self-construal was
analyzed within each culture, collective-interdependent self-construal only significantly

predicted communal orientation in Taiwan (3 = .23, t [173] = 2.97, p < .01), but not in

the United States (5: .11, t [174] = 1.57, a = .12, n.s.).

Hypothesis 6 predicted that relational-interdependent self-construal would be
positively related to communal orientation. Results were partially consistent with the

hypothesis. Overall, the beta weight of relational-interdependent self-construal was

significant in the predicted direction (,8 = .17, t [351] = 3.89, p < .001). In the United

States, relational-interdependent self-construal was also significantly predictive of

communal orientation (6 = .26, t [174] = 3.87, p < .001). In Taiwan, unexpectedly, the
beta weight of relational interdependent self—construal was not significant (6 = .10, t [173]

= 1.29, p = .20, n.s.).

33

R3: Are there any sigpificant differences in exchange and communal orientations among

 

participants in the United States and TM

It was asked if there are any significant differences in exchange and communal
orientations among participants in the United States and Taiwan. As shown in Table 16,
Americans scored higher on exchange orientation (M = 2.60, S_D = .58) than did
Taiwanese (M = 2.43, S_D_ = .56 with t [355] = 2.74, p < .01). Moreover, Americans also
showed higher scores on communal orientation (M = 4.08, SD = .49) than did Taiwanese
(M = 3.43, SD = .56 with; [355] = -.11.77, p < .001). Within each culture, both
Americans and Taiwanese emphasized communal orientation over exchange orientation.

[Table 16 and Figure 6 about here]

Additional Analysis

Hypothesis 4 predicted that independent self-construal would be positively related
to exchange orientation. Regardless of hypothesis 4, the regression analysis showed one
interesting result. That is, independent self-construal was positively related to communal

orientation across cultures (B = .18, t [351] = 3.01 , p < .01), and in the United States (B

= .26, t [174] = 3.69, p < .001). Possible explanations for this result are provided in

discussion.

34

Chapter 6

Discussion

The purpose of this study was to seek a fuller understanding of the self-construal
construct. The results of the current study show that the construct of self-construal is
more complex than previously conceptualized by Markus and Kitayama (1991).
Distinguishing the three dimensions of self-construal, rather than the two dimensions,
provides an alternative way to define the self in relation to others.

Several major findings in this study are summarized here. First, cultural and
gender differences were unlikely to be characterized by the same dimensions of the self.
Second, two forms of interdependence predicted preference for communal orientation,
but manifested different effects in two cultures. Third, people with an independent
self-construal showed a preference for communal orientation, rather than exchange
orientation. Finally, Americans scored relatively higher on exchange and communal
orientations compared to Taiwanese. A further discussion of these findings will explain
why the three dimensions of self-construal may provide a fuller understanding of
self-construal construct.

Consistent with previous research findings (Kashima et al., 1995; Wang et al,
2000), this study indicated that cultural and gender effects did not have a great deal of
overlap. Specifically, cultures were best differentiated on independent and
collective-interdependent self-construals. Men and women were more likely to be
differentiated on relational-interdependent self-construal even though the effect size for

gender was minimal.

35

These results suggest that self-construals may be quite sensitive to relationship
types. More precisely, when personal relationships are close, cultural influence on
self-construals may become minor. For example, compared the two types of
interdependence, this study revealed that the major dimension that distinguished
Taiwanese from Americans was collective-interdependent self-construal. On the other
hand, no cross-cultural differences were found in the relational dimension of the self. So,
it is possible that the relationships with people with whom we are neither extremely close
nor extremely distant would illustrate cultural differences in self-construals (Park et al.,
1999)

In addition, gender differences also indicated that self-construals may vary
according to the change of relationships. For instance, although American men construed
themselves as highly independent (Madson et al., 1997), the current study found that
American men held a more group-oriented interdependence, while women held a more
relational-centered interdependence. Similarly, Gabriel et al. (1999) reported that
American men were more likely to put their personal desires aside to help their group
compared to American women. However, these differentiations were not significant
between Taiwanese men and women in this study. The further investigation for gender
differences in two types of interdependence outside the United States might be needed.

Based on the previous findings, this research suggests that people activate
different dimensions of self-construals depending on with whom they are interacting.
Including the third construct, relational dimension of the self, allows a fuller
understanding how people shift their views of the self in different relationship types and

how the consequences of shifts may influence one’s behavior.

36

Considering independent self-construal as a predictor for relationship orientations,
this study found that people with an independent self-construal did not show a preference
for exchange orientation. On the other hand, independent self-construal was positively
related to communal orientation.

One possible explanation for these unexpected findings is that independent
self-construal may not be a good predictor for relationship orientations. As Park et al.’s
(1999) concern, the view of the self as a unique person with one’s own thoughts, feelings,
and behavior may be a universal, and natural facet of human development. Consequently,
independent self-construal is not as sensitive to relationship types as interdependent or
relational-interdependent self-construal. In this view, a person with a very independent
self-construal can be communal oriented in personal relationships.

When collective-interdependent self-construal was analyzed as a predictor, the
results, across cultures and in Taiwan, showed that people who scored higher on
collective-interdependent self-construal tended to prefer communal orientation in
personal relationships. In the United States, however, collective-interdependent
self-construal did not significantly predict any preferences for relationship orientations.

Compared to collective-interdependent self-construal, relational-interdependent
self-construal worked better in the United States than in Taiwan. The results, across
cultures and in the United States, indicated that people with a very
relational-interdependent self-construal showed a preference for communal orientation.
In Taiwan, although the association between relational-interdependent self-construal and

communal orientation was in the expected direction, the strength was not significant.

37

Together, these results showed that, across cultures, two forms of interdependence
were useful for predicting communal orientation. On the other hand, within each culture,
communal orientation was attributed to different forms of interdependence. Specifically,
whereas relational-interdependent self-construal best predicted communal orientation in
the United State, collective—interdependent self-construal accounted for communal
orientation in Taiwan. This finding is important because it reveals that both Americans
and Taiwanese concern about their connections with others, but pursue the connections in
different spheres. As argued earlier, individualistic and collectivistic cultures tend to
emphasize different dimensions of interdependent self-view. Americans are more likely
to include significant others rather than group membership into the self (Cross et al.,
2000). In other words, Americans are relational but not necessarily collectivists
(Oyserrnan et al., 2001).

Finally, regardless of self-construals, Americans scored relatively higher on
exchange and communal orientations compared to Taiwanese. This result suggests that a
person with a very independent self-view, like American participants in this study, can
behave communally in personal relationships. Thus, whether we tested communal
orientation based on one’s cultural or individual identities (self-construals), the results all
suggest that emphasizing one’s thoughts, feelings, and behavior does not mean that the
person will be unlikely to Show concern for others’ needs and feelings.

In sum, the previous discussion led us to conclude that the construct of
self-construals is more complex than previously thought. The three dimensions of
self-construal more accurately explain cultural and gender differences in self-construals

compared to the two dimensions. One important reason is because self-construals are

38

sensitive to relationship types. With a relational dimension of the self, researchers are
allowed to perceive how people from various cultures and gender shift the views of the
self when the relationship types change in different social contexts.

Conclusion

As both previous research and the current study show, the distinction of the three
dimensions of self-construal is an alternative construct of self-construals. In particular,
considering the sensitivity to relationship types, the three dimensions of self-construal
provide a fuller understanding how people construe themselves and behave differently.
More importantly, one should note that this is not the only alternative. Future research
needs to rethink the definition of self-construals commonly discussed in the literature,
and expanding the current construct of self-construals is needed.

Limitations

There are three limitations in this study. First, this research employed self-report
scales and closed-end measures. With such measurement, it is possible that people
over-report prosocial behavior and under-report socially undesirable behavior. Thus,
using a self-report scale may result in lower scores on exchange orientation.

Second, the results of confirmatory factor analyses suggest that there is a problem
with scale unidimensionality. As noted earlier, if a scale is unidimensional, all the items
in the scale should measure the same trait. However, the five scales of the study had to
delete a total of 38 (of 63) items in order to obtain unidimensionality. After deleting a
number of problematic items, all scales passed the internal consistency test, but only the
communal orientation scale passed the parallelism test. Thus, one should take this

measurement evaluation into account while reading the results of the current research.

39

Finally, this study has bearing on the issue of whether Leung and Kim’s (1997)
self-construal scale is valid to investigate one’s self-concept. Levine et al. (2001) argued
that most studies using Leung and Kim’s scale only consider scale alpha reliability as
evidence for scale validity. However, the value alpha should be interpreted only after the
scale has been demonstrated as unidimensional (Hunter et al, 1982). One study
considering both the internal consistency and parallelism of the scale revealed dropping
75% of the items in order to obtain acceptable fit (Ohashi, 2000). Similar results were
found in the current study. Moreover, Levine et a1. (2001) claimed that Leung and Kim’s
self-construal scale is, in fact, multidimensional and highly unstable.

Rather than letting these measurement difficulties impede research, investigators
may consider different research designs in the future. For example, a group of researchers
have utilized priming studies in assessing one’s self-construal (e. g., Gardner, Lee, &
Aaker, 2000; Gardner et a1, 1999; Brewer et al., 1996; Madson et al., 2001; Ybarra &
Trafimow, 1998). In addition, Matsumoto (1999) suggested the use of qualitative
methods. He argued that these investigations might tell us about self and culture in ways
that cannot be captured through questionnaire studies. Although the current self-construal
scales seem problematic and there are alternative methods to measure one’s self-construal,
it does not mean that we should abandon the self-construal scales completely. The further

refinement is still needed.

40

Chapter 7

Footnotes

. The approximate number was obtained through computer searches using ProQuest.
From 1986 to present, a total number of 101 studies include the word “self-construal”
in the text.

. Among 3704 Asian subjects, there were 1526 Japanese, 1024 Koreans, and 1134 from
Hong Kong.

. It should be noted that these three studies employed different measurements to test
self-construals. Cross (1995) used the Ego-Tasks Subscales (Breckler, Greenwald, &
Wiggins, 1986) and the Collectivism Scale (Yamaguchi, 1990, 1994). Krull et al.
(1999) used the Self-Construal Scale developed by Singelis (1994) and by Gudykunst
et al. (1996). Witte et al. (1993) used the INDCOL Scale by Hui (1988).

. Cross et al. simply asked participants if they were US. citizens or not. Anyone who
replied “no” was defined as a non-citizen. Among 2483 subjects, there were 2,374
Americans and only 109 noncitizens.

. Although the Leung and Kim’s self-construal scale (Leung & Kim, 1997) has been
argued as lacking in validity (Levine et al., 2001), this study still employed this scale
for three reasons. First, previous self-construal research only provided little evidence
from Taiwan and most Asian subjects were recruited from Japan and Korea. Second,
this study conducted the CFA and parallelism tests to check the unidimensionality of
Leung and Kim’s (1997) self-construal scale. Third, whether the problems concerning

the invalidity of self-construal stem from measurement or theory remains debatable

41

(Levine et al., 2001). The current study focused on conceptual discussion rather than
measurement modification.

. Before performing the CFA procedure, the standard score coefficient alphas of the
Communal Orientation Scale were .75 overall, .74 in the United States and .74 in
Taiwan. The standard score coefficient alphas of the Exchange Orientation Scale
were .63 overall, .62 in the United States and .70 in Taiwan. The standard score
coefficient alphas of the Relational-Interdependent Self-Construal Scale were .82
overall, .86 in the United States and .79 in Taiwan. The standard score coefficient
alphas of the Independent Self-Construal Scale were .81 overall, .79 in the United
States and .72 in Taiwan. The standard score coefficient alphas of the Interdependent
Self-Construal Scale were .87 overall, .76 in the United States and .72 in Taiwan.
After the CFA procedure, a total of thirty-eight out of sixty-three items were deleted.

However, this procedure did not cause significant reduction in reliability.

42

REFERENCES

43

References

Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as
including other in the self. Journal of Person_ality and Social Psychology. 60. 241-253.

Baumeister, R. F., & Sommer, K. L. (1997). What-d0 mmant? Gender
differences and two Spheres of belongingness: Comment on Cross and Madson.
Psychological Bulletin, 117, 497-529.

1310) 3% r W

Breckler, S. J ., Greenwald, A. G, & Wiggins, E. C. (1986, April). Public, private,
and collective self-eval_uation: Measurement of individual differences. Paper presented at
the International Research and Exchanges Board conference on Self and Social
Involvement, Princeton, U.S.

Bresnahan, M. J., Liu, W. Y., Nebashi, R., & Liao, C. C. (1999). A comparison of
response styles in Singapore and Taiwan. Journal of Cross-Cultural Psychology 30.
342-357.

Brewer, M. B., & Gardner, W. (1996). Who is the “we”? Levels of collective
identity and self-representations. Journal of Personality and Social Psychology. 71, 83-93.

Brislin, R. W. (1986). The wording and translation of research instruments. In W.
J. Lonner & J. W. Berry (Eds.), Field methods in cross-cultugral reseafih (pp. 137-164).
Beverly Hills, CA: Sage.

Clark, M. S., Dubash, P., & Mills J. (1998). Interest in another’s consideration of
one’s needs in communal and exchange relationships. Journal of Experimental Social
Psychology. 34. 246-264.

, Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and
communal relationships. Journal of Person_ality and Social Psychology. 37, 12-24.

Clark, M. 8., Mills, J., & Corcoran, D. (1989). Keeping track of needs and inputs
of friends and strangers. Personality and Social Psychology Bulletin. 15, 533-542.

44

Clark, M. 8., Mills, J., & Powell, M. ( 1986). Keeping track of needs in communal
and exchange relationships. Joumal of Personality and Social Psychology. 51. 333-338.

Clark, M. S., Ouellette, R., Powell, M., & Milberg, S. (1987). Recipient’s mood,
relationship type, and helping. Journal of Persopality and SocLapl Psychology, 53, 94-103.

Clark, M. S., & Taraban, C. B. (1991). Reactions to and willingness to express
emotion in two types of relationships. Journal of Experimental Social Psychology. 27.
324-336.

Clark, M. S., & Waddell, B. (1985). Perceptions of exploitation in communal and
exchange relationships. Journal of Sociefiand Personral RelationshipsLZ, 403-418.

Cross, S. E. (1995). Self-construals, coping, and stress in cross-cultural adaptation.
Jourpal of Cross-Cultural Psychology. 26. 637-697.

Cross, 8. E., Bacon, P. L., & Morris, M. L. (2000). The relational-interdependent
self-construal and relationships. Journal of Person_ality apd Social Psychology, 140.
210-217. 112“.) 9511:5259

Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender.
Psychological Bulletin, 122, 5-37.

Gabriel, 8., & Gardner, W. L. (1999). Are there “his” and “hers” types of
interdependence? The implications of gender differences in collective versus relational
interdependence for affect, behavior, and cognition. Journal of Persoaalipy and Social
Psychology. 77, 642-655.

Gardner, W. L., Gabriel, 8., & Lee, A. Y. (1999). “I” value freedom, but “we”
value relationships: Self-construal priming mirrors cultural differences in judgment.

Psychological Science, 10, 321-326.

Gardner, W. L., Lee, A. Y., & Aaker, J. L. (2000). The pleasures and pains of
distinct self-construals: The role of interdependence in regulatory focus. Journal of
Person_ality and Social Psychology. 78. 1122-1134.

45

Gudykunst, W., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., &
Heyman, S. (1996). The influence of cultural individualism-collectivism, self-construals,
and individual values on communication styles across cultures. Human Communication
Resgtrch. 22. 510-543.

Hartley, W. S., (1970). Manual for the twenty statements problem. Kansas City,
MO: Department of Research, Greater Kansas City Mental Health Foundation.

Hofstede, G (1980). Culture’s consequences: International differences in
work-related values. Beverly Hill, CA: Sage.

Hui, C. H. (1988). Measurement of individualism —collectivism. Journal of
Research in Personality, 22. 17-36.

Hunter, J. E. & Gerbing, D. W. (1982). Unidimensional measurement, second
order factor analysis, and causal models. Resegch in Organizational Behavior. 4.
267-320.

Kanagawa, C., Cross, S. E., & Markus, H. R. (2001). “Who am I?” The cultural
psychology of the conceptual self. Personality and Social Psychology Bulletin. 27.
90-103.

Kashima, Y., Yamaguchi, S., Kim, U., Choi, S-C, Gelfand, M. J ., & Yuki, M.
(1995). Culture, gender, and self: A perspective from individualism-collectivism research.
Journal of Personality and Social Psychology. 69. 925-937.

Kim, M. 8., Hunter, J. E., Miyahara, A., Harvath, A. M., Bresnahan, M., & Yoon,
H. J. (1996). Individual vs. culture-level dimensions of individualism and collectivism:
Effects on preferred conversational styles. Communication Monoggaphs, 63, 29-49.

Kitani, K., & Kim, M. S. (1996). The effects on self-construals on conflict
management styles in romantic relationships. Paper presented at the International
Communication Association ( International and Intercultural Communication Division)
Chicago, U.S.

Krull, D. S., Loy, M. H. M., Lin J., Wang, C. F., Chen, S., & Zhao, X. (1999). The
fundamental attribution error: Correspondence bias in individualist and collectivist

46

cultures. Personality and Social Psychology Bulletin. 25, 1208-1219.

Leung, T., & Kim, M. S. (1997). A revised self-construal scale. Honolulu, HI:
University of Hawaii at Manoa.

Levine, T. R., Bresnahan, M. J ., Park, H. S., Lapinski, M. K., & Wittenbaum, G.
M. (2001). Self-constraal lackLvalidity. Unpublished manuscript. East Lansing, MI:
Michigan State University.

Lin, Y. W., & Rusbult, C. E. (1995). Commitment to dating relationships and
cross-sex friendships in America and China. Journal of Social and Personal Relationships,
_1_2, 7-26.

Madson, L., & Trafimow, D. (2001). Gender comparisons in the private,
collective, and allocentric selves. Journfi of Social Psychology. 141, 551-557.

’ Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for
cognition, emotion, and motivation. Psychological Review. 98, 224-253.

Matsumoto, D. (1999). Culture and self: An empirical assessment of Markus and
Kitayama’s theory of independent and interdependent self-construals. _A__s_i_an Journal of
Social Psycholtgy, 2, 289-310.

McPartland, T. S., Cumming, J. H., & Garreston, W. (1961). Self-conception and
ward behavior in two psychiatric hospitals. Sociometry. 21, 111-124.

Oetzel, J. G. (1998 a). Explaining individual communication processes in
homogeneous and heterogeneous groups through individualism-collectivism and
self-construal. Human Communication Research 25, 202-224.

 

Oetzel, J. G (1998 b). The effects of self-construal and ethnicity on self-reported
conflict styles. Communication Reports, 11, 133-144.

Ohashi, R. (2000). High/low-context communication: Conceptualization and scale
development. Unpublished doctoral dissertation. East Lansing, MI: Michigan State

University.

 

47

Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2001). Rethinking

individualism and collectivism: Evaluation of theoretical assumptions and meta—analyses.
Unpublished manuscript. Ann Arbor, MI: University of Michigan.

 

Park H. S., & Levine T. R. (1999). The theory of reasoned action and
self-construal: Evidence from three cultures. Communication Monograpth, 66, 199-218.

Sato, T., & Cameron, J. E. (1999). The relationship between collective self-esteem
and self-construal in Japan and Canada. Journal of Social Psychology, 139. 426-435.

Sharkey, W. F., & Singelis, T. M. (1995). Embarrassability and self-construal: A
theoretical integration. Personality and In_dividual Differences. 19, 919-926.

Singelis, T. M. (1994). The measurement of independent and interdependent
self-construals. Persopality and Social Psychological Bulletin, 20, 580-591.

Singelis, T. M., Bond, M. H., Sharkey, W. F., & Lai, C. S. Y. (1999). Unpackaging
culture’s influence on self-esteem and embarrassability: The role of self-construals.
Journal of Cross-Cultural Psychology. 30, 315-341.

Singelis, T. M., & Brown, W. J. (1995). Culture, self, and collectivist
communication: Linking culture to individual behavior. Human Communication
Research. 21, 354-389.

Singelis, T. M., & Sharkey, W. F. (1995). Culture, self-construal, and
embarrassability. Journal of Cross-Cultural Psychology. 26, 622-644.

. - Trafimow, D., Triandis, H. C., & Goto, S. G (1991). Some tests of the distinction
between private self and the collective self. Journal of Personaliy and Social Psychology,
_60, 649-655.

Triandis, H. C. (1989). The self and social behavior in differing cultural contexts.
Psychological Review, 96, 506-520.

48

' Triandis, H. C., Bontempo, R., Villareal, M., Asai, M., & Lucca, N. (1988).
Individualism-collectivism: Cross-cultural perspective on self-in —group relationships.
Journflf Perscmality and Social Psychology, 54, 323-338.

Vanyperen, N. W., & Buunk, B. (1991). Equity theory and exchange and
communal orientation from a cross-national perspective. Journal of Social Psychology.
1M, 5-20.

/ fWang, C. L., Bristol, T., Mowen, J. C., & Chakraborty, G. (2000). Alternative
models of self-construal: Dimensions of connectedness-separateness and advertising
appeals to the culture and gender-Specific self. Journal of Con_surner Psycholggy. 9.
107—115.

Witte, K., Murray L., Liu, W. Y., & Hubbell, A. P. (1993, November). Addressing
cultural orientations in fear appeals: PromotingAIDS-jaotective behaviors among
Hispanic immigrant and African-American adolescents, and American and Taiwanese
college students. Paper presented as the annual meeting of the Speech Communication
Association, Miami, U.S.

Yamaguchi, S. (1990). Personality and cognitive correlates of collectivism among
the Japanese: Validation of collectiviam sca1_e. Unpublished paper, University of Tokyo,
Department of Social Psychology.

Yamaguchi, S. (1994). Collectivism among the Japanese: A perspective from the
self. In U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon (Eds.),
Individualism and collectivism: Theory. method, and applications (pp. 175-188).
Thousand Oaks, CA: Sage.

Ybarra, O., & Trafimow, D. (1998). How priming the private self or collective
self affects the relative weights of attitudes and subjective norms. Personality and Social
Psychology Bulletin. 24, 362-370.

Yum, Y. 0., & Canary, D. J. (2000, November). The effect of culture m
exchange orientation on problem-solving behavior in close relationships: Europeaa
Americans and Koreana. Paper presented at the National Communication Conference,
Seattle, US.

49

APPENDICES

50

 

Table 1: T-Test for Means of Self-Construals between the United States and

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Taiwan
U.S. Taiwan
(n=179) (n=178)
Self-Construal M SD M SD t value r
Independent 4.04 .45 3.18 .49 -17.26*** -.68
Collective-Interdependent 2.94 .55 3.71 .43 l4.78*** .62
Relational-Interdependent 3.89 .50 3.84 .45 -.96 -.05
*** p < .001
Figure 1: Means for Self-Construals across Countries
I U.S. El Taiwan
5 4 04
4 , - 3.7L 3.89 3.84
2.94
3 .
2 .
1 .
o , 1
Independent Self- Collective- Relational-
Construal Interdependent Self- Interdependent Self-

Construal

51

Construal

 

Table 2: T-Test for Means between Types of Self-Construals in the United
States

 

Table 3: T-Test for Means between Types of Self-Construals in Taiwan
value r
1

1

 

Figure 2: Means for Self-Construals within Countries

 

El Independent Self-
Construal

 

I Collective-
Interdependent Self-
Construal

Relational
Interdependent Self-
construal

 

 

 

 

 

U.S. Taiwan

52

 

Table 4: T-Test for Means of Self-Construals between Men and Women

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Men Women
(n=128) (n=229)

Self-Construal M SD M SD t value r
Independent 3.52 .64 3.66 .63 -1.96 -.10
Collective-Interdependent 3.39 .62 3.29 .63 1.48 .08
Relational-Interdependent 3.79 .50 3.91 .45 -2.37* -. 12
* p < .05
Fi ure 3: Means for Self-Construals of Men and Women

I Men E] Women

3.79 3.91

 

GANG-ha!

 

 

 

3.39 3.29

 

Construal

Independent Self-

 

Collective-

lnterdependent Self-

Construal

53

 

 

 

 

 

 

 

 

Relational-
lnterdependent Self-

Construal

 

 

Table 5: T-Test for Means of Self-Construals between Men and Women in tap

 

 

 

 

 

 

 

 

 

 

United States
Men Women
(n=60) (n=1 19)
Self-Construal M SD M SD t value r
Independent 3.95 .52 4.08 .41 -1.91 -.14
Collective-Interdepgndent 3.07 .56 2.88 .53 2.21 * .16
Relational-Interdependent 3.78 .57 3.93 .46 -1.74 -.13

 

 

 

*p<.05

Fi ure 4: Means for Self-Construals of Men and Women in the United States

 

 

I Men 121 Women

 

 

 

3.78 3-93

 

 

 

 

 

3.07 2.88

 

5 3_g5 4.08

4 .

3 .

2 .

1

O .

Independent Self-
Construal

 

 

Collective-

Interdependent Self-

Construal

54

 

 

 

 

 

 

 

 

Relational-
lnterdependent Self-
Construal

 

Table 6: T-Test for Means of Self-Construals between Men and Women in
Taiwan

 

 

 

 

 

 

 

 

 

Men Women

(n=68) (n=1 10)
Self-Construal M SD M SD t value r
Independent 3.14 .47 3.20 .50 -.72 -.05
Collective-Interdependent 3.68 .52 3.74 .37 -.85 -.06
Relational-Interdependent 3.77 .44 3.88 .45 -1.54 -.12

 

 

 

 

 

Figure 5: Means for Self-Construals of Men and Women in Taiwan

 

I Men El Women

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5
4 3.68 3.74 3.77 3-88
3.14 3.2 . ‘
3
2
1 .
O , l 1
Independent Self- Collective- Relational-
Construal Interdependent Self— Interdependent Self-
Construal Construal

55

 

Table 7: Two-way ANOVA Results for Independent Self-Construal

 

 

 

 

 

 

Independent Sum of DF Mean F p
chiares Square
Country 65.08 1 65.08 295.86 < .001 .45
Gender .72 1 .72 3.28 .07 .01
Country by .14 1 .14 .62 .43 .00
Gender
Residual 77.65 353 .22
Total 144.42 356 .41

 

 

 

 

 

 

 

Eagle 8: Two- ay ANOVA Results for Collective-Interdependent

 

 

 

 

 

 

 

Self-Construal
Interdependent Sum of DF Mean F p
Squares Square
Country 52.37 1 52.37 218.90 < .001 .38
Gender .32 l .32 1.36 .25 .00
Country by 1.23 1 1.23 5.15 .02* .01
Gender
Residual 84.46 353 .24
Total 138.92 356 .39

 

 

 

 

 

 

 

 

Table 9: Two-way ANOVA Results for Relational-Interdependent

 

 

 

 

 

 

Self-Construal
Relational Sum of DF Mean F p
Interdependent Squares chiare
Country .16 1 .16 .73 .39 .00
Gender 1.20 1 1.20 5.38 .02 .01
Country by .02 1 .02 .09 .77 .00
Gender
Residual 78.62 353 .22
Total 80.05 356 .23

 

 

 

 

 

 

 

 

56

 

 

 

 

Table 10: Standardized Multiple Regression Coefficients Predicting
Exchange Orientation—Overall

 

 

 

 

 

 

b t p
Independent Self-Construal .00 .01 .99
Collective-Interdependent Self-Construal -.16 -2.23 .03
Relational-Interdependent Self-Construal -.O3 -.58 .56
Cultural .05 .62 .54
Gender -.04 -.73 .47

 

 

 

 

 

N = 357

Regression Model: R2 = .04, F (5, 351) = 3.02, p < .05
Respondent’s culture was coded Taiwan = 0, American = 1.
Gender was coded male = 0, female = 1.

57

Table 11: Standardized Multiple Regression Coefficients Predicting
Exchange Orientation—U.S.

 

 

 

 

 

 

 

 

 

b t p
Independent Self-Construal -. 14 -1.70 .09
Collective-Interdependent Self-Construal -.04 -.47 .64
Relational-Interdependent Self-Construal -.02 -.23 .82
Gender .03 .41 .68

 

N = 179
Regression Model: R2 = .02, F (4, 174) = .83, p = .51, n.s.
Gender was coded male = 0, female = I.

Table 12: Standardized Multiple Regression Coefficients Predicting
Exchange Orientation—Taiwan

 

 

 

 

 

 

 

 

 

b t p
Independent Self-Construal .13 1.73 .09
Collective-Interdependent Self-Construal -.28 -3.54 .001
Relational-Interdtmendent Self-Construal .00 .04 .97
Gender -.08 -1.05 .28

 

N = 178
Regression Model: R2= .11, F (4, 173) = 5.41, p < .001.
Gender was coded male = 0, female = l.

58

 

 

 

Table 13: Standardized Multiple Regression Coefficients Predicting

Communal Orientation—Overall

 

 

 

 

 

 

b t
Independent Self-Construal .18 3.01 .003
Collective-Interdependent Self-Construal .15 2.71 .007
Relational-Interdependent Self-Construal .17 3.89 < .001
Culture .49 7.94 < .001
Gender .21 5.11 < .001

 

 

 

 

 

N = 357

Regression Model: R2 = .41, F (5, 351) = 48.57, p < .001
Respondent’s culture was coded Taiwan = 0, American = 1.
Gender was coded male = 0, female = l.

59

 

 

Table 14: Standardized Multiple Regression Coefficients Predicting
Communal Orientation—U.S.

 

 

 

 

 

 

 

 

b t
Independent Self-Construal .26 3.70 < .001
Collective-Interdependent Self-Construal .11 1.57 .12
Relational-Interdependent Self-Construal .26 3.87 < .001
Gender .28 4.22 < .001

 

 

N = 179
Regression Model: R2 = .27, F (4, 174) =16.36, p < .001.
Gender was coded male = 0, female = 1.

Table 15: Standardized Multiple Regression Coefficients Predicting
Communal Orientation—Taiwan

 

 

Independent Self-Construal .07 .99 32

 

Collective-Interdependent Self-Construal .23 2.97 .003

 

Relational-Interdependent Self-Construal .10 1.29 20

 

 

 

 

Gender .21 2.93 .004

 

 

N = 179
Regression Model: R2= .14, F (4, 173) =7.25, p < .001.
Gender was coded male = 0, female = 1.

60

Eagle 16: T-Test for Means of Exchange and Communal Orientations
between the United States and Taiwan

 

 

 

 

 

 

 

 

 

 

 

 

 

U.S. Taiwan
(n=l 79) (n=1 78)
M SD M SD t value r
Exchange Orientation 2.60 .58 2.43 .56 2.74" .14
Communal Orientation 4.08 .49 3.43 .56 -11.77*** -.53
** p < .01
*** p < .001

Figure 6: Means for Exchange and Communal Orientations in the United
States and Taiwan

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I U.S. ElTaiwan

5
4
3 2.43
2 .
1
O , .

Exchange Orientation Communal Orientation

61

QUESTIONNAIRE
1. Abbreviations:

COM: Communal Orientation Scale

EX: Exchange Orientation Scale

IND: Independent Self-Construal Scale

INTER: Interdependent Self-Construal Scale

RISC: Relational-Interdependent Self-Construal Scale

2. *ltems retained after the CFA procedure.
INSTRUCTIONS

The purpose of this questionnaire is to examine people’s attitude toward personal
relationships. DO M WRITE YOUR NAME ON THIS QUESTIONNAIRE. For each
of the items below, please read carefully and mark one response option that best
represents how much you agree/disagree with the statement that is presented.

EXAMPLE:

 

Strongly Disagree Neutral Agree Strongly
Disagree Agree

 

1. What happens to me is my own doing. x

 

 

 

Strongly Disagree Neutral Agree Strongly
Disagree Agree

 

1. It bothers me when other people neglect
my needs. [COM]

 

2. When making a decision, I take other
people’s needs and feelings into
account.[COM]

 

3. I’m not especially sensitive to other
people’s feelings. [COM*]

 

4. I don’t consider myself to be a
particularly helpful person. [COM*]

 

5. I believe people should go out of their
way to be helpful. [COM]

 

6. I don’t especially enjoy giving others aid.
[COM*]

 

7. I expect people I know to be responsive
to my needs and feelings. [COM]

 

 

8. I often go out of my way to help another
person. [COM*]

 

 

 

 

 

 

 

62

 

Strongly
Disagree

Disagree

Neutral

Agree

Strongly
Agree

 

9. I believe it’s best not get involved taking
care of other people’s personal needs.
[COM]

 

10. I’m not the sort of person who often
comes to the aid of others. [COM]

 

11. When I have a need, I turn to others I
know for help. [COM]

 

12. When people get emotionally upset, I
tend to avoid them. [COM]

 

13. People should keep their troubles to
themselves. [COM]

 

14. When I have a need that others ignore,
I’m hurt. [COM]

 

15. When I give something to another
person, I generally expect something in
return. [EX*]

 

16. When someone buys me a gift, I try to
buy that person as comparable a gift as
possible. [EX]

 

17. I don’t think people should feel
obligated to repay others for favors.
[EX*]

 

18. I wouldn’t feel exploited if someone
failed to repay me for a favor. [EX*]

 

19. I don’t bother to keep track of benefits I
have given others. [EX*]

 

20.When people receive benefits from
others, they ought to repay those others
right away. [EX]

 

21. It’s best to make sure things are always
kept “even” between two people in a
relationship. [EX]

 

22.1 usually give gifts only to people who
have given me gifts in the past. [EX]

 

23.When someone I know helps me out on a
project, I don’t feel I have to pay him or
her back. [EX]

 

 

24. My close relationships are an important
reflection of who I am. [RISC]

 

 

 

 

 

 

63

 

 

Strongly
Disage

Disagree

Neutral

Agree

Strongly
Agree

 

25.

When I feel very close to someone, it
often feels to me like that person is an
important part of who I am. [RISC*]

 

26.

I usually feel a strong sense of pride
when someone close to me has an
important accomplishment. [RISC*]

 

27.

I think one of the most important parts
of who I am can be captured by looking
at my close friends and understanding
who they are. [RISC*]

 

28.

When I think of myself, I often think of
my close fi'iends or family also. [RISC]

 

29.

If a person hurt someone close to me, I
feel personally hurt as well. [RISC*]

 

30.

In general, my close relationships are an
important part of my self-image.
[RISC*]

 

31.

Overall, my close relationships have
very little to do with how I feel about
myself. [RISC]

 

32.

My close relationships are unimportant
to my sense of what kind of person I am.
[RISC]

 

33.

My sense of pride comes from knowing
who I have as close friends. [RISC]

 

34.

When I establish a close relationship
with someone, I usually develop a
strong sense of identification with that
person.[RISCfl

 

35.

I should be judged on my own merit.
[IND*]

 

36.

1 voice my own opinion in group
discussions. [IND]

 

37.

I feel uncomfortable disagreeing with
rpy group. [INTER]

 

 

38.

I conceal my negative emotions so I
won’t cause unhappiness in my group.
[INTER*]

 

 

 

 

 

 

64

 

 

Strongly
Disagree

Disagree

Neutral

Agree

Strongly
Agree

 

40.

I prefer to be self-reliant rather than
dependent upon others. [IND]

 

41.

I act as a unique person, separate from
others. [IND]

 

42.

I don’t like depending upon others.
[IND]

 

43.

My relationships with others in my
group are more important than my
personal accomplishments. [INTER]

 

44.

My happiness depends on the happiness
of those in my group. [INTER*]

 

45.

I often consider how I can be helpful to
specific others in my group. [INTER]

 

46.

I take responsibility for my own actions.

[IND]

 

47.

It is very important for me to act as an
independent person. [IND]

 

48.

I have an opinion about most things; I
know what I like, and I know what I
don’t like. [IND*]

 

49.

I enjoy being unique and different from
others. [IND*]

 

50.

I don’t change my opinions in
conformity with those of the majority.
[1ND*1

 

51.

Speaking up in a work/task group is not
a problem for me. [IND*]

 

52.

Having a lively imagination is important
to me. [IND]

 

53.

Understanding myself is a major goal in
my life. [IND]

 

54.

I enjoy being admired for my unique
qualities. [IND]

 

55.

I am careful to maintain harmony in my
group. [INTER*]

 

 

56.

When I’m with my group, I watch my
words so I won’t offend anyone.

[INTER*]

 

 

 

 

 

 

65

 

 

Strongly
Dispgree

Disagree

Neutral

Agree

Strongly
Agree

 

57.

I would sacrifice my self-interest for the
benefit of my group. [INTERL

 

58.

I try to meet demands of my group, even

if it means controlling my own desires.
[INTER*]

 

59.

It is important to consult close friends
and get their ideas before making
decisions. [INTER]

 

60.

I should take into consideration my
parents’ advice when making education
and career plans. [INTER]

 

61.

I act as fellow group members would
prefer me to. [INTER]*

 

62.

The security of being an accepted
member of a group is very important to
me. [INTER]

 

 

63.

If my brother or sister fails, I feel
responsible. [INTERL

 

 

 

 

 

 

Personal Data

1.

9'95"!"

Male Female

Age

 

Are you an international student?

Nationality: U.S.A. Other (please specify)

 

Ethnic Identity:

_ Caucasian _ Asian/Pacific Islander
_ African-American _ Native American
__ Hispanic _ Other (please specify)

66

 

CONSENT FORM

We are interested in the effects of culture and self-construals on personal
relationships. You will be asked to complete a questionnaire that asks about your attitude
toward interpersonal relationships. Then, you will be asked to provide some information
about your background. The total time necessary is about 15 to 20 minutes.

*Participation in this study is voluntary, and you may withdraw your consent to
participate at any time without penalty.

*You will receive one point of extra credit for your participation in this study. This extra
point is not transferable to any other course.

*Your participation in this study does not guarantee any beneficial results to you.

*You have the right to have this study explained to you to your satisfaction.

*Your privacy will be protected to the maximum extent allowable by law. The results of
this study will be treated with strict confidence with regard to the data of any participant.

With this restriction, the results will be made available to you at your request.

*The data you provide to the researcher may be used by other scientists for secondary
analysis. Again data will be treated with strict confidence.

Should you have any questions, problems, complaints, or if you desire further
information, you have the right to contact the following person.

 

For questions about this study For questions about your rights as a
please contact human subject of research please contact
Hui-Chih Chiu Dr. David E. Wright

Email: chiuhuic@msu.edu Email: ucrihs@msu.edu

Department of Communication UCRIHS

Michigan State University 246 Administration Building

East Lansing, MI 48824 Michigan State University

(517) 355-4004 East Lansing, MI 48824

(517) 355-2180
Please indicate if you are willing to participate in this study by signing your name.
Signature Printed Name

 

 

Date Student Number

67

|IuAN HIAIF LIBHAPIL

 
   

MHH (
t ‘ .

3 1293 02318 2532