Reaction to conflict in Chinese/U.S.A. interrelationships

MISSING IMAGE

Material Information

Title:
Reaction to conflict in Chinese/U.S.A. interrelationships the nature of discontented responces
Physical Description:
viii, 130 leaves : ill. ; 28 cm.
Language:
English
Creator:
Ma, Chen-Lung Ringo, 1952-
Publication Date:

Subjects

Subjects / Keywords:
Intercultural communication   ( lcsh )
Relations -- United States -- China   ( lcsh )
Relations -- China -- United States   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1987.
Bibliography:
Includes bibliographical references (leaves 100-103).
Statement of Responsibility:
by Chen-Lung Ringo Ma.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000968193
notis - AEU3409
oclc - 17394267
System ID:
AA00003792:00001


This item is only available as the following downloads:


Full Text














REACTION TO CONFLICT IN CHINESE/U.S.A. INTERRELATIONSHIPS:
THE NATURE OF DISCONTENTED RESPONSES







BY

CHEN-LUNG RINGO MA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA


1987















ACKNOWLEDGMENTS


Without being offered a graduate teaching

assistantship by the Department of Speech, completion of my

doctoral program would have been impossible. The present

dissertation could not have been written either. I have

never taken this opportunity for granted, since I know my

potential is from time to time overshadowed by my

linguistic and cultural background. Only bravest educators

make such an offer.

Undoubtedly Dr. Donald E. Williams, the supervisory

committee chair, has played an important role in my life.

His encouragement, endeavor, and empathy expedited my

dissertation research and made it a more enjoyable

experience. In addition, I have learned much from his work

ethics, critical methods, pedagogical philosophy, and

communication skills.

Several scholars have expanded the universe of my

knowledge in specific ways. Dr. Anthony J. Clark

enlightened me about research alternatives, and provided me

an opportunity for developing my quantitative research

skills in a significant research project. Dr. Lynne Webb's

positive feedback and realistic suggestions fit my









cognitive style quite well. Dr. Mark P. Becker armed me

with many statistical methods, which have become an

important asset in my academic life.

I wish to express my appreciation and respect to Dr.

Thomas B. Abbott and Dr. Chauncey C. Chu for their guidance

and friendship while serving on my supervisory committee.

I also wish to thank Drs. Robert C. Ziller and Mary A.

Fukuyama for making valuable input toward the design of

this study at its inception.

My special acknowledgment, with gratitude and love, is

to my mother, Hsiu-Wen Chang Ma, and my wife, Shu-Wen

Louisa Ma. Their never-ending support, encouragement, and

confidence in me make seemingly remote goals accessible.

The assistance of my brother, Chen-Chi Albert Ma, and

brother-in-law, Shu-Cheng Chen, in the task of collecting

data for this study is also greatly appreciated.

Finally, I wish to thank Drs. Norman N. Markel,

Kenneth J. Gerhardt, Rebecca J. Cline, Mr. Gerald Kish,

Miss Anita Raghavan, as well as other faculty and staff

members in the Department of Speech, for their assistance

and friendship in a variety of occasions. All the subjects

participating in my research have my gratitude, though

their names are unknown.


iii

















TABLE OF CONTENTS


Page


ACKNOWLEDGMENTS . .


ABSTRACT . . .

CHAPTER

I INTRODUCTION . .

The Purpose: Its Significance and Rationale
Definitions of Terms . .
Generalizability Considerations .
Literature Review . .
Hypotheses . .


II METHODOLOGY . .

Research Design . .
Preliminary Study . .
Validity and Reliability Considerations
Instrument . .
Episodes Used in Major Study .
Subjects . .
The Assistant in Taiwan .
Analysis of Results . .

III RESULTS . .


. 21
. 26
. 37
. 39
. 42
. 44
. 48
. 49


. 51


Results of Testing Hypotheses .
Summary . .


. 51
. 76


IV DISCUSSION . . 81

Conclusions . . 81
Limitations of the Study . .. 88
Implications for Future Research. .. .94

REFERENCES . . 100


. vi












Page
APPENDICES

A REQUEST FOR PROJECT APPROVAL FROM UNIVERSITY
OF FLORIDA INSTITUTIONAL REVIEW BOARD 105

B LETTER APPROVING PROJECT FROM UNIVERSITY OF
FLORIDA INSTITUTIONAL REVIEW BOARD 108

C QUESTIONNAIRE TO ASSESS DISCONTENTED RESPONSES. 110

D SUBJECTS' RAW SCORES . 125

BIOGRAPHICAL SKETCH . ... .129















A Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


REACTION TO CONFLICT IN CHINESE/U.S.A. INTERRELATIONSHIPS:
THE NATURE OF DISCONTENTED RESPONSES

by

Chen-Lung Ringo Ma

August 1987

Chair: Donald E. Williams
Major Department: Speech

This study investigated possible impact of the

cultures of the Republic of China and the United States of

America on the expression of discontented responses within

these cultures, with regard to three relationships of

communication interactants, "intimate," "acquaintance," and

"stranger," and three determinants of discontent, "injury,"

"disappointment," and "disagreement." Previous studies

suggest that, in contrast to Chinese, Americans rely on

more explicit communication codes and are less likely

influenced by communication contexts. Investigations also

indicate that while "disagreement" prompts less strong

discontented responses than feelings of "injury" or

"disappointment" by people in both cultures, Chinese

respond to "intimates" more explicitly than to

"acquaintances" or "strangers."

vi









Nine hypothetical episodes were developed from a

preliminary study; a questionnaire calling for responses to

these episodes was then structured. These episodes, chosen

on basis of high discontented response scores registered by

both American and Chinese groups, represent various

combinations of the considered relationships and

determinants. One hundred and fifty American and Chinese

students at major universities in Florida and Taiwan were

asked to respond to the nine episodes, written in their

respective native languages, by indicating their degree of

discontent on a Likert-type scale. Subjects in each

culture were classified in three groups according to

contrasting variations in the cultural memberships of

"offenders" and in the cultural settings of episodes.

Their response scores were analyzed by conducting repeated-

measures analyses, accommodating one between-subjects

variable (groups), and two within-subjects variables

(relationships and determinants). Scheffe confidence

intervals were constructed for multiple comparisons of

means.

Posited conclusions include (1) in intracultural

situations, Americans are not significantly different from

Chinese in expressing discontented responses, when

relationships and determinants are experimentally

controlled; (2) discontented responses by both Americans

and Chinese are not influenced by the culture represented


vii









by the "offender," or by the cultural setting; (3)

discontented responses by Chinese as well as by Americans

do not appear to be influenced on the basis of their

relationships with communication interactants; (4)

"disagreement" generates less strong responses than

"injury" or "disappointment" in more circumstances for

Chinese than for Americans.


viii















CHAPTER I
INTRODUCTION



The Purpose: Its Significance and Rationale

In stressing the importance of human communication,

Emmert and Donaghy (1981) emphasized that "communication is

a tool we use to cope with our environment" (p. 9). Many

factors in our physical environment, such as the climate

and geographical location that usually imply a limited

choice of means to satisfy our physiological needs,

accommodate a particular group of people who coexist in the

same environment for a long time, developing many shared

techniques by which they can communicate to one another.

There are other factors, such as norms and values, that are

evolved from our socio-psychological environment. These

factors make it possible for people in the same environment

to establish and utilize carefully observed guidelines for

communication. Most of these guidelines are a result of

their continuous communication per se--they are reinforced

through the process of communication. If the broad term,

"culture," is suggested by this "mutual-sharing" phenomenon

(as a consequence of human as well as non-human factors),

culture is not merely an important factor in communication,

but is indeed an omnipresent form of communication. A

1










2

likelihood for communication breakdown to occur between

people from different cultures who adopt different

guidelines is also obvious.

A good example for adoption of different guidelines

for communication can be found between Americans (people of

the United States of America) and Chinese. Berne (1964),

in his transactional analysis book, Games People Play,

proposed that a person's behavioral changes are "often

accompanied by shifts in feeling." Therefore the games

people play are ascertained if their motives can be

determined. A more accurate title for this book, according

to Stover (1974), should be "Games Americans Play," since

"such games as the good doctor has formulated for the

insight of his patients he would not find among Chinese

elite persons of traditional upbringing" (p. 242). The

following description can explain why the contrasting

patterns of communication between Americans and Chinese

make Berne's American "games" inapplicable to Chinese at

large:

The probing of motives is repugnant to Chinese
sensibilities. The Chinese want to know what the
other person is going to do (as a matter of
predictability), not why he does it (the unpredictable
element). Inquiry into the inner lives of others is
regarded as dirty. One need not look into these
matters in any case. The Chinese make it a conscious
policy to keep action apart from feelings. The
Chinese see two things, and keep them apart, where the
Americans see only one thing. (Stover, 1974, p. 244)










3

As disclosed in this example, the guidelines (what Berne

called "games" people play) governing communication among

Americans are not necessarily followed among Chinese.

The differences in communication between Americans and

Chinese, as well as between people from any other two

differing cultures, become a more and more important topic

as Americans and Chinese are interacting with each other

more frequently than before. As Taiwan has become the

United States' sixth-ranking trading partner since 1981,

surpassing France and Italy (Reed, 1986), the frequency of

communication between people from these two areas is

expected to increase greatly. In addition, the

normalization of the diplomatic relations between the

United States and Mainland China since 1979 has intensified

the relationships among people from these two areas.

Understanding of the similarities and differences in

communication patterns between American people and Chinese

people is urgent.

As a matter of fact, cultural differences between

Americans and Chinese are always of interest in various

academic fields. Among these differences, American

people's preference of explicit communication codes and

Chinese people's emphasis on social harmony, at the expense

of direct expression of feelings and verbal aggression,

have been discussed by many scholars (e.g., Stover, 1974,

pp. 242-263; Stover & Stover, 1976, pp.202-207; Toupin,


j











1980; Wolfson & Norden, 1984). Yet although comparisons

have been made between these two cultures, no study

provides empirical data regarding communication behavior

between Americans and Chinese in an intercultural context.

Discontented responses (discontent expressed via

verbal and/or nonverbal channels) seem to be a good

indicator for implicitness as opposed to explicitness in

communication. Nomura and Barnlund (1983) found that

discontented responses were influenced by two basic

variables: the psychological nature of the triggering

discontent in the respondent, and the character of the

relationship with the other person in the interaction (p.

4). Discontented responses in Japan have been found to

differ from those in the United States, with these two

basic variables being experimentally controlled (Nomura &

Barnlund, 1983). The same investigation has not yet been

applied to Americans and Chinese.

In responding to the above-mentioned lack of insight,

the purpose of this study was to identify the effects of

the two different cultures, American culture and Chinese

culture, on communicators' discontented responses in the

context of intercultural communication, with regard to the

relationships between communication interactants, and the

determinants of discontent. The relationships between the

communication interactants were reflected by the

categories, intimates, acquaintances, or strangers, whereas









5

the determinants of discontent included perception of

injury, disappointment, and disagreement.



Definitions of Terms

"American" and "Chinese"

"American," as an adjective used in this study, means

"of the United States of America." "Americans" are limited

to the people from the United States of America. "American

culture" indicates the mainstream culture of the United

States of America (U.S.A.). Though the Chinese people in

Mainland China and Taiwan share the same cultural heritage,

different political-economic systems make it difficult to

present a consolidated overall picture for both areas.

"Chinese," as used in this study, means "of the Republic of

China (R.O.C.) on Taiwan," or "the Chinese people in

Taiwan."

"Intimate," "Acquaintance," and "Stranger"

"Intimate" denotes the relationship characterized by

close association and warm friendship between two persons,

who refer to each other as "my good friend," "my best

friend," "my close friend," "my intimate," etc.

"Acquaintance" denotes the relationship between people that

is present when people know each other's name but lack

close association and warm friendship. In other words, if

two persons know each other's name but do not refer to each

other as "my good friend," "my best friend," "my close











friend," "my intimate," etc., they are acquaintances.

"Stranger" denotes the type of relationship between people

that is characterized by their inability to identify each

other's name and their lack of previous interpersonal

encounter.

"Discontent"

"Discontent" is defined as a mental state of being

displeased or uncomfortable. To "show discontent" is to

express, verbally and/or nonverbally, this mental state to

another person.

"Injury," "disappointment," and "disagreement"

"Injury," "disappointment," and "disagreement" are the

three determinants of perceived discontent. Although there

is usually no clear borderline among these three concepts,

a unique emphasis is proposed for each of them in this

research. According to Nomura and Barnlund (1983),

"injury" occurs in a situation in which a person has been

wronged without apparent justification. "Disappointment"

pertains to failure of a person's communication interactant

to live up to his/her expectation. "Disagreement" occurs

when interpretations and/or convictions of a person's

communication interactant are significantly different from

those of him/her.

"Discontented Responses"

The "discontented response" refers to the discontent

expressed by a person via verbal and/or nonverbal channels









7

in response to a situation involving injury,

disappointment, or disagreement.

"Intercultural Communication"

According to Gudykunst and Kim (1984), "Intercultural

communication is a transactional, symbolic process

involving the attribution of meaning between people from

different cultures" (p. 14). Kim (1984), in a different

article, also indicated that "Intercultural communication

refers to the communication phenomena in which

participants, different in cultural backgrounds, come into

direct or indirect contact with one another" (p. 16). A

definition for "intercultural communication" proposed for

this study is a message-sending and message-receiving

process affected by people with different cultural

backgrounds.



Generalizability Considerations

This investigation was limited to the culture of the

United States of America as it exists in the State of

Florida, and to the Chinese culture in Taiwan, the Republic

of China, for several reasons. That subjects were

available to the author in these two places by direct

involvement or with the help of a faithful assistant was

the major reason for this limitation. However, the

generalizability of the results and conclusions of this

study is not likely to be greatly restricted by this scope









8

of data collection. The representativeness of people in

Taiwan in reflecting the Chinese culture has been

acknowledged explicitly or implicitly by previous

researchers (e.g., Alexander et al., 1986; Grove, 1982;

Schneider, 1985; Wolf, 1974; Yang, in press), though in a

strict sense this investigation does not intend to

generalize observations of the culture as it exists in

Mainland China. In Florida, the population has risen from

2.8 million people in 1950 to more than 9.5 million in 1985

(Information Please Almanac, p. 686). The population

increase between 1970 and 1980 was 2,949,000, within which

migration constituted 92% (2,712,000) (Florida Handbook,

1985-1986, p. 546). According to Proctor (1984), "A

substantial percentage of Florida's population was not born

in the state," and many people from the North and the

Middle West have settled in Florida in recent years (p.

426). Shofner (1982) also indicated that "the population

of Florida has grown largely through migration of Americans

from other states" (p. 79). Hence, it may not be

unreasonable to consider the sample of Floridians as being

representative of a larger segment of the American

population.









9

Literature Review

High-context Culture (HCC) vs.
Low-context Culture (LCC)

Hall (1976) indicated that cultures vary in the degree

to which communication relies on the context (the

"contexting process") (pp. 74-90). A high-context

communication is "one in which most of the information is

either in the physical context or internalized in the

person, while very little is in the coded, explicit,

transmitted part of the message" (Hall, 1976, p. 79). A

low-context communication is "just the opposite, i.e., the

mass of the information is vested in the explicit code"

(Hall, 1976, p. 79). The American culture, according to

Hall (1976), is "toward the low end of the scale," while

Chinese culture is an example of high-context cultures (p.

79).

The high-context culture (HCC) versus low-context

culture (LCC) theory has been investigated in several

empirical studies. Gudykunst (1983), in an exploratory

investigation, found that subjects in an HCC were more

cautious in initial interactions, had a greater tendency to

make assumptions based upon a stranger's cultural

background, and asked more questions about a stranger's

background, than did subjects in an LCC. A follow-up study

disclosed that Japanese respondents (HCC) indicated a lower

preference to use self-disclosure and interrogation as

uncertainty reduction strategies, but at the same time









10

reported higher levels of attributional confidence than did

U.S.A. respondents (LCC) (Gudykunst & Nishida, 1984). In a

later study comparing and contrasting acquaintances and

people involved in dating relationships in Japan, Korea,

and the United States, a fitted statistical model suggested

that more variance was explained in attributional

confidence among acquaintances than among people in dating

relationships for the United States participants (LCC);

however, the findings were opposite for the Japanese and

Korean participants (HCC) (Gudykunst et al., 1985).

In a project resulting in the constructing of theories

pertaining to conflict management in intercultural

contexts, Ting-Toomey (1985) claimed that "individuals in a

[sic] LCC are more likely to assume a confrontational,

direct attitude toward conflicts," and "in a [sic] HCC are

more likely to assume a non-confrontational, indirect

attitude toward conflicts" (p. 81). Associated with this

theoretical framework, Ting-Toomey (1986) engaged in a

semi-structured interview survey to investigate the core

Japanese communication patterns as viewed from an insider

perspective, and as viewed from an outsider perspective.

She found that to Japanese "concealment is used

functionally to humble oneself in front of others, to not

overexpose oneself to the public world, and to make sure

one does not hurt another person's feelings" (p. 117).









11

Without mentioning Hall's theory, Matsumoto and

Kishimoto (1983), in an emotional judgment experiment,

revealed that Japanese children's accuracy rates for

identifying anger through color photographs, prepared

corresponding to Ekman and Friesen's muscle patterns for

different emotions, was lower than identification accuracy

regarding happiness, surprise, and sadness, as well as

lower than American children's accuracies of identification

for all four of these emotions.

In summary, communication in an HCC is more likely to

be influenced by its context than communication in an LCC

is. People in an HCC tend to adopt more implicit and non-

confrontational codes; while people in an LCC prefer the

use of relatively explicit and confrontational codes.

American Culture vs. Chinese Culture

Hsu (1953) indicated that Americans are more

individual-centered, and tend to display strong

emotionality, whereas Chinese are more situation-centered,

and underplay all matters of the heart (p. 10).

According to Kluckhohn and Strodtbeck (1961), the

"harmony-with-nature" value held by Chinese people dictates

that "there is no real separation of man, nature, and

supernature," whereas the "mastery-over-nature" value

preferred by Americans demands that "natural forces of all

kinds are to be overcome and put to the use of human

beings" (p. 13). This concept has been elaborated by









12
Condon and Yousef (1975) as they provide suggestions for

facilitating intercultural communication.

As a pioneer in rhetorical analysis of Chinese

thought, Oliver (1961) discovered that the principal

concept of Taoism is harmony, which demands that persuasive

effort to be non-assertive and non-argumentative, rather

than prescriptive or categorical.

Lidin (1974), drawing from traditional philosophies in

China and West, argued that harmony is the key word to

describe the beliefs held by Chinese people, whereas

"conflict" is central to the inclinations of Western

people.

The difference between Western and Chinese concepts

of sincerity is, according to Stover (1974), that "western

[sic] sincerity is aimed at gratifying the personal

feelings of others," but "Chinese sincerity is aimed at

harmony (ho)" (p.246). Harmony is regarded as a reflection

of "the success of the on-going social system--as set

against the Westerner's psychological interest in human

motives and the honest display of feelings" (Stover &

Stover, 1976, p. 203).

Yang (1978), in a research project investigating

communicative competence, has identified the ideal

personality in Taiwan to be a non-aggressive and self-

negating type of person.










13

The harmony-based Confucian heritage, according to

Kang and Pearce (1983), makes a degree of reticence

appropriate for Asians including Chinese, Japanese, and

Koreans, that exceeds that of Americans in comparable

situations; and "the rationalistic heritage of contemporary

Americans predisposes them to dispute, self-disclose, and

express disagreement to an extent far exceeding Asians"

(p.101).

Wolfson and Norden (1984), in an interpersonal

conflict study, discovered that Chinese subjects seem to

demonstrate relatively passive responses to conflict or

threat, while North American subjects tend to react more

actively to threat. They proposed that the value the

Chinese subjects place on harmony "prevents them from

'seeing' conflict" (Wolfson & Norden, 1984, p. 163).

Cline and Puhl (1984), in an empirical study comparing

seating arrangements in the United States and Taiwan,

concluded that "a primary overriding cultural theme in the

U.S. is the valuing of competition; in Taiwan it is the

valuing of harmony" (p. 216).

Chinese, according to Schneider (1985), express less

emotion than Americans and are less likely to express

emotions with strangers or acquaintances than with close

friends in the context of their communication system.

Becker (1986), in his qualitative study, claimed that

the absence of argumentation in the Far East, including










14

China and Japan, can be explained by the social,

linguistic, and philosophical background of the people in

this area.

Proceeding from empirical data, Yang (in press)

indicated that Chinese are more restrained, cautious,

patient, and self-contained, and less impulsive, excitable,

spontaneous, and natural, than are Americans.

Although there is no indication for any collaborative

effort or coordination among the authors of the above-

mentioned research publications, they seem to have reached

a consensus--"confrontation" versus "non-confrontation"

reflects a major difference between the American way of

communication and the Chinese way of communication.

Relationships of Communication Interactants,
Determinants of Discontent, and Discontented
Responses in HCC vs. LCC

The cross-cultural study by Nomura and Barnlund (1983)

yielded a highly significant difference in preferred forms

of discontented responses (which they referred to as forms

of "interpersonal criticism") between Americans and

Japanese. Gender was found to be a non-significant factor

in the choice of different forms of discontented responses

(p. 9). In responding to a questionnaire designed to point

out preferred forms of discontented responses, Japanese

scored consistently higher in registering preference for

all forms of passive responses, such as "attempt not to

show dissatisfaction" and "express dissatisfaction to a











third person [rather than to the offending person]";

American subjects scored consistently higher in expressing

preference for all forms of active responses, such as

through "constructive suggestions," and through "sarcastic

remarks" (p. 9). In addition, the study revealed that, for

both American and Japanese subjects, more "active-

aggressing forms of criticism," as dominant responses to

the situations of "injury" and "disappointment," were

shifted "in favor of more passive forms of criticism in

cases of disagreement" (Nomura & Barnlund, 1983, pp. 9-12).

In the same study, the relationships among communication

interactants were found to reflect different roles in the

two cultures. Among Japanese, the closer the personal

relationship, the more active the forms of discontented

responses became (p.13). Yet among American people,

differences in the relationships were not linked reliably

to preferred forms of discontented responses (p. 13).

Ting-Toomey (1981), when investigating the ethnic

identity and close friendship among Chinese-American

college students, found that once friendships were

established, cultural dissimilarity did not seem to hinder

friendship relationships, nor did friendship network size

have any effect on intensity of intimacy or degree of

perceived similarity. This finding that intercultural

friendship will transcend cultural dissimilarity has been

confirmed in two later studies (Gudykunst, 1985a, 1985b).









16

In a comparative study of development of relationships

among interactants (Alexander et al., 1986), American

respondents were found to introduce new topics into a

conversation at a faster rate in initial interactions than

did Chinese. Meanwhile, Chinese respondents categorized

significantly more messages as representing "prohibited"

topics of conversation, than did American subjects. This

difference is because, according to the authors, "Oriental

cultures have a tendency to reveal information about the

self at a substantially lower rate," and the acquaintance

process for Chinese is more likely to be "circumscribed" by

felt cultural prohibitions (Alexander et al., 1986, p. 71).

On the basis of these research findings, people in an

HCC, such as the Japanese, tend to adopt passive forms of

discontented responses more frequently than do people in an

LCC, such as Americans, while people in an LCC use active

forms of discontented responses more often than people in

an HCC. The relationships among communication interactants

appear to have played different roles in causing

discontented responses for different cultures. People in

an HCC are more likely to be influenced by their

relationships with communication interactants than are

people in an LCC. Chinese, as well as Japanese, were found

to be more passive, cautious, and hesitant in

communication, especially with strangers, than were

Americans. To Chinese, the more intimate the relationship,









17

the less likely will the cultural dissimilarity influence

the communication process, and the more active will be the

discontented responses. Communicators, in general, in

spite of cultural dissimilarities, are inclined to respond

to "disagreement" less strongly than they are to "injury"

or "disappointment."

The Influences of Cultural Environments
on Communication

As an overview, a communication event, as described by

Emmert and Donaghy (1981), always occurs within a

particular environment, and "we must frequently consider

environmental effects in order to explain the interaction

that occurs within any human communication system" (pp. 42-

43).

That the environment is a factor in behavior and a

part of context has also been emphasized by Hall (1976), in

parallel with his concept of HCC and LCC (p. 84).

Gudykunst and Kim (1984) proposed that strangers are likely

to interpret behavior differently in the same setting,

since "they may have different perceptions of and

orientations toward the environment," and that

environmental influences on behavior "can be expected to

vary systematically across cultures, having less impact in

low-context cultures and more impact in high-context

cultures" (p. 34).











Summary

This review of relevant literature provides a basis

for the observation that people in an HCC (including the

Chinese culture) tend to adopt more implicit than explicit

codes in communication, and assume indirect and

nonconfrontational attitudes in conflict, while the reverse

is true for people in an LCC (including the American

culture). Harmony is the key word to summarize many

different Chinese values, but confrontation and competition

are central concepts for American people. Chinese and

Japanese communicative behaviors appear to be influenced by

their relationships with communication interactants, i.e.,

the closer the relationship is, the more active form of

responses will be adopted. Yet relationships do not seem

to indicate reliably how American subjects will respond.

That communicators respond to "disagreement" less strongly

than they do to "injury" or "disappointment" is not likely

subject to cultural variations. Environmental influences

are expected to have more impact in an HCC than in an LCC.

In brief, a central theme seems to have been evolved

from the review: In communication, Americans are expected

to adopt more explicit codes and to be less influenced by

relationships with communication interactants and

communication environments, than Chinese. While

"disagreement" is responded to less strongly than "injury"

or "disappointment" by Americans as well as Chinese,











Chinese alone respond to familiar persons and environments

more explicitly than to strangers and new environments.



Hypotheses

Based on the review of the literature related to the

differences between high-context cultures and low-context

cultures and between the American culture and Chinese

culture, relationships of communication interactants,

determinants of discontent, and discontented responses, the

following hypotheses were proposed for investigation:

Hypothesis One

When the two variables, relationships of communication

interactants and determinants of discontent, are

controlled, United States of American (U.S.A.) subjects in

the U.S.A. respond to U.S.A. communication interactants

more strongly than Republic of China (R.O.C.) subjects in

the R.O.C. respond to R.O.C. communication interactants.

Hypothesis Two

When the two variables, relationships of communication

interactants and determinants of discontent, are

controlled, responses in each of the following three

situations will reflect the same degree of discontent:

1. U.S.A. subjects responding to U.S.A. communication

interactants in the U.S.A. ("A-A in USA")

2. U.S.A. subjects responding to R.O.C. communication

interactants in the U.S.A. ("A-C in USA")










20

3. U.S.A. subjects responding to R.O.C. communication

interactants in the R.O.C. ("A-C in ROC")

Hypothesis Three

When the two variables, relationships of communication

interactants and determinants of discontent, are

controlled, responses in situation 1, described below, will

be stronger than situation 2, and responses in situation 2

will be stronger than in situation 3:

1. R.O.C. subjects responding to R.O.C. communication

interactants in the R.O.C. ("C-C in ROC")

2. R.O.C. subjects responding to U.S.A. communication

interactants in the R.O.C. ("C-A in ROC")

3. R.O.C. subjects responding to U.S.A. communication

interactants in the U.S.A. ("C-A in USA")

Hypothesis Four

When determinants of discontent are controlled, U.S.A.

subjects will respond to communication interactants of

different relationships with the same degree of discontent,

but R.O.C. subjects will respond to "intimates" more

strongly than "acquaintances" or "strangers."

Hypothesis Five

When cultural memberships of U.S.A. and R.O.C.

subjects and relationships of communication interactants

are controlled, the responses to "injury" or

"disappointment" will be stronger than "disagreement."
















CHAPTER II
METHODOLOGY



Research Design

As stated in the preceding chapter, the purpose of

this study was to identify the effects of two different

cultures, the American culture and Chinese culture, on

communicators' discontented responses in these two

different cultural environments, with regard to the

relationships of communication interactants and the

determinants of discontent. Operationally, the

relationships between four independent variables and one

dependent variable were investigated. The four independent

variables were cultural pairs, cultural environments of

communication, relationships of communication interactants,

and determinants of discontent. The dependent variable was

discontented responses, measured by a Likert-type five-

point scale.

Variables

The four independent variables investigated in this

study and the various levels of each of these variables

were










22

1. Respondent-offender cultural pairings (American-Chinese,

Chinese-American, American-American, and Chinese-

Chinese).

2. Cultural environments of communication (the Republic of

China on Taiwan, and the United States of America).

3. Relationships of the communication interactants

(intimate, acquaintance, and stranger).

4. Determinants of discontent (injury, disappointment,

and disagreement).

The dependent variable, discontented responses, was

measured by the following five-point Likert-type scale:

Concept to be rated: "I would show discontent."

Strongly Strongly
Disagree Disagree Undecided Agree Agree
1 2 3 4 5

Procedures

The first and the second independent variables yield

the formation of four respondent-offender intercultural

pairings and four respondent-offender intracultural

pairings. The four intercultural pairings are "A-C in

USA," "A-C in ROC," "C-A in ROC," and "C-A in USA." The

four intracultural pairings are "A-A in USA," "A-A in ROC,"

"C-C in ROC," and "C-C in USA." The "A-C in USA," for

example, stands for an American subject's discontented

response to a Chinese when the communication event occurs

in the United States. The "C-C in ROC" means a Chinese

subject's discontented response to another Chinese when the









23
communication event happens in the Republic of China on

Taiwan. Since intercultural communication was the major

concern of this study, and in accordance with the first

three hypotheses, only two intracultural pairings, "A-A in

USA" and "C-C in ROC," and all four of the intercultural

pairings have been investigated (see Table 1). These two

intracultural pairings can be regarded as control groups.

They provided a base for comparison and contrast with the

intercultural pairings. As a result, data from a total of

six groups corresponding to these pairings were collected.

Nine "episodes" reflecting a 3 x 3 combination of the

third and the fourth independent variables (Relationships

of Communication Interactants x Determinants of Discontent)

(see Table 2) were selected based on the results of the

preliminary study (see Preliminary Study, pp. 26-37, and

Episodes Used in Major Study, pp. 42-44).

Each subject was asked to respond by indicating the

appropriate personal reaction on the scale (see p. 22) to

each of the nine episodes in terms of the dependent

measurement. Each of the nine episodes was adapted to the

requirements of the first independent variable; i.e., the

cultural membership of the communication interactant who

caused the subject's discontent was specified. For

example, the episode reflecting the combination of

"intimate" and "injury" for American subjects in "A-C"

groups to respond was











Table 1

Investigated Respondent-Offender
Cultural Pairings


Respondent-Offender
Intercultural Pairings



Cultural Environments A-C C-A



USA A-C in USA C-A in USA

ROC A-C in ROC C-A in ROC





Respondent-Offender
Intracultural Pairings



Cultural Environments A-A C-C



USA A-A in USA

ROC C-C in ROC



My good Chinese friend, Chen, falsely accused me of
damaging a video cassette recorder (VCR): "Why did
you damage my VCR?" I told Chen that I didn't touch
it at all. Chen said, "I am sure you did. Nobody
else would use my VCR."

To accommodate the second independent variable, all

data which belonged to the "in USA" groups were collected

in the United States, and data which belonged to the "in

ROC" groups were collected in Taiwan. This means the data











Table 2

The Various Combinations of the Relationships


of Communication Interactants and the
Determinants of Discontent


Determinants of Discontent



Relationships Injury Disappointment Disagreement



Intimate Intimate + Intimate + Intimate +
Injury Disappointment Disagreement

Acquaintance Acquaintance + Acquaintance + Acquaintance
Injury Disappointment + Disagree-
ment
Stranger Stranger + Stranger + Stranger +
Injury Disappointment Disagreement



of two intercultural groups and one intracultural group

were collected in each of the two countries. In order to

clarify the location of communication for the subject, a

statement such as, "All the following episodes are presumed

to occur in the country where you are answering these

questions, NOT in any other country," appeared at the

beginning of each page of the questionnaire used to

investigate the subject's responses.











Preliminary Study
Purposes

Within each of the nine categories reflecting the

various combinations of the two variables, relationships of

communication interactants, and determinants of discontent,

there might be variations in terms of the degree of

offensiveness to subjects. Different degrees of

offensiveness were likely to elicit different responses,

which would cloud the differences among the nine

categories. In order to eliminate the possibility of using

a more offensive episode for one category and a less

offensive one for another, it was decided that within each

category an episode with high degree of offensiveness to

both Americans and Chinese would be created, although the

degree of offensiveness might not be as intense in one

culture as in the other. Only a preliminary survey could

assure the achieving of this goal.

In addition, a preliminary survey served the purpose

of testing the feasibility of the original design,

providing a preview of the study, and suggesting directions

for improving the major study.

Methods

Twenty-five University of Florida students enrolled

in a junior-level "Public Speaking" class were asked to

recall and write down episodes which could be categorized

as causing or reflecting "injury," "disappointment," or










27

"disagreement." A definition of each of these three terms

was provided at the beginning of the experiment. Students

were instructed that these episodes were to represent

happenings in their real lives, but were not limited to

recent happenings. They were also requested to choose a

category from the three types of relationships, "intimate,"

"acquaintance," or "stranger," to identify their relation

to the offender in each of their episodes. Each of these

students recalled at least one episode reflecting each of

the three determinants of discontent. The collected

episodes were classified according to the nine above-

mentioned categories, with some questionable or unworkable

episodes deleted. Deletions were made when cited instances

referred to any one of these three situations: (1) When

two, or more than two, episodes were identical or similar,

only one was retained. (2) When a special relationship,

such as parent-child or boyfriend-girlfriend, was involved,

deletion was dependent on whether the episode could also

apply to an intimate friend. It was retained if change of

the relation from the special relationship to a good-friend

relationship was reasonably appropriate. Any episode

limited to a special relationship would not be relevant to

this study. (3) Semantically unclear episodes were

deleted.

The same type of survey was done with five Chinese

students. All the episodes they provided were found










28

redundant or similar in essence to those recalled by

American students, so it was decided that a suitable range

within the collection of episodes had been achieved.

A questionnaire based on forty-five episodes was

constructed. There were five episodes representing each of

the nine categories. Twenty-five American students and

twenty-five Chinese students not participating in the

episode-recalling session were asked to respond to the

forty-five episodes in terms of the dependent measurement

indicated before. Immigrants from other cultures and black

people were excluded from the American group, since they

were regarded as non-representative of the mainstream

culture in the United States. Among the twenty-five

American subjects, eleven of them were male, and fourteen

were female. Their mean age was 22.04, while the standard

deviation in age was 2.51. Within the Chinese group, there

were sixteen males and nine females, with the mean age of

26.6 and the standard deviation in age of 2.94. Most

American subjects were higher-level undergraduate students,

while most Chinese respondents were first-year graduate

students. There was no indication of significant

differences in academic maturity between these two groups

of students. In addition, students in each of the two

groups represented a wide variety of academic fields. They

reflected the same characteristics as those involved in the

major study.











Since all the subjects in both groups were proficient

in reading English, only one version of the questionnaire

was prepared, and that was in English. All the Chinese

subjects were allowed to consult a dictionary in case of

need, but were instructed not to discuss their responses

with each other.

Discontented responses investigated in this study

were, according to the definition used in this study, the

verbal and/or nonverbal expressions associated with

discontent. That is, the responses to hypothetical

situations indicated in this questionnaire were external

responses, rather than internal feelings. This key concept

was defined clearly to all the fifty subjects before they

answered the questionnaire, in order to avoid possible

confusion between "internal feelings" and "external

responses."

One episode was chosen from the five representing each

of the nine categories. A total of nine episodes were

chosen corresponding to the nine categories. The criteria

for choosing each of these nine episodes are as follows:

(1) In both groups, the episode having the highest mean

score among the five representing the same category was

chosen. (2) If the highest-scored episodes were different

in two groups, the one having the highest mean score in one

group and second highest in another was chosen. (3) If the

above procedure was not workable, then a check was made to









30

ascertain if the second-highest episode in one group was

also the second-highest one, or the third highest one, in

another group; a selection was made accordingly. (4) This

procedure was continued until an episode for each category

claiming similar responses from both groups was identified.

The SAS GLM-REPEATED procedure (SAS User's Guide:

Statistics, 1985, pp. 433-506) was used to analyze the

between-subjects and within-subjects effects. The two

groups were treated as a between-subjects variable. The

relationships of communication interactants and

determinants of discontent were two within-subjects

variables. The nine chosen episodes were the nine

combinations of the two within-subjects variables, each of

which had three levels. For example, a response score for

the episode representing the combination of "intimate +

injury" indicated a score for both the first level of the

relationship variable and the first level of the

determinant variable.

Results

Among the nine episodes chosen based on the above-

mentioned criteria, two of them registered the highest

score in both groups, four registered the highest in one

group and the second highest in the other group, two

registered the second highest in both groups, and the

remaining one has the highest in one group and the third

highest in the other group (see Table 3).












Table 3

The Mean Scores of the Episodes Selected
for Use in the Major Study (Mean 1)
and the Highest Mean Score (Mean 2)
within Each of the Nine Categories in
the Preliminary Study


Episodes Americans Chinese

mean 1 mean 2 mean 1 mean 2


Intimate + Injury 4.40 4.40 2.72 2.76

Intimate + Disappointment 4.68 4.68 2.64 2.84

Intimate + Disagreement 3.16 3.80 2.60 2.76

Acquaintance + Injury 4.48 4.68 2.72 3.24

Acquaintance + Disappoint- 4.04 4.04 2.68 2.72
ment

Acquaintance + Disagree- 3.28 3.88 2.60 2.68
ment

Stranger + Injury 4.16 4.16 2.96 2.96

Stranger + Disappoint- 4.44 4.72 3.16 3.16
ment

Stranger + Disagreement 2.68 2.68 2.56 2.56



By adopting the SAS GLM-REPEATED procedure to analyze

the nine chosen items, significant interactions were

detected between the groups and the relationships of

communication interactants, between the groups and the

determinants of discontent, and between the relationships

of communication interactants and the determinants of

discontent (see Table 4).












Table 4

The Effects of Cultural Groups, Relationships of


Communication Interactants, and Determinants of


Discontent in the Preliminary Study


Source


df SS


F value


Groups


Subjects within Groups



Relationships

Groups x Relationships

Relationships x Subjects
within Groups


1 158.4200


48 143.0222


53.17


2 0.3378 0.29

2 6.2400 5.45

96 54.9778


Determinants 2 60.4044 20.87 .0001*

Group x Determinants 2 30.2533 10.45 .0001*

Determinants x Subjects 96 138.8978
within Groups



Relationships x 4 7.6622 2.50 .0437
Determinants

Group x Relationships x 4 2.5867 0.85 .4980
Determinants

Relationships x 192 146.8622
Determinants x
Subjects within Groups


Indicates significant effect


.0001*


.7453

.0057"










33

The mean scores for the American group were

consistently higher than those for the Chinese group.

Further comparisons of means by adopting Scheffe type

confidence intervals indicated that for all but the three

"disagreement" categories, the means in the American group

were significantly higher than those in the Chinese group.

Separate repeated-measures analyses detected a

significant interaction between the relationships of

communication interactants and the determinants of

discontent in the American group (p = .0204), but not in

the Chinese group (p = .6036).

Within the American group, the mean scores were higher

for "disappointment" than for "injury" or "disagreement,"

when the relationships of communication interactants were

"intimate" or "stranger" ( For "intimate," mean scores for

"injury," "disappointment," and "disagreement" were 4.4,

4.68, and 3.16, respectively; for "stranger," mean scores

were 4.16, 4.44, and 2.68, respectively.). For

"Acquaintance," the mean score of "injury" (4.48) was

higher than that of "disappointment" (4.04) or

"disagreement" (3.28). And in general, the mean scores for

three "disagreement" categories were lower than any of the

six "injury" and "disappointment" categories (see Figure

1).

Since no significant interaction was detected between

the relationships of communication interactants and the















Disappointment

Injury



Disagreement


Intimate


Acquaintance


Stranger


Figure 1


The Interaction between the


Relationships of


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5


Communication Interactants and the Determinants
of Discontent for the American Group in the
Preliminary Study


I ~I











determinants of discontent, no similar figure will be

presented for the Chinese group. The repeated-measures

analysis for the Chinese group disclosed that significant

effects were found in neither the three relationships (p =

.1232) nor the three determinants (p = .5365).

Implications for the Major Study

Since the nine episodes were selected from a wide

collection of real life situations and each of them

represented a "strong" point within one of the nine

categories for both Americans and Chinese, they were likely

to induce reliable comparisons between the two cultural

groups in terms of the nine different situations. However,

due to the attempt to make the whole questionnaire shorter

and less problematic for the subjects, most of the forty-

five episodes included in the preliminary study were

composed by using brief statements. These brief statements

might not be descriptive and precise enough to elicit the

subjects' genuine responses. In the major study, each of

the nine selected episodes was refined to provide more

concrete and realistic contexts. Only nine episodes were

responded to by each subject in the major study, so the

more elaborated episodes were not likely to tire any

subject. Furthermore, because the language barrier could

have influenced the way Chinese subjects responded to the

different situations to some extent (though nothing in this

way was indicated in the study), it was advisable to











prepare a Chinese version of the questionnaire for Chinese

subjects in the major study.

The preliminary results from the comparison between

the two cultural groups, in terms of the relationships of

communication interactants and the determinants of

discontent, were encouraging as a further study had been

planned. The suggestion that the two cultural groups were

consistently, and in most situations significantly,

different in terms of their discontented responses

supported the existing "high-context culture versus low-

context culture" and "American culture versus Chinese

culture" theories. That is, Americans tended to use more

explicit and direct codes in communication, while the

reverse was true for Chinese. In addition, a contrasting

pattern within the nine categories was detected between the

two cultural groups. There was a significant interaction

between the relationships of communication interactants and

the determinants of discontent for the American group. For

the Chinese subjects, though no interaction was found, they

did not respond to any situation strongly. It appears that

they responded to all the different situations with the

same type of response, a response concealing discontent.

Whether the use of more elaborated episodes, and the

use of the Chinese version of the questionnaire for Chinese

subjects, would alter the results, became a legitimate

question raised after the preliminary study. In addition,










37

whether placing the episodes in intercultural contexts

would change the degree of discontented responses for both

cultural groups, and whether a different cultural

environment would influence the way they responded, were

also questions to be answered in the major study.



Validity and Reliability Considerations

The cultural memberships, cultural environments, and

relationships of communication interactants were explicitly

stated in each of the nine episodes selected in the

preliminary study, so the validity of the episodes in terms

of these three variables was obvious. The determinants of

discontent were, however, only implied in each episode. In

order to assess if each of the episodes was a believable

representation of the determinant of discontent assigned, a

validity evaluation was arranged. At first, the nine

episodes selected from the preliminary study were refined

to indicate a more detailed context. Four professors and a

doctoral candidate (a Japanese) in the Department of

Psychology, University of Florida, were then asked to

classify each of these elaborated episodes in one of the

three categories corresponding to the three determinants of

discontent. The definitions for the three categories were

provided in advance. The preassigned category for each

episode was not known to any of the evaluators. The

results showed that all of the five persons agreed with the











preassigned categories for four episodes. Another four

episodes were categorized by four out of the five

evaluators as preassigned. The remaining episode was

identified to be the preassigned category by three

evaluators. In other words, for eight episodes, the

identification of the determinants as provided by at least

four validity evaluators correlated with the determinants

projected in advance. This suggested that these eight

episodes had a strong indication of validity. Further

investigation was done for the one episode which had only

three interpretations correlating with the prior

classification. This episode involved a situation in which

the repair job for a motorcycle was underestimated, so the

respondent, when calling for the motorcycle, had to pay

much more money than originally estimated. Among the five

validity evaluators, three categorized it as a

"disappointment," and two identified it as an "injury."

This episode was regarded as a "disappointment" case in the

preliminary study, based on the rationale that what might

dissatisfy the respondent is mainly the unexpected raise of

the repair bill. That two evaluators interpreted it as an

"injury" indicates that the degree of offensiveness implied

in the episode might be a little too "strong" for a

"disappointment" classification. After consulting with one

of these evaluators, this episode was modified in order to

eliminate some elements which might make it more or less









39

like an "injury." The modification included the reduction

of the amount of increased charges in the final repair bill

and the offering of more justification for the increase on

the final repair bill by the motorcycle shop manager.

In general, a high degree of validity regarding the

determinant of discontent pertaining to each of the nine

episodes used in the major study seemed to have been

confirmed by this process of validity evaluation.

Test-retest and alternate form are the two major

methods for reliability check (Selltiz et al., 1967,

pp.167-182; Tucker et al., 1981, pp. 167-169). Since the

instrument used in this research has never been adopted in

previous studies, and the survey was not repeated within

the major study, a test-retest reliability check does not

apply to this investigation. As discussed earlier, the

nine episodes were developed from the preliminary study.

These episodes constituted the only form used in this

study; an alternate form reliability check does not apply

to this study either. A reliability check, however, can be

done in future studies, if a test-retest method will be

used, or if a similar instrument based on the same

rationale utilized for this study will be developed.



Instrument

The testing instrument used in the major study was a

questionnaire consisting of the following sections: (1)









40

instruction, (2) personal data, (3) the nine episodes based

on the preliminary survey, and (4) the scheme by which

subjects could respond. The questions pertaining to

personal data focused on age, gender, native culture, and

educational level. Information regarding each subject's

age, gender, native culture, and educational level made it

possible to retrieve his/her personal data easily in any

possible effort involving subsequent analysis. This

information also provided an opportunity to check the

background information for each subject. To avoid possible

gender bias, the episodes used to elicit subjects'

discontented responses were gender-free, i.e., only the

family name of the offender in each episode was used. For

the subjects answering the questionnaire in a cultural

environment other than their native culture, the length of

their stay in this culture was ascertained to establish

their degree of time proximity to their native culture.

A table of random numbers was used to decide the

sequence of the nine episodes as presented in the

questionnaire. This method resulted in the following

sequence of relationship/determinant combinations:

1. stranger + disagreement

2. acquaintance + injury

3. intimate + disagreement

4. acquaintance + disappointment

5. stranger + disappointment












6. intimate + disappointment

7. intimate + injury

8. acquaintance + disagreement

9. stranger + injury

Four different versions of the questionnaire were

constructed: Chinese version with Americans as offenders,

Chinese version with Chinese as offenders, English version

with Americans as offenders, and English version with

Chinese as offenders. The "offender" was the person who

caused discontent for the subject. The cultural membership

of the offender could be identified by the subject through

the statement such as "My good Chinese friend, Chen, .

included in each episode. The four versions of

questionnaire forms were utilized for the different groups

as follows:

1. Chinese version with Americans as offenders--for "C-A in

ROC" and "C-A in USA" groups,

2. Chinese version with Chinese as offenders--for "C-C in

ROC" group,

3. English version with Americans as offenders--for "A-A in

USA" group, and

4. English version with Chinese as offenders--for "A-C in

USA" and "A-C in ROC" groups.

The English version was composed first. The Chinese

version translated from the English version was back-

translated by a qualified bilingual blind to the study.










42

The back-translated English version was then compared with

the original English version for detecting inconsistency

between the Chinese version and the English version;

warranted revisions in wording were then made.



Episodes Used in Major Study

The episodes used in the major study were the nine

selected in the preliminary study, with elaborations and

modifications based on the validity considerations. Each

episode represented one of the nine possible combinations

of the relationship variable and the determinant variable.

The English version with Chinese as offenders was as

follows:

1. Intimate + Injury

My good Chinese friend, Chen, falsely accused me of
damaging a video cassette recorder (VCR): "Why did you
damage my VCR?" I told Chen that I didn't touch it at
all. Chen said, "I am sure you did. Nobody else
would use my VCR."

2. Intimate + Disappointment

After my car broke down on the highway, one of my good
Chinese friends, Wang, told me on the phone, "I'll
come to pick you up very soon." Yet Wang has been
three hours late.

3. Intimate + Disagreement

I had a conversation with one of my good Chinese
friends, Liu, about the proposed plan to purchase a
new piece of equipment for our tennis club. I
expressed my concern to Liu, "Everyone in our club
should be involved in the process of decision making
regarding this purchase." "I don't think so," Liu
claimed. "It would be time-wasting if everyone is
involved. The decision for this purchase should come
from the top."











4. Acquaintance + Injury

The owner of the store at which I have been working
for just a few days, Tai, is a Chinese. We treat each
other as acquaintances without any close contact.
Today, Tai told me, "I am sorry you can't continue
your job at this store beginning tomorrow. You were
the last person leaving here yesterday, but you didn't
lock the back door as you should." I am "fired," but
I was not informed in advanced that the last person
leaving the store should lock the back door.

5. Acquaintance + Disappointment

One of my Chinese acquaintances, Chang, borrowed a
book from me the day before yesterday. I told Chang
that I wanted to use it today, in conjunction with one
of my classes. Chang promised to return the book to
me by this morning, but didn't. When we met on the
street, Chang said, "Since I haven't finished reading
your book, I'll return the book to you tomorrow. I
hope you don't have to use it today."

6. Acquaintance + Disagreement

Though I always respect and accept my father's advice,
Wu, my Chinese acquaintance, said to me, "I don't
think you should allow your father to be so
domineering as you make plans to pursue your
professional career."

7. Stranger + Injury

While my friends and I were playing a baseball game, a
stranger, who is obviously a Chinese, kept shouting at
me: "You don't have to play the game if you can't!
You need to practice more!"

8. Stranger + Disappointment

The Chinese manager of a motorcycle shop (a stranger
to me) estimated that the repair job for my motorcycle
would cost thirty dollars. Since my budget allowed
for this estimate, I left my motorcycle there for
repairs. When I picked up my motorcycle, the manager
said, "Sir, I am sorry, the problem with your
motorcycle was much more serious than what I told you
this morning. A lot of work has been done. So the
total charge is forty, instead of thirty, dollars."











9. Stranger + Disagreement

A stranger, who is a Chinese, while talking with me at
a bus station, emphasized, "Canada is an
underdeveloped country in many respects. Nobody there
lives a happy life." What this person says is
certainly not true.

To illustrate how the other English version of the

questionnaire (with Americans as offenders) was constructed

based on the same prototypes, one episode representing the

combination of "acquaintance + disappointment" is presented

below:

One of my acquaintances, Jones, borrowed a book from
me the day before yesterday. I told Jones that I
wanted to use it today, in conjunction with one of my
classes. Jones promised to return the book to me by
this morning, but didn't. When we met on the street,
Jones said, "Since I haven't finished reading your
book, I'll return the book to you tomorrow. I hope
you don't have to use it today."



Subjects

The subjects used in the major study consisted of

college students in both Gainesville, Florida, and Taipei,

Taiwan. In Florida, twenty-five American students at the

University of Florida, Gainesville, were used for the "A-A

in USA" group, the same number of American students at the

same University were used for the "A-C in USA" group, and

the same number of Chinese students from Taiwan who were

studying at the same University formed the "C-A in USA"

group. In Taiwan, twenty-five Chinese students studying at

Fujen University, Taipei, were investigated as the "C-C in

ROC" group, the same number of Chinese students at the same









45

University were used for the "C-A in ROC" group, and the

same number of American students from both Fujen University

and National Taiwan Normal University, Taipei, constituted

the "A-C in ROC" group.

The use of college students of both cultures for

subjects had the following advantages: (1) To use the same

"type" of people in both cultures may tend to neutralize

those confounding factors such as age, working experiences

and educational level, which would make the explanation of

results more difficult. (2) Students were the only age

group that could be easily found in one setting. (3)

College students had no problem understanding and

responding to the nine episodes written in their native

languages.

Students studying English at Fujen University were

used as subjects in the "C-A in ROC" group, whereas the "A-

C in USA" group was comprised of students who were studying

Chinese or had completed Chinese courses at the University

of Florida. This choice not only made the questionnaire

involving intercultural encounters between Americans and

Chinese relevant to subjects (since they must have had

experiences communicating with people from the culture to

which the offenders in the episodes belong), but also made

the above two groups, with a similar degree of association

with the culture of the offenders, comparable. As

indicated before, black people in the United States tend to











develop a sub-culture different from the predominant sub-

culture to some extent, so they were not included in any of

the American samples. Immigrants from other cultures were

also excluded from any of the American samples, since they

were regarded as significantly non-representative of the

American culture. No subject in the "A-C in ROC" or "C-A

in USA" group had been living in the country where the

questionnaire was answered for over five years.

Each of the "A-A in USA," "C-C in ROC," and "C-A in

ROC" groups was tested in a classroom at the university

where the subjects were studying. In both the "A-C in USA"

and "A-C in ROC" groups, almost 75% of the subjects

answered the questionnaire in their classes, while the

remaining students answered individually. Subjects in the

"C-A in USA" group gathered as small groups of three to six

people to respond to the questionnaire at different times.

To avoid introducing unwanted factors to the investigation

due to instructional discrepancies, the message each

subject orally received before answering the questionnaire,

no matter what kind of environment in which the

questionnaire was answered, was: "Please read the

instruction on the first page of this questionnaire

carefully before you answer the nine questions." All the

subjects involved in the study answered the questionnaire

in writing. No subject in any group needed further

clarification in the process of answering questions.










47

Although gender was not treated as an independent

variable, an approximately equal number of male and female

students was used in each group to avoid any possible

biased conclusion based on a mono-gender sample. Age was

not an independent variable either, yet the use of college

students as subjects, as indicated before, made different

groups comparable. Within the "A-A in USA" group, there

were twelve males and thirteen females, with the mean age

of 21.32 and the standard deviation in age of 1.41. Within

the "A-C in USA" group, twelve were male, and thirteen were

female. Their mean age of this group was 22.44, while the

standard deviation in age was 3.58. In the "A-C in ROC"

group, fourteen males and eleven females were included,

with the mean age of 25.16 and the standard deviation in

age of 3.61. The "C-C in ROC" group was comprised of

twelve males and thirteen females. The mean age was 23.84,

and the standard deviation in age was 2.71. In the "C-A in

ROC" group, twelve males and thirteen females, with the

mean age of 24.76 and the standard deviation in age of

2.35, were used as subjects. And the mean age of the

thirteen males and the twelve females in the "C-A in USA"

group was 26.28, while the standard deviation in age was

2.97.









48

The Assistant in Taiwan

The person who collected data in Taiwan was a college

graduate with bilingual ability in Chinese and English. As

indicated by his current status as the sales manager in a

large publishing company, he had little problem in

complying with the specified operational procedure. He was

provided the appropriate questionnaire forms, which were

identical with those used in Florida, as well as directions

for the procedure. He was also requested to remind each

subject of carefully reading the instruction presented at

the top of the questionnaire (see Appendix C) before the

questionnaire was answered.

Before the process was initiated, the author conferred

orally with the assistant in Taiwan via telephone to assure

the procedure was understood. After the questionnaire had

been administered to the participating subjects, the

assistant was asked to record in writing all the responses

collected and to keep them securely in Taiwan. This

recording, prior to the mailing of the completed

questionnaires, preserved the data from possible loss due

to mailing irregularities (though the data were safely

received). No unusual phenomena were observed as the

subjects responded to the questionnaire, according to the

assistant's report when he was consulted via telephone

after the procedure was completed.











Analysis of Results

The statistical method, repeated-measures analysis,

was adopted to analyze the results (Pedhazur, 1982, pp.

551-573). More specifically, a factorial model with one

between-subjects variable (the different groups to be

analyzed) and two within-subjects variables (the

relationships of communication interactants and the

determinants of discontent) was fit using the "REPEATED"

option of the general linear models procedure (PROC GLM) of

the Statistical Analysis System (SAS) computer package (SAS

User's Guide: Statistics, 1985, pp. 433-506). The level of

significance for testing any hypothesis in this study was

pre-set at .05.

The Scheffe multiple comparison procedure was used to

compare means. By adopting this procedure for constructing

95% confidence intervals for a large number of pairwise

comparisons of means, "the probability that at least one of

the intervals is in error is controlled at the level of

.05" (Agresti & Finlay, 1986, p. 408). For many other

procedures, such as t-intervals, the error rate is .05 for

each comparison, not for the entire set of pairwise

comparisons (Agresti & Finlay, 1986, p. 410). In Scheffe's

procedure, a 95% confidence interval was given by

(Yi Yj) (MSe)1/2 x [(i/ni + 1/nj) x (# of

Groups 1) x (# of Relationships 1) x (# of

Determinants 1) x (F.05)]1/2









50


where "Yi" and "Yj" denote the two mean scores to be

compared. The error mean square, "MSe," is an estimate of

the variance associated with the observations. The "ni"

and "nj" represent the sample sizes of the two groups,

whereas "F.05" indicates the value from the F distribution

at the level of .05 with corresponding degrees of freedom.
















CHAPTER III
RESULTS



Results of Testing Hypotheses

The hypotheses were tested by applying the general

linear models procedure (PROC GLM) of the Statistical

Analysis System (SAS) (SAS User's Guide: Statistics, 1985,

pp. 433-506). The level of significance for testing

hypotheses was pre-set at .05.

Hypothesis One compared the discontented responses

between the "Americans responding to Americans in the

United States of America" ("A-A in USA") group and the

"Chinese responding to Chinese in the Republic of China"

("C-C in ROC") group:

H1: When the two variables, relationships of

communication interactants and determinants of

discontent, are controlled, United States of

America (U.S.A.) subjects in the U.S.A. respond

to U.S.A. communication interactants more

strongly than Republic of China (R.O.C.) subjects

in the R.O.C. respond to R.O.C. communication

interactants.









52

From the SAS GLM procedure, a significant three-way

interaction among the groups, the relationships and the

determinants was detected (see Table 5).

Since no conclusion can be drawn from the repeated-

measures analysis due to interaction, the mean scores for

each of the nine categories, reflecting combinations of the

relationships and determinants, between the "A-A in USA"

and "C-C in ROC" groups were compared.

The mean scores for the "A-A in USA" group are higher

than the "C-C in ROC" group in all the nine categories

except "stranger + injury" (see p. 56, Table 7, & p. 60,

Table 9). In order to test significance of these

differences, the Scheff' multiple comparison procedure was

adopted to compare mean scores between the two groups for

each of the nine categories. The results indicated that

zero is included in all nine of the 95% confidence

intervals, i.e., none of the nine pairs can be regarded as

significantly different from each other. Therefore,

Hypothesis One was not evidenced in this analysis.

The discontented responses of the three American

groups were compared in Hypothesis Two. The differences

among the three groups were in the cultural membership of

the offenders and the cultural environment in which their

responses were elicited:











Table 5

The Effects of Groups, Relationships of
Communication Interactants, and Determinants
of Discontent in "A-A in USA" and "C-C in
ROC" Groups


Source df SS F value p > F



Groups 1 21.7800 12.61 0.0009*

Subjects within Groups 48 82.9333



Relationships 2 0.7511 0.47 .6239

Groups x Relationships 2 3.6400 2.30 .1060

Relationships x Subjects 96 76.0533
within Groups



Determinants 2 66.2044 34.04 .0000*

Groups x Determinants 2 4.2133 2.17 .1202

Determinants x Subjects 96 93.3600
within Groups



Relationships x 4 22.1956 6.12 .0001
Determinants

Group x Relationships x 4 9.3067 2.57 .0395
Determinants

Relationships x 192 174.0533
Determinants x
Subjects within Groups


Indicates significant effect









54

H2: When the two variables, relationships of

communication interactants and determinants of

discontent, are controlled, responses in each of

the following three situations will reflect the

same degree of discontent:

1. U.S.A. subjects responding to U.S.A.

communication interactants in the U.S.A. ("A-A in

USA")

2. U.S.A. subjects responding to R.O.C.

communication interactants in the U.S.A. ("A-C in

USA")

3. U.S.A. subjects responding to R.O.C.

communication interactants in the R.O.C. ("A-C in

ROC" )

The results from the SAS GLM procedure disclosed that

interactions between the determinants and the groups,

between the relationships and the determinants, and among

the relationships, the determinants and the groups, were

all significant (see Table 6).

Due to significance of the interactions, the mean

scores among the three American groups, in terms of each of

the nine categories, were compared by adopting the Scheffe

type confidence intervals. The mean scores for the three

groups are presented in Table 7.

From the Scheffe multiple comparisons, no significant

difference was detected between any two of the three











Table 6

The Effects of Groups, Relationships of
Communication Interactants, and Determinants


of Discontent in the Three American Groups


Source df SS F value p > F



Groups 2 9.9763 1.70 .1904

Subjects within Groups 72 211.5733



Relationships 2 1.4252 0.69 .5035

Groups x Relationships 4 3.4992 0.85 .4981

Relationships x Subjects 144 148.8533
within Groups



Determinants 2 87.7541 40.60 .0001*

Groups x Determinants 4 20.3970 4.72 .0013*

Determinants x Subjects 144 155.6267
within Groups



Relationships x 4 37.5348 10.82 .0001*
Determinants

Group x Relationships x 8 14.1541 2.04 .0419*
Determinants

Relationships x 288 249.8667
Determinants x
Subjects within Groups


indicates significant effect











Table 7

Mean Scores in the Three American Groups
(The mean scores of these three groups are also
presented in Figures 2, 3, and 4, on pages 65, 66,
and 67.)


"A-A in USA" Group


Injury Disappointment Disagreement


Intimate

Acquaintance

Stranger


4.44

4.04

3.48


3.80

4.56

4.04


3.28

3.28

3.60


"A-C in USA" Group


Injury Disappointment Disagreement


Intimate

Acquaintance


4.44

3.88


Stranger 4.12


3.52

4.12

3.68


3.16

2.40

2.92











Table 7--continued



"A-C in ROC" Group


Injury Disappointment Disagreement


Intimate 4.12 3.52 3.24

Acquaintance 4.16 3.60 2.84

Stranger 3.48 3.40 3.80



American groups for any of the nine items representing

various combinations of the three relationships and the

three determinants. Since the result does not contradict

Hypothesis Two, there is no reason to believe that

Hypothesis Two is not true. In other words, Hypothesis Two

is strongly suggested in this analysis.

Hypothesis Three investigated the possible differences

among the three Chinese groups due to the differences in

the cultural memberships of the offenders and in the

cultural environments of the episodes:

H3: When the two variables, relationships of

communication interactants and determinants of

discontent, are controlled, responses in

situation 1, described below, will be stronger

than situation 2, and responses in situation 2

will be stronger than in situation 3:









58

1. R.O.C. subjects responding to R.O.C.

communication interactants in the R.O.C. ("C-C in

ROC")

2. R.O.C. subjects responding to U.S.A.

communication interactants in the R.O.C. ("C-A in

ROC")

3. R.O.C. subjects responding to U.S.A.

communication interactants in the U.S.A. ("C-A in

USA")

The SAS GLM procedure detected significant

interactions between the determinants and the groups, as

well as between the relationships and determinants (see

Table 8).

Further analysis of the mean scores in the three

Chinese groups was required because of the existence of

significant interactions. The means scores for these

groups are presented in Table 9.

The Scheffe confidence intervals indicated no

significant difference between any two of the three Chinese

groups for any of the nine categories reflecting

combinations of the two within-subjects variables, since

zero can not be excluded from any of these intervals.

Hypothesis Three is not supported.

Hypotheses Four and Five concerned the differences in

the responses to each level of the two within-subjects

variables, relationships of communication interactants and











Table 8

The Effects of Groups, Relationships of
Communication Interactants, and Determinants


of Discontent in the Three Chinese Groups


Source df SS F value p > F



Groups 2 1.4074 0.40 .6712

Subjects within Groups 72 126.3644



Relationships 2 3.6296 2.24 .1099

Groups x Relationships 4 3.6059 1.11 .3524

Relationships x Subjects 144 116.5422
within Groups



Determinants 2 189.2385 89.27 .0001*

Groups x Determinants 4 13.2504 3.13 .0168*

Determinants x Subjects 144 152.6222
within Groups



Relationships x 4 27.4015 7.28 .0001
Determinants

Group x Relationships x 8 9.7896 1.30 .2430
Determinants

Relationships x 288 271.0311
Determinants x
Subjects within Groups


indicates significant effect












Table 9

Mean Scores in the Three Chinese Groups
(The mean scores of these three groups are also
presented in Figures 5, 6, and 7, on pages 73, 74,
and 75.)


"C-C in ROC" Group


Injury Disappointment Disagreement


Intimate

Acquaintance

Stranger


3.92

3.40

4.12


3.40

4.00

3.44


2.56

2.72

3.00


"C-A in ROC" Group


Injury Disappointment Disagreement


Intimate

Acquaintance


4.00

3.72


3.72

4.08


2.88

2.64

2.88


3.92 3.72


Stranger











Table 9--continued



"C-A in USA" Group


Injury Disappointment Disagreement


Intimate 4.52 2.76 2.40

Acquaintance 4.04 3.96 2.60

Stranger 4.16 3.68 2.84



determinants of discontent, in both American groups and

Chinese groups:

H4: When determinants of discontent are controlled,

U.S.A. subjects will respond to communication

interactants of different relationships with the

same degree of discontent, but R.O.C. subjects

will respond to "intimates" more strongly than

"acquaintances" or "strangers."

H5: When cultural memberships of U.S.A. and R.O.C.

subjects and relationships of communication

interactants are controlled, the responses to

"injury" or "disappointment" will be stronger

than "disagreement."

The fifth hypothesis predicted that the responses to

"disagreement" would be weaker than to "injury" or

"disappointment" across all the six investigated groups, so









62

a repeated-measures analysis accommodating all six groups

was adopted. Since the results indicated that interactions

between the determinants and the six groups, between the

relationships and the determinants, and among the

relationships, the determinants and the six groups, had

significant influences on discontented responses, these

groups deserve to be discussed separately (see Table 10).

In comparing three American groups alone, a similar

pattern was detected; i.e., the effects of the

determinants, as well as interactions between the

determinants and the groups, between the relationships and

the determinants, and among the relationships, the

determinants and the groups, were significant (see p. 55,

Table 6).

Separate repeated-measures analysis for each of the

three American groups further indicated that the

interaction between the relationships and the determinants

was significant in all these groups (p = .0002 for the "A-A

in USA" group, p = .0141 for the "A-C in USA" group, and p

=.0002 for the "A-C in ROC" group).

In the "A-A in USA" group, the mean score for "injury"

was the highest among the three determinants when the

relationship was "intimate," was higher than "disagreement"

and lower than "disappointment" when the relationship was

"acquaintance," and was the lowest in the case of

"stranger." The mean score for "disappointment" was the











Table 10

The Effects of Relationships of Communication
Interactants and Determinants of Discontent
across All the Six Groups


Source df SS F value p > F



Relationships 2 0.2533 0.14 .8716

Groups x Relationships 10 11.9067 1.29 .2343

Relationships x Subjects 288 265.3956
within Groups



Determinants 2 267.3378 124.89 .0000

Groups x Determinants 10 43.3022 4.05 .0001*

Determinants x Subjects 288 308.2489
within Groups



Relationships x 4 54.6356 15.10 .0001*
Determinants

Group x Relationships x 20 34.2444 1.89 .0109*
Determinants

Relationships x 576 520.8978
Determinants x
Subjects within Groups


Indicates significant effect

highest among the three when the relationships were

"acquaintance" and "stranger," and was higher than

"disagreement" and lower than "injury" when the

relationship was "intimate." The mean score for

"disagreement" was the lowest in all relationships but









64

"stranger" (see Figure 2). The mean scores in the "A-C in

USA" group were similar to those in the "A-A in USA" group.

The only difference was that, in the "A-C in USA" group,

"stranger" was the lowest across all three relationships

(see Figure 3). Within the "A-C in ROC" group, the mean

scores for "injury" were the highest, and "disagreement"

were the lowest, when relationships were "intimate" and

"acquaintance." For "stranger," "disagreement" was the

highest, and "disappointment" was the lowest (see Figure

4).

Scheffe multiple comparisons were utilized for testing

significance of differences of means among the three

relationships in each of the three determinants. Only two

significant differences were detected in the twenty-seven

comparisons for the three American groups: the difference

between "intimate + injury" and "stranger + injury" in the

"A-A in USA" group (4.44 and 3.48), and the difference

between "acquaintance + disagreement" and "stranger +

disagreement" in the "A-C in ROC" group (2.84 and 3.80,

respectively).

In comparing the means between "injury" and

"disagreement," and between "disappointment" and

"disagreement," at each level of the relationships, in the

three American groups, zero was found to be included in ten

of a total of eighteen confidence intervals. The eight

pairs having significantly different mean scores are
















- Injury


Disappointment

Disagreement


Intimate


Acquaintance


Stranger


Figure 2


The Interaction between the


Relationships of


5.0 -


4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5


Communication Interactants and the Determinants
of Discontent for the "A-A in USA" Group


--
















- Injury



- Disappointment

- Disagreement


Intimate


Acquaintance


Stranger


Figure 3

The Interaction between the Relationships of
Communication Interactants and the Determinants


of Discontent for the "A-C in USA" Group


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5


1


















- Injury

- Disappointment
Disagreement


Intimate


Acquaintance


Stranger


Figure 4


The Interaction between the


Relationships of


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0


Communication Interactants and the Determinants
of Discontent for the "A-C in ROC" Group









68

"intimate + injury" versus "intimate + disagreement" in all

the three groups, "acquaintance + injury" versus

"acquaintance + disagreement" in the two "A-C" groups,

"acquaintance + disappointment" versus "acquaintance +

disagreement" in the two "in USA" groups, and "stranger +

injury" versus "stranger + disagreement" in the "A-C in

USA" group (see Table 11).

From the Scheffe multiple comparisons, significant

differences were detected for only two comparisons related

to the three relationships in the three American groups.

Results of most comparisons regarding relationships in the

American groups were not in contradiction to Hypothesis

Four. Hypothesis Four is thus suggested in the three

American groups, except for the two above-mentioned

instances. Hypothesis Five is, on the other hand, largely

unevidenced in the American groups.

Significant effects of the determinants of discontent

as well as interactions between the determinants and the

groups, and between the determinants and the relationships

were also detected in the three Chinese groups (see p. 59,

Table 8).

In separate repeated-measures analyses of the three

Chinese groups, interaction between the relationships and

the determinants was found significant in the "C-C in ROC"

group (p = .0185) and the "C-A in USA" group (p = .0002),

but not in the "C-A in ROC" group (p = .3791). The











Table 11

Scheffe Multiple Comparisons of the Means
of the Three Determinants of Discontent among
the Three American Groups



"A-A in USA" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. NSD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. NSD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. SD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. NSD
"Stranger + Disagreement"











Table 11--continued



"A-C in USA" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. SD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. SD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. SD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. NSD
"Stranger + Disagreement"











Table 11--continued



"A-C in ROC" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. SD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. NSD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. NSD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. NSD
"Stranger + Disagreement"


SD indicates "significantly different."
NSD indicates "not significantly different."

repeated measures for the "C-A in ROC" group also detected

significant effects for the determinants (p = .0001) but

not for the relationships (p = .9366).

In the "C-C in ROC" and "C-A in ROC" groups, the mean

scores were higher for "injury" than for "disappointment"

or "disagreement" when the relationships of communication

interactants were "intimate" or "stranger." When the

relationships was "acquaintance," the mean score for

"disappointment" was higher than that for "injury" or

"disagreement" (see Figures 5 and 6). Within the "C-A in

USA" group, the mean scores for "injury" was the highest,









72
and those for "disagreement" was the lowest, among the

three determinants in all the three different relationships

(see Figure 7).

The Scheffe multiple comparison procedure was adopted

to compare means of the three relationships at each of the

three levels of the other within-subjects variable,

determinants, for the three Chinese groups. Differences

were found significant in only two of the twenty-seven

possible comparisons. These two comparisons are "intimate

+ disappointment" versus "acquaintance + disappointment,"

and "intimate + disappointment" versus "stranger +

disappointment," in the "C-A in USA" group. Contrary to

the prediction of Hypothesis Four, the mean score of

"intimate + disappointment" in the "C-A in USA" group

(2.76) was found to be significantly lower than that of

"acquaintance + disappointment" or "stranger +

disappointment" in the same group (3.96 and 3.68,

respectively).

The Scheffe confidence intervals discovered twelve

significant differences between the determinants, "injury"

and "disagreement," and between "disappointment" and

"disagreement," in the three Chinese groups, with the three

relationships controlled. The six pairs which were not

significantly different from each other based on Scheffe's

method include "intimate + disappointment" versus "intimate

+ disagreement" in all the three groups, "stranger +


















- Injury

- Disappointment



- Disagreement


Intimate


Acquaintance


Stranger


Figure 5

The Interaction between the Relationships of
Communication Interactants and the Determinants
of Discontent for the "C-C in ROC" Group


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5


1

















Injury
Disappointment



Disagreement


Intimate Acquaintance Stranger


Figure 6


The Interaction between the


Relationships of


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0


Communication Interactants and the Determinants
of Discontent for the "C-A in ROC" Group
















-Injury


--/--



Disappointment
- Disagreement


Intimate


Acquaintance


Stranger


Figure 7

The Interaction between the Relationships of
Communication Interactants and the Determinants


of Discontent for the "C-A in USA" Group


5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5


i i I










76

disappointment" versus "stranger + disagreement" in the two

"in ROC" groups, and "acquaintance + injury" versus

"acquaintance + disagreement" in the "C-C in ROC" group

(see Table 12).

Based on Scheffe type confidence intervals, Hypothesis

Four is not supported in the Chinese groups. Among the

three determinants, mean scores for "stranger" were

significantly lower than other two determinants in twelve

of eighteen comparisons. Hypothesis Five is supported in

these comparisons only.



Summary

A summary of the findings based on statistical

analyses of the subjects' responses is presented in Table

13.

The differences in discontented responses between "A-

A in USA" and "C-C in ROC" groups were not found to be

statistically significant, when relationships of

communication interactants and determinants of discontent

were controlled. Hypothesis One has not been supported.

Discontented responses for the three American groups were,

as predicted in Hypothesis Two, not significantly different

from one another, when relationships and determinants were

controlled. The three Chinese groups were similarly not

significantly different from one another, contrary to the

prediction of Hypothesis Three. For each of the two











Table 12

Scheffe Multiple Comparisons of the Means
of the Three Determinants of Discontent among
the Three Chinese Groups



"C-C in ROC" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. NSD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. SD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. SD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. NSD
"Stranger + Disagreement"











Table 12--continued



"C-A in ROC" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. SD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. SD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. SD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. NSD
"Stranger + Disagreement"











Table 12--continued



"C-A in USA" Group Results



"Intimate + Injury" vs. SD
"Intimate + Disagreement"

"Acquaintance + Injury" vs. SD
"Acquaintance + Disagreement"

"Stranger + Injury" vs. SD
"Stranger + Disagreement"

"Intimate + Disappointment" vs. NSD
"Intimate + Disagreement"

"Acquaintance + Disappointment" vs. SD
"Acquaintance + Disagreement"

"Stranger + Disappointment" vs. SD
"Stranger + Disagreement"


SD indicates "significantly different."
NSD indicates "not significantly different."

cultural groups, the differences among the responses to

three relationships of communication interactants, with

determinants of discontent controlled, were found

significant in only two of twenty-seven instances. This

finding is quite concordant with what was predicted in

Hypothesis Four for Americans, but discordant with the

prediction for Chinese in the same Hypothesis. Hypothesis

Five predicted that "disagreement" will be responded to by

both Americans and Chinese less strongly than to "injury"

or "disappointment." It is supported in twelve of eighteen

comparisons in the Chinese groups, but only in eight of the











Table 13

Summary of Hypotheses Test Results


Hypotheses


Results


page 51)

page 54)

page 57)

page 61)


H5 (see page 61)


Non-supported

Supported

Non-supported

Largely supported for American groups,
and non-supported for Chinese groups

Supported in 2/3 of the comparisons for
Chinese groups,
and only in 4/9 of the comparisons for
American groups


same number of comparisons within the American groups.


(see

(see

(see

(see















CHAPTER IV
DISCUSSION



Conclusions

This study investigated the differences in

discontented responses between Americans and Chinese, with

regard to three relationships of communication interactants

and three determinants of discontent. The three

relationships of communication interactants were intimate,

acquaintance, and stranger, whereas the determinants of

discontent were injury, disappointment, and disagreement.

The hypotheses generated from current research suggested

that the discontented responses in the group, "Americans

responding to Americans in the United States of America"

("A-A in USA"), would be stronger than those in the group,

"Chinese responding to Chinese in the Republic of China on

Taiwan" ("C-C in ROC") (Hypothesis One). The discontented

responses in the three American groups, "A-A in USA," "A-C

in USA," and "A-C in ROC," would reflect the same degree of

discontent (Hypothesis Two). Yet different degrees of

discontent were likely to be discovered for the responses

in the three Chinese groups, "C-C in ROC," "C-A in ROC,"

and "C-A in USA" (Hypothesis Three). The hypotheses also

proposed that American discontented responses would not be

81









82

subject to relational variations, while the reverse would

be true for Chinese (Hypothesis Four). In addition, both

Americans and Chinese would respond to "disagreement" less

strongly than to "injury" or "disappointment" (Hypothesis

Five). On the basis of data collected both in Taiwan and

in the United States of America, these hypotheses were

tested by applying the statistical methods, repeated-

measures analysis and multiple comparisons of means. As

results of these statistical analyses, five conclusions

corresponding to the five hypotheses are presented below.

The first conclusion that can be drawn from this study

is that, in intracultural communication, Americans are not

significantly different from Chinese in expressing

discontented responses. Although overall mean degree of

strength of response tends to be higher for Americans than

for Chinese, the differences can not be regarded as

significant when relationships of communication

interactants and determinants of discontent are properly

controlled. This finding is not consonant with current

theories about American and Chinese cultures. There are

several possible explanations for this inconsistency.

First, it seems that most existing theories regarding

differences between high- and low-context cultures, and

between American and Chinese cultures, were constructed on

the basis of the communication of people in the two

cultures in non-discontented situations, such as initial









83

interactions and typical daily interactions. Non-

discontented interactions can be quite different from

discontented ones in the sense that people are less likely

to hide or control their emotional states in the latter,

whatever their cultural identification is. Researchers'

lack of attention to the two important variables,

relationships and determinants, could also prevent existing

theories from revealing more precise information about

communication behaviors in these two cultures. The other

possible reason is that cultural values in both American

and Chinese cultures have been more or less harmonized

because of the increased interactions and understanding

between the two cultures in recent years; a degree of

shared acculturation may have occurred. Finally, in terms

of statistical interpretation, Scheffe type confidence

intervals for comparing means, which were utilized in this

study, are more conservative than many others. That is,

Scheffe's method would reject null hypotheses less

frequently than many others. For example, as indicated by

Agresti and Finlay (1986), "The Scheffe intervals tend to

be wider than the corresponding intervals using the t

distribution" (p. 410). The moderate size of the sample in

each of the investigated groups (twenty-five subjects in

each group) also tended to make confidence intervals wider

than they could have been if a larger sample had been used.

A wider interval is more likely to include zero (the null









84

hypothesis will not be rejected if zero is included in the

interval) and less likely to detect differences. In a

strict sense, "Failure to reject the null hypothesis

implies only the difference between population means, if

any, is not large enough to be detected with the given

sample size" (SAS User's Guide: Statistics, 1985, p. 470).

In consideration of the questions raised previously

regarding the refinement of the questionnaire used in the

preliminary study, the following observations appear

warranted. In the preliminary study, significant

differences between the American group and the Chinese

group were detected for six of the nine situations

reflecting various combinations of the three relationships

and the three determinants. Yet differences were found

significant for none of the nine situations in the major

study. One apparently valid explanation for this

discrepancy is that different questionnaires were adopted.

The refined episodes used in the major study were very

likely to elicit more realistic responses from the

subjects, since the subjects would be better apprised of

the communication contexts by the episodes with additional

details. Furthermore, the use of two Chinese version

questionnaires in the major study appears to have largely

increased the degree of strength in discontented responses

for the three Chinese groups, especially when the

determinants were "injury" and "disappointment." The











finding is concordant with the discovery of effects of

language versions of a questionnaire by Yang and Bond

(1980). In an investigation concerning "ethnic

affirmation" by Chinese bilinguals in Hong Kong, the

authors found "a tendency for Chinese students to respond

in a more 'Chinese' way when the questionnaire was printed

in English than when it was printed in Chinese" (p. 419).

Discontented responses expressed by the Chinese subjects in

the preliminary study can be regarded, in one sense, as a

more accurate reflection of basic Chinese cultural values,

such as harmony, and implicitness of communication, than

the somewhat different responses expressed by the Chinese

subjects in the major study.

The second conclusion for this study is that

discontented responses by Americans are not likely to be

influenced by the culture represented by the "offender," or

by the cultural environment in which the communication

event occurs. This finding supports many current theories,

such as Hall's "low-context cultures" theory (1976) and

Hsu's argument that Americans are "individual-centered,"

rather than "situation-centered" (1953).

The third conclusion drawn from this investigation is

that Chinese people's discontented responses, similarly to

those of Americans, are not likely to reflect significant

variations due to different cultural memberships of

"offenders" or different cultural environments of









86

communication events. This finding is contrary to what has

been suggested by current theories. The discordance can be

attributed to several possible factors. First, Hsu's claim

that Chinese are "situation-centered" (1953), as well as

other theories concerning the values held by Chinese

people, may not have relevance to circumstances involving

discontented responses. It is also possible that many

Chinese have shifted more or less from being "situation-

centered" to being "individual-centered," as a result of

rapid westernization. Another possible explanation is that

the variations due to different cultural memberships of

"offenders" and different cultural environments are not

easily detected in self-report forms of investigation, such

as the questionnaires that were utilized in this

investigation.

The fourth conclusion drawn from this investigation is

that relationships of communication interactants do not

seem to generate any significant effects for either

Americans or Chinese, in terms of their discontented

responses. Nomura and Barnlund's finding (1983) that the

closeness of personal relationships is positively

correlated with the activeness of discontented responses in

Japan has not been reflected by Chinese subjects in this

study. For American subjects in this study, neither was an

association revealed between relationships and discontented

responses. This result is in accordance with Nomura and









87

Barnlund's conclusions. From Nomura and Barnlund's study,

as well as from this study, it seems that, while

discontented responses by both Americans and Chinese seem

to be free from influences by their relationships with

communication interactants, Japanese, in contrast to

Americans and Chinese, tend to be sensitive to their

relationships with others in communicating discontented

responses.

The fifth and last conclusion is concerned with the

role of the three discontent determinants in discontented

responses involved in intercultural interrelationships.

Significant differences were detected in eight of the nine

comparisons of means between "injury" and "disagreement"

responses for the Chinese groups investigated in this study

(see p. 77, Table 12). To Americans, the difference

between "injury" and "disagreement" responses was found

significant in six of nine comparisons (see p. 69, Table

11). Significant differences between responses indicating

"disappointment" and "disagreement" were found in four of

nine comparisons for the Chinese groups, but only in two of

nine similar comparisons for the American groups (see p.

69, Table 11, & p. 77, Table 12). In addition, the mean

response scores related to "disagreement" were consistently

the lowest among the three determinants for the Chinese but

not the American groups (see p. 56, Table 7, & p. 60, Table

9). From this overall picture, Chinese are more likely









88

than Americans to differentiate "disagreement" from

feelings of "injury" or "disappointment" by registering

"disagreement" responses less strongly than others, though

this variation between the two cultural groups is by no

means impressive. The relatively non-differentiating

attitude toward the three determinants of discontent among

the Americans may support other research findings to some

extent, though it is in conflict with Nomura and Barnlund's

study (1983). Hsu (1953) concluded that Americans are more

capable of attacking "wrongs" or fighting for "truth" in

society, than are Chinese; Americans' concern might be due

to many possible factors, such as, they are introduced to

the world of adults later in life, and, they are less bound

to the primary group and to traditional ways, than are

Chinese (pp. 70-83 & 355-380). Relatively stronger

"disagreement" responses by Americans can, in a sense, be

regarded as a characteristic of communication behavior

reflective of "truth-fighting," since advantages of a more

personal nature are usually not involved in these

situations.



Limitations of the Study

This study was designed to test five hypotheses on the

basis of derived, sufficient quantitative data obtained

from available subjects. The strength of the research,

however, does not preclude the existence of possible









89

inherent limitations. Three possible limitations of this

study are presented below.

Research Method

If any intercultural communication research endeavor

can incorporate both qualitative and quantitative

approaches, this study was designed to reflect the best

features of both, but is obviously more quantitative in

nature. Ting-Toomey (1984) indicated that the primary

distinction between these two approaches is "the

fundamental principles that govern and guide each research

paradigm" (p. 169). While the major task of quantitative

research is to project "the generalizability and

predictability of human behavior," qualitative research

intends to map "the interpretive and situational schemata

of social actors" (Ting-Toomey, 1984, p. 182). Ting-Toomey

(1984) indicated that the strength of qualitative studies

stems from the special attention paid to the "interpretive

principles" which "social actors used to cognitively 'make

sense' of the meanings that are attached to their symbolic

activities," and the "contextual principles" that

situationallyy govern and guide the interpretation of

discourse in a speech community" (pp. 170-171). These

"principles" are usually deemphasized in quantitative

research. Although efforts were made in this study to

bridge the gap between qualitative and quantitative

research methods, restraints in accomplishing this goal may









90
still be apparent. The inclusion of two within-subjects

variables, relationships of communication interactants and

determinants of discontent, was likely to have enhanced the

understanding of the interpretive and contextual principles

in each of the two cultures investigated. Yet the

complexity of these principles was not likely to be

completely revealed by the three levels of relationships

(intimate, acquaintance, and stranger) and the three levels

of determinants (injury, disappointment, and disagreement).

There seem to be many possible variations within each of

the nine categories representing different combinations of

the two within-subjects variables. For instance, a

situation featuring the factor of "acquaintance" can vary

from a respected professor to a disliked neighbor; a

subject's identification of an "acquaintance," therefore,

could very likely determine the intensity of discontented

responses. There can also be other conditions, such as the

respondent's emotional state, the special meaning of a

described occasion to the respondent, or the appearance of

a third person in the communication setting, any or all of

which could influence this person's discontented responses

to some extent. The controlled investigation featured in

any quantitative research usually necessitates slighting of

some "insider's meanings" and oversimplifying the process

of human interactions.










Data Collection

The advantages of a questionnaire for recording

responses in research have been recognized for years.

These advantages include ensured uniformity through

standardized wording, instruction and order of questions,

and low administrative costs (Selltiz et al., 1967, pp.

238-239). In addition, a questionnaire can create gender-

free situations by relying on last names only, as was done

in this study.

A major problem, however, could also arise from using

a questionnaires in research. What the subjects thought

they would do, in responding in writing to hypothetical

situations, might not characterize their behavior in actual

situations. The discrepancy between what they thought they

would do and what they would actually do does not represent

an uncommon occurrence in society. For example, very few

so-called "cowards" would think they were less brave than

others while they were merely responding to hypothetical

situations. What a questionnaire can elicit may usually be

more related to "what respondents think they would do" than

"what they would actually do." As one projected example of

difficulty, this discrepancy could cause problems in

testing potential differences in discontented responses

between the "C-C in ROC" and "C-A in ROC" groups, in view

of the variations in cultural memberships of "offenders"

(Hypothesis Three). Chinese subjects might also have had









92

difficulties in differentiating responses to situations

with different relationships of communication interactants

involved (Hypothesis Four).

Another possible problem is that the reality to which

this study refers might have been more or less clouded by

the questionnaires. In order to compare the strength of

discontented responses among different groups, only the

question pertaining to whether the subjects would express

discontent in reaction to the "offender" in each of the

nine episodes was asked. Yet, insight into the seriousness

of discontent perceived by the subject and into the

inherent nature of the response may not be gauged

accurately until this second, open-ended question is

investigated: "What type(s) of discontented response would

characterize subjects' thinking and behavior as they

responded to described situations?" That both American and

Chinese subjects "would show discontent" to a specific

situation does not necessarily mean that they would

manifest the same types of responses in registering their

discontent. Physical violence as a means of showing

discontent, for instance, would be sharply different from

showing discontent through facial expressions unaccompanied

by verbalization. A range of alternative communication

behaviors would be available to the discontented person as

the interaction proceeds.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ESK9W594Q_FC2H69 INGEST_TIME 2012-02-07T17:48:29Z PACKAGE AA00003792_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES