"Education, Faculty of"@en . "Educational and Counselling Psychology, and Special Education (ECPS), Department of"@en . "DSpace"@en . "UBCV"@en . "Lau, Siu-Ling Bonita"@en . "2010-09-11T16:58:53Z"@en . "1989"@en . "Master of Arts - MA"@en . "University of British Columbia"@en . "The present study is a cross cultural study comparing North Americans and Chinese on field dependence-independence, locus of control and uncertainty orientation. The relationships among these constructs and cultural based differences in those relationships were also explored. Chinese were speculated to be more field dependent, external locus of control and certainty-oriented than North Americans. This study was the first exploratory investigation of a suggested model, which provided a perspective of how field dependence-independence, locus of control, uncertainty orientation, moral judgement and learned helplessness interact. The study may also have some contribution to broadening our view on how Chinese express themselves on the three cognitive constructs mentioned above.\r\nA sample of 39 Hong Kong Chinese and 41 Canadian college students were selected according to their cultural backgrounds. The results partially supported cultural differences in locus of control and uncertainty orientation. A strong cultural difference was found on the two subscales of the locus of control instrument and the two component measures for uncertainty orientation. Cultural differences for field dependence-independence was not confirmed. Results concerning the interrelationships among these cognitive constructs suggested a moderate association between locus of control and uncertainty orientation for Canadian subjects, but not for Chinese subjects. No other significant correlations among these constructs was obtained in this study. Because of the methodological problems inherent in the instruments, it is difficult to interpret the obtained results unambiguously, especially for the Hong Kong group. Further empirical work should be done before any firm conclusion can be drawn form the current results. Some implications for future cross-cultural studies were also discussed."@en . "https://circle.library.ubc.ca/rest/handle/2429/28414?expand=metadata"@en . "CROSS CULTURAL DIFFERENCES IN LOCUS OF CONTROL, FIELD DEPENDENCE-INDEPENDENCE AND UNCERTAINTY ORIENTATION AMONG CANADIANS AND CHINESE by SIU-LING BONITA L A U A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Educational Psychology & Special Education We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA, 5 June 1989 \u00C2\u00AE Siu-Ling Bonita Lau, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT The present study is a cross cultural study comparing North Americans and Chinese on field dependence-independence, locus of control and uncertainty orientation. The relationships among these constructs and cultural based differences in those relationships were also explored. Chinese were speculated to be more field dependent, external locus of control and certainty-oriented than North Americans. This study was the first exploratory investigation of a suggested model, which provided a perspective of how field dependence-independence, locus of control, uncertainty orientation, moral judgement and learned helplessness interact. The study may also have some contribution to broadening our view on how Chinese express themselves on the three cognitive constructs mentioned above. A sample of 39 Hong Kong Chinese and 41 Canadian college students were selected according to their cultural backgrounds. The results partially supported cultural differences in locus of control and uncertainty orientation. A strong cultural difference was found on the two subscales of the locus of control instrument and the two component measures for uncertainty orientation. Cultural differences for field dependence-independence was not confirmed. Results concerning the interrelationships among these cognitive constructs suggested a moderate association between locus of control and uncertainty orientation for Canadian subjects, but not for Chinese subjects. No other significant correlations among these constructs was obtained in this study. Because of the methodological problems inherent in the instruments, it is difficult to interpret the obtained ii results unambiguously, especially for the Hong Kong group. Further empirical work should be done before any firm conclusion can be drawn form the current results. Some implications for future cross-cultural studies were also discussed. iii T A B L E OF CONTENTS ABSTRACT ii LIST OF TABLES vi LIST OF FIGURES vii ACKNOWLEDGEMENTS viii I. RESEARCH PROBLEM AND RELATED RESEARCH 1 A. A MODEL OF THE RELATIONSHIPS AMONG FIELD DEPENDENCE-INDEPENDENCE, LOCUS OF CONTROL, UNCERTAINTY ORIENTATION, MORAL JUDGEMENT AND LEARNED HELPLESSNESS 3 B. LITERATURE REVIEW ....... 5 1. Proposition l a - lc: Chinese are more field dependence, external locus of control & certainty-oriented than North Americans ; 5 a. Field dependence-independence and cross cultural studies comparing Chinese and North Americans 5 b. Locus of control and cross cultural studies comparing Chinese and North Americans 9 c. Uncertainty orientation and cross cultural studies comparing Chinese and North Americans 13 d. Chinese cultural characteristics 14 e. Empirical findings on Chinese cultural characteristics 20 f. Relationships among field dependence-independence, locus of control and uncertainty orientation 25 2. Proposition 2a - 2c: Field dependent, external controlled and certainty-oriented individuals are more conventional in moral judgement than field independent, internal controlled and uncertainty-oriented individuals 27 a. Conventional versus autonomous moral judgement 27 b. Relationships among field dependence-independence, locus of control and moral judgement 28 c. Relationship between uncertainty orientation and moral judgement 31 3. Proposition 3: Chinese have less learned helplessness experience than North Americans 31 4. Proposition 4. Individuals with conventional moral judgement are less susceptible to learned helplessness than those with autonomous moral judgement 33 C. CHAPTER SUMMARY 33 H. NATURE OF THE STUDY... 34 A. RESEARCH OBJECTIVES 34 B. RESEARCH QUESTIONS 36 C. CHAPTER SUMMARY 37 iv III. METHOD 38 A. SUBJECTS & SAMPLING PROCEDURES 38 B. INSTRUMENTS 39 1. Field Dependence-Independence Instrument 40 2. Locus of Control Instrument 41 3. Assessment of Uncertainty Orientation 43 4. Demographic Questionnaire 45 C. PROCEDURE 46 D. CHAPTER SUMMARY 47 IV. RESULTS 49 A. DEMOGRAPHIC CHARACTERISTICS OF SUBJECTS 49 1. Cultural Background 49 2. Gender, Age and Educational Background 51 3. Socioeconomic Status 52 B. MAIN ANALYSES ....... 55 1. Confounding Effects 55 2. Cultural Differences in Field Dependence-Independence, locus of control and uncertainty orientation 57 3. Correlations among Field Dependence-Independence, locus of control and uncertainty orientation 59 C. . ADDITIONAL ANALYSES 64 D. CHAPTER SUMMARY 67 V. DISCUSSION 69 A. CULTURAL DIFFERENCES IN FIELD DEPENDENCE-INDEPENDENCE. LOCUS OF CONTROL & UNCERTAINTY ORIENTATION 69 B. RELATIONSHIPS AMONG FIELD DEPENDENCE-INDEPENDENCE, LOCUS OF CONTROL & UNCERTAINTY ORIENTATION 73 C. IMPLICATIONS FOR CROSS-CULTURAL RESEARCH STUDIES 74 D. CONCLUSION 76 REFERENCES 77 APPENDIX A. CONSENT FORM 85 APPENDIX B. ENGLISH VERSION OF QUESTIONNAIRES 90 APPENDIX C. CHINESE VERSION OF QUESTIONNAIRES 115 APPENDIX D. DEBRIEFING MESSAGE 136 APPENDIX E. THE SOCIOECONOMIC INDEX OF OCCUPATIONS 140 v LIST OF TABLES Table 1. Cultural Background of the Subjects 50 Table 2. Demographic Characteristics of the Subjects 52 Table 3. Socioeconomic Status by Cultural Group 53 Table 4. Correlation of Gender, Age & SES with Criterion Measures 56 Table 5. Means, Standard Deviations and Group Differences 59 Table 6. Intercorrelation Among FD-I, LOC and UNCERT Scores 60 Table 7. Summary of Standard Multiple Regression Analyses on GEFT 61 Table 8. Summary of Canonical Correlation Analysis , 64 Table 9. Summary of Item Statistics by I, P and C 66 vi LIST OF FIGURES Figure 1. A Model of the Relationships among FD-I, LOC, UNCERT, MJ and L H 4 vii ACKNOWLEDGEMENTS I would like to express my appreciation to my committee members, Dr. R. Conry, Dr. V. D'Oyley and Dr, J. W. Kehoe, for their guidance and assistance during the preparation of this thesis. I would also like to thank the teachers and the students who participated in this study. Without their help, I could not finish this thesis. My grateful thanks are also extended to Bernice Yam and Carmen Rida who scored the projective test for me. Last of all, my sincere thanks also go to my friends for their sharing when I was down. From them, I am learning to accept things that I can never change. viii I. R E S E A R C H P R O B L E M A N D R E L A T E D R E S E A R C H Cognitive functioning is a research area that currently demands a great deal of attention in the fields of psychology and education. Thousands of studies examining cognitive functioning have been conducted over the past three decades. The present study was undertaken to investigate three related cognitive constructs in a cross cultural perspective. The constructs chosen are: field dependence-independence (FD-I) (Witkin, Dyk, Paterson, Goodenough & Karp, 1962; Witkin & Goodenough, 1981), locus of control (LOC) (Rotter, 1966) and uncertainty orientation (UNCERT) (Sorrentino, Short & Raynor, 1984). FD-I and LOC are two popular constructs which have stimulated a wide range of research studies in social and behavioral science. FD-I, defined by Witkin et. al. (1962), is a bipolar dimension reflecting characteristic styles of intellectual functioning. Field independent (FI) persons tend to use internal referents as their primarj' guides in information processing, whereas field dependent (FD) persons are more likely to use external referents. Similarly, Rotter (1966) proposed LOC construct. Internal-external LOC was proposed as generalized beliefs or perceptions about how positive and negative reinforcements are obtained. An internal individual perceives outcomes or events to be based on his own personal actions. On the other hand, an external person perceives outcomes to be randomly controlled by outside forces such as chance, fate, luck, powerful others or social systems. UNCERT, proposed by Sorrentino et. al. (1984), is a recently developed construct. It was portrayed as a cognitive dimension with \"uncertainty orientation\" and \"certainty orientation\" at the two extremes. Uncertainty oriented individuals are hypothesized to be dominated by a cognitive need to maintian clarity by 1 RESEARCH PROBLEM AND RELATED RESEARCH 1-2 discovering the unknown, while certainty oriented individuals try to maintain clarity by avoiding confusion. Close examination of past studies in these areas discloses that most have been conducted in North America, an uncomfortably narrow cultural context, given the generality of some of the inferences drawn. Little information is available about how people in the other cultures function in these cognitive domains. (Cross cultural studies in these areas are discussed in the following sections.) As a consequence, the generalizability of these psychological constructs may be ethnocentrically limited to North Americans. North American samples are perhaps not entirely representative of humankind; having different cultural values and world views, people in the other cultures may function differently from North Americans. The present study was undertaken to examine how Chinese and North Americans express themselves in the domains of FD-I, LOC and UNCERT. The relationships among these constructs and culturally based differences in those relationships were also explored. Finalty, this study was the First exploratory investigation of the model suggested below. This paper is organized into two parts. The first part presents a tentative model of inter-relationship among field dependence-independence (FD-I) (Witkin, Dyk, Paterson, Goodenough & Karp, 1962; Witkin & Goodenough, 1981), locus of control (LOC) (Rotter, 1966), uncertainty orientation (UNCERT) (Sorrentino, Short & Raynor, 1984), moral judgement (MJ) and learned helplessness (LH) (Seligman, RESEARCH PROBLEM AND RELATED RESEARCH / 3 1975). The second part presents a study testing Proposition 1 in the model. A. A MODEL OF THE RELATIONSHIPS AMONG FIELD DEPENDENCE-INDEPENDENCE, LOCUS OF CONTROL, UNCERTAINTY ORIENTATION, MORAL JUDGEMENT AND LEARNED HELPLESSNESS This section portrays a model of the relationships among the five cognitive constructs identified above. Previously, these constructs have been treated in isolation or, at best, in groups of two or three. Here, they are treated as essential components of a comprehensive model. In the model, the variable \"cultural values\", generally defined as cultural conceptions about what are desirable goals to pursue and what are appropriate standards for judging actions, is another essential component. All the constructs in the model are considered as pertaining to different cultures, \u00E2\u0080\u0094 in this instance, to Chinese and North American populations. (Characteristics of Chinese and North American cultural values are elaborated later in this chapter.) All of these constructs interact, and the model posits that each is understood only when examining the influence of culture. The listed \"propositions\" are consistent with the logic of the model showed; the following chapter addressed the scarcity of empirical findings which might be used to make the propositions more specific. For example, in the face of present knowledge, it is difficult to predict whether Proposition 2 or 4 is true in the same way for both Chinese and Western cultures. RESEARCH PROBLEM AND RELATED RESEARCH / 4 - autonomous MJ ~ conventional MJ Figure 1. A model of the inter-relationship among field dependence-independence, locus of control, uncertainty-orientation, moral judgement and learned helplessness NOTE. FD-I: Field Dependence-Independence LOC: Locus of Control MJ: Moral Judgement LH: Learned Helplessness UNCERT: Uncertainty Orientation The solid arrows indicate that there are causal relationships between the corresponding constructs. The dotted arrows indicate that there may not be a direct causal relationships between the corresponding constructs. The dotted rectangular in the model represents the hypotheses of the present study. R E S E A R C H P R O B L E M A N D R E L A T E D R E S E A R C H / 5 Propositions l a Chinese are more F D than North Americans, lb Chinese are more external L O C than North Americans. Ic Chinese are more certainty-oriented than North Americans. 2a Individuals with F D are more conventional in M J than those with FI. 2b Individuals with external L O C are more conventional in M J than those with internal L O C . 2c Certainty-oriented individuals are more conventional in M J than uncertainty-oriented individuals. 3 Chinese are less susceptible to L H than North Americans. 4 Individuals with conventional M J are less susceptible to L H than those with autonomous M J . In this study, Chinese people refers to those who are born and educated in Hong Kong, Taiwan or Mainland China. B. LITERATURE REVIEW 1. Proposi t ion l a - l c : Chinese are more field dependence, external locus of control & certainty-oriented than Nor th Americans a. Field dependence-independence and cross cultural studies comparing Chinese and North Americans FD-I is a name given to a cognitive style defined by Witkin and his colleagues RESEARCH PROBLEM AND RELATED RESEARCH / 6 as an integral part of his broader concept of psychological differentiation (see Witkin & Goodenough, 1981). Psychological differentiation involves the development of specialization and articulation of psychological functions. The more differentiated the individual is, the more autonomy he has in his psychological functionings. Differentiation theory emphasizes that articulation and specialization are characteristic of the development of the organism as a whole. That means, greater differentiation in one psychological domain is indicative of a similar tendency in another domain, making for \"self consistency\" (Witkin & Berry, 1975). If an individual easily differentiated figures hidden or \"embedded\" in an organized background, the individual would also easily carry out an intellectual differentiation task. The basic assumption of self-consistency across all domains allows psychologists to make inferences about an individual's 'cognitive style' on the basis of his/her 'perceptual style'. FD-I, one of a number of cognitive styles, is a bipolar dimension reflecting characteristic styles of intellectual functioning. Field independent (FI) individuals would be those whose cognitive styles reflect more differentiation while field dependent (FD) individuals are those with relatively less differentiation. It is proposed that FI persons are more likely to use internal referents as primary guides in information processing, whereas FD persons use more external referents. A number of studies have demonstrated differences in behaviors as a function of FD-I (see Witkin & Goodenough, 1977; Witkin & Goodenough, 1981), those which are most relevant to this study indicate that FI persons have an ability to restructure a perceptual situation or impose a structure by ignoring the irrelevant RESEARCH PROBLEM AND RELATED RESEARCH / 7 context (e.g., Fineley, Solla & Cowan, 1977; Messick & French, 1975). Because FI persons use internal referents to solve a problem, they are not easily disturbed by irrelevant external information. On the other hand, FD individuals find it more difficult to ignore irrelevant information and to impose a structure on unstructured situations. Progressing a more articulated experience of self, the FI person is also found to have a more developed sense of separate identity; such persons recognize their own needs, feelings, attributes as being separate from those of others. This means that FI persons are guided more by their own feelings, thoughts and needs than by those of others in their actions and self-definition. As a consequence, they are less likely sensitive to social cues provided by others, less concerned with people, show both physically and emotionally distancing from people and prefer nonsocial situations. Conversely, FD persons tend to rely more an external referents for a definiton of their feelings, attitudes, thoughts and sentiments. They are also more attentive to social cues, more interested in others and prefer interpersonal situations. Many studies have found association between FD-I and the way the individual deals with social environment (e.g., Birnhaum, 1975; Holley, 1972). It seems that FD and FI people are different in an array of characteristics that make FD people have a set of social skills that are less evident in FI people. On the other hand, FI people have greater skill in cognitive analysis and structuring. Thus, each of these cognitive styles has characteristics that are adaptive to particular situations. With respect to origins of differences in FD-I, Berry (1966) and Berry & Witkin (1975) suggested an eco-cultural perspective to Witkin's psychological differentiation theory and provided researchers interested in other cultures with a very useful RESEARCH PROBLEM AND RELATED RESEARCH / 8 guide for investigation. According to the Berry's perspective, social, cultural and ecological factors are three interactive factors that directly influence the development of FD-I orientation. Only social and cultural factors are discussed here since these are the two relevant factors in the present study. The work of socialization as a social antecedent of FD-I focuses mainly on child-rearing practices. By and large, childrearing practices which foster the development of FD are those which encourage conformity to parental and social norms, and do not allow for flexibility in parental authority (e.g., Baran, 1971; Dawson et. al., 1974). \"Culture\" has been conceived to affect the development of FD-I through influencing the choice of preferred ideology and mode of socialization. In the literature of psjrchological differentiation, types of culture are classified by the criteria of \"social tightness\". \"Tight\" cultures are those with many roles in a structured hierarchy and conformity as their positive cultural values. On the other hand, \"loose\" cultures are those have fewer rigidly assigned roles and a less rigidly structured hierarchy. Training for autonomy is emphasized in these cultures. Man}' studies have showed that \"tight\" cultures are associated with FD and \"loose\" cultures are associated with FI (e.g., Amir, 1972; Okonji, 1969; Witkin et. al, 1974). Since Berry's work (Berry, 1966; Berry & Witkin, 1975), cross cultural studies in FD-I have proliferated. However, only two studies have involved Chinese subjects and neither of these directly examined the influence of Chinese culture on FD-I. Dawson et. al. (1974), in their study of 11- to 13-year old Hong Kong Chinese, found that severity of upbringing, as assessed by Dawson's Chinese RESEARCH PROBLEM AND RELATED RESEARCH / 9 Socialization Scale completed by the children, was linked to performance on Rod-and-Frame Test, Embedded Figure Test and Block Design (three different measures of FD-I). The more severe the child's upbringing, the more likely that he/she would develop FD orientation. To investigate the ecological influence in FD-I development, Dawson (1970) also conducted a study to examine differences between Hong Kong Boat People and Hong Kong Hakka agriculturalists. As expected, he found that Hong Kong Boat People were more FI on Embedded Figure Test and Pre-school Embedded Figure Test than Hong Kong Hakka agriculturalists. b. Locus of control and cross cultural studies comparing Chinese and North Americans Based on social learning theory, Rotter (1966) proposed that an internal-external (I-E) locus of control construct moderates the effects of reinforcement and expectancy. Under the assumption that everyone seeks to account for and to explain his own behavior, LOC is hypothesized as generalized beliefs or perceptions about how positive and negative reinforcements are obtained. Such perceptions, in turn, directly influence the expectancy that a particular reinforcement will reoccur. Once an expectancy for a particular reinforcement is built up, it tends to generalize from a specific situation to a series of situations which are perceived as related or similar. This generalized expectancy or belief becomes a stable characteristic of the personality, and might affect a variety of behavioral choices in a broad spectrum of life situations. Generalized expectancies RESEARCH PROBLEM AND RELATED RESEARCH / 10 in two different persons will result in characteristic differences in behavior in a particular situation. At one extreme is internality in which a person has a general expectancy that he/she can obtain reinforcements by his/her own action. At the opposite extreme is externality in which a person has a general expectancy that reinforcements depend on such uncontrollable outside forces as chance, fate, luck or the behavior of those more powerful. According to Rotter's theortical conception (1975), the I-E LOC construct is not a xlear-cut good-bad dichotomy. Rather, it is a .continuum with no inherent value judgement imposed on either internal or external control. It would be incorrect to assume that internals have all good traits or that externals have all bad traits. Concerning the origins of different LOC orientations, interestingly, LOC and FD-I have similar social and cultural antecedents (see Lefcourt, 1982). Parents' rejecting and dominating behaviors are reported to encourage their children to believe that factors outside their own control are responsible for their action, and that fosters the development of an external LOC. The role of culture has also been examined. The research in this area suggests that external LOC is found in those cultures with social emphases on authoritarian submission, obedience to tradition and conformity to group norms. Conversely, internal LOC is associated with cultures stressing autonomy, self-determination and individualism as positive cultural values. These postulations have stimulated a considerable number of cross cultural studies on LOC. The application of the LOC concept to cross-cultural comparison has resulted in theoretical and psychometric reinfinement of the construct. As RESEARCH PROBLEM AND RELATED RESEARCH / 11 originally conceptualized by Rotter (1966), the Internal-External (I-E) Scale, the first measure of LOC orientation, was thought to be unidimensional. However, the issue of the dimensionality of the scale was raised by Gurin et. al., (1969), who found that a three-factor structure best accounted for the responses of U.S. black college students. These factors were labelled: academic performance, self confidence and aspiration. Independently, Mirels (1970) also obtained two factors, one pertaining to the personal control and the other to the political system control. This two-factor structure has been successfully replicated in several studies (Bond & Tornatsky, 1973; Cherlin & Bourque, 1974; Joe & Jahn, 1973). In their comparative study of Denmark and U.S. students, Schneider and Parsons (1970) also conceptualized five factors (i.e., luck-fate, personal respect, politics, leadership-success and academics) on the basis of a content analysis of the I-E Scale. Taken together, these results required a re-evaluation of the dimensionality of the LOC construct. The LOC studies between Chinese and North Americans further suggest the multidimensionality of the LOC construct. A review of research revealed five cross cultural research studies involving Chinese in the area of LOC prior to 1987. Hsieh et. al. (1969) performed the first cross cultural study between American and Chinese on LOC. He and his associates administered the Rotter's I-E Scale to three groups of high school students (Hong Kong Chinese, U.S.-born Chinese and U.S. Caucasians), matched on age and statistically controlled for SES. As expected, Hong Kong Chinese students were the most external; whereas U.S.-born Chinese were in between Hong Kong Chinese and the more internal U.S. Caucasians in their control orientation. However, Tsui's study (1977) . RESEARCH PROBLEM AND RELATED RESEARCH / 12 partially contradicted what Hsieh et. al. found by demontrating that Hong Kong Chinese college women were more internal than American-Chinese women on the Rotter's I-E Scale. Using a different LOC instrument \u00E2\u0080\u0094 Lefcourt's Multi-dimensional Multi-attributional Causality Scales (1979) \u00E2\u0080\u0094 Chan (1981) compared different cultural groups (Anglo-Canadian, Hong Kong Chinese, and Hong Kong Chinese foreign students in Canada) and different age level (college students), and found that Hong Kong Chinese students were more externally controlled than Canadian students on the overall I-E scores. Subscale analysis indicated that this difference was mainly due to differences in an achievement domain, where Hong Kong Chinese students were more external than the Canadian students. Foreign students were in the middle between Hong Kong and Canadian students. No significant difference was obtained in an affiliation scores. A similar result was also obtained by Lao (1977), who compared Taiwan Chinese and American college students. She found that Taiwan Chinese ( females, but not males) to be more external on the 'Internal' subscale of the Levenson IPC Scale. However, the relationship was reversed on the 'Powerful Others' subscale, where U.S. subjects (males and females) appeared to be more external than Taiwan subjects. No significant difference was reported on the 'Chance' subscale. These inconsistent findings on the 'Powerful others' and the 'Chance' subscales may be due to the unusual external orientation of the Lao's white sample ( see Dyal, 1984, p.226). RESEARCH PROBLEM AND RELATED RESEARCH / 13 Using Crandall's IAR Scale (Crandall, Katkovsky & Crandall, 1965), Chiu (1986) conducted another study comparing Taiwan Chinese and U.S. Caucasians. His subjects were 6th and 8th graders and they were matched on SES, age and grade. Results indicated that American children were more internally controlled for success than their Chinese counterparts; however, they were more externally controlled for failure than Chinese children. All these studies clearly showed that multidimensional LOC instruments (e.g., Levenson's IPC Scale, Lefcourt's MMCS) have provided greater insight into differences between cultural groups than unidimensional instruments (e.g., Rotter's I-E Scale) wherein only total scores are compared. c. Uncertainty orientation and cross cultural studies comparing Chinese and North Americans UNCERT is a recently developed construct. To date, little information is available about the characteristics of this construct, or its relationship with other phenomena. Sorrentino et. al. (1984) and Raynor and Zubek (1984) proposed UNCERT as a cognitive construct. At one end of the hypothesized dimensions, uncertainty-oriented individuals are dominated by a cognitive need to attain clarity by discovering the unknown. They are more likely to integrate new events or beliefs into already existing belief systems, changing the new or old belief accordingly. Logically, they are more rational and have a 'need to know'. Uncertainty-oriented persons can exhibit negative as well as positive behaviors in order to attain clarity about themselves or their environment. At the other end of the dimension, certainty-oriented individuals are characterized by a cognitive RESEARCH PROBLEM AND RELATED RESEARCH / 14 need to maintain clarity by avoiding confusion. In order to avoid confusion, certainty-oriented individuals are thought to adhere to familiar events and traditional beliefs, to be dependent, and to have a low sense of autonomy in an unfamiliar environment. Research in this area has been restricted to the North American cultural context; no cross cultural research has been done except one unpublished study (Lau, 1985). To investigate cultural differences in learned helplessness as a function of uncertainty orientation, Lau (1985) conducted an experiment on approximate 41 Hong Kong foreign students and 43 Canadian Caucasians at the University of Western Ontario, Canada. As predicted, Hong Kong foreign students were found to be more certainty-oriented than the Canadian counterparts; however, no cultural difference was found in learned helplessness. The cross cultural study in this area is so sparse that further cross-cultural investigation is required to clarify the role of cultural influence on UNCERT. d. Chinese Cultural Characteristics Literature reviewed earlier in this chapter indicated that the social and cultural environment are very important to the individual's development of FD-I, LOC and UNCERT orientations. Chinese culture, with more than 4,000 years of civilization, is one of the ancient cultures. As the literature showed, Chinese culture with its emphasis on collectivism and harmony is quite different from North American culture, which emphasizes individualism and competitiveness (see Hofstede, 1980; Hsu, 1970, 1981; Needham, 1956; Yang, 1981). Such broad differences in RESEARCH PROBLEM AND RELATED RESEARCH / 15 cultural values may well lead to an alternative cognitive structures or processes. In the next section, an attempt is made to employ ecocultural perspective to illustrate some major Chinese cultural qualities. According to this perspective, different ecologies make different kinds of demands on the psycholgical and physical resources of the individuals who inhabit them. In the process of adaptation to given ecologies, members of successive generations of an ethnic group develop a special type of economj' which in turn necessitates establishement of an appropriate social structure. A unique socialization pattern then evolves to train members of subsequent generations in a set of personality and behavior traits that will make them compatible with the particular social structure and function. Approximately 4,000 years ago, Chinese began to establish their society along the Yellow River which was an ideal place for the development of a self-substaining agricultural economy. This agricultural economy had particular features which led directly to the socialization of certain behavioral characteristics. For example, a traditional agricultural economy requires a large group of people working on the farm. To maintain the effectiveness of group effort, harmony and cooperation among some collective, which is usually a stable ingroup (e.g., family) are more important than the individual's needs and goals. Individuals are induced to sacrifice their personal goals for attaining collective goals, and much of the behavior of individuals is directed toward the goals of this ingroup. In other words, the individual sublimates collective goals as personal goals. Thus, collectivism, with the family as its basic unit developed gradually in China. The RESEARCH PROBLEM AND RELATED RESEARCH / 16 main feature of this social structure is its hierarchical network. In this hierarchical system, each individual has his defined roles and social power according to his/her gender, age, birth order and position in a group. Those in higher positions in the hierarchy are given more social power in controlling economic resources. In order to share these economic resources and obtain social support, those in lower positions are required to respect and obey those in higher positions. For this reason, socialization practices which emphasize obedience to constituted authority and conformity come to be valued. Consequently, individual autonomy and independence are reduced through socialization. In response to this agricultural ecology, Chinese, as a whole, are therefore relatively more passive, dependent, collective and extremely subject to group influence. These eco-cultural qualities, interacting with certain dominant moral and religious thoughts and doctrines (e.g., Confucianism, Taoism and Buddhism), would lead to a special mode of socialization through which a traditional Chinese learns how to live effectively in the social network of this agricultural society. A detailed discussion of Chinese thought is not possible here; and any such analysis would be complicated by the enormous diversity of Chinese thought in different places and times. Therefore, the following discussion is limited to Confucianism, which has dominated Chinese thought for the past twenty-five centuries. Confucianism was advocated by most rulers thoroughout the history of China not only because it emphasized the loyalty of officers, but also because its ideology was congruent with the cultural system of traditional China, basically an agrarian state. Being a theory of society and the individual, Confucianism is tremendously rich and complex. Numerous English-language publications are RESEARCH PROBLEM AND RELATED RESEARCH / 17 available for those wishing to pursue the philosophy of Confucius in detail (e.g., Fung, 1948; Needham, 1956; Wright, 1962). Only Confucian themes which have immediate implications for the daily functioning of Chinese people are highlighted here. The concept of the universe is an essential concept in Chinese thought. In Confucianism, the concept of the universe is captured in its conception of Tien (Heaven, ). Tien is a total cosmic process; whatever is in the universe must be good, simply because it is. Confucians perceive the universe as a harmoniously functioning organism consisting of an orderly hierarchy of interrelated parts and forces, which though unequal in their status, are all equally essential for the whole. In other words, the world of man and the world of nature constitute one great indivisible unity. Man is a part, but a vital part, of the universe as a whole. This belief is further strengthened by the philosophy of Taoism, the basic principle of which is that Man must subordinate himself to Too (Nature 2^,). In the West, happiness is to be found by harnessing the forces of nature to the will of man. For the Chinese, on the contrary, when one integrates oneself into the universe as one finds it, one gains the happiness of contentment in Ching (simplicity, quiescence ^ ^ ) . At the social level, society is, or at least should be, a reflection of this universal harmony. An ideal society should be an ordered hierarchy of individuals with unequal status, all of whom, however, have essential roles to perform. As a result, cooperation and harmonious relationship are developed among people. According to Confucianism, social harmony would be realized when man practices RESEARCH PROBLEM AND RELATED RESEARCH / 18 and perfects virtues such as li (propriety, rules of correct behavior,-75! ) yi (righteousness, 4^ ,' ), chi (moral self, wisdom in a moral sense, ^ ) and jen Li, which can be conceptualized as the guidelines for performing the various roles, entails both rights and responsibilities of a man toward another man in a situation. Each individual is expected to perform his role according to the norms prescribed by li. Consequent!}', a harmonious society would be achieved. Failure to follow li would jeopardize the relationships between/among individuals and disrupt the harmon}' of society. Among these relationships, certain relationships are of paramount importance. between sovereign and subordinate, father and son, elder brother and younger brother, husband and wife, and friend and friend. As showed in this order, these relationships, even those between friends, are constructed in a hierarchical pattern. In this hierarchical structure, sovereign, father, elder brother, husband and older friend were given a wide range of privileges and authority with respect to subordinate, son, younger brother, wife and younger friend. It should be emphasized that this hierarchical structure is maintained by the inter-dependence of both parties in each relationship. Each. party in the relation is speculated to follow his role requirement on the basis of li. For example, the sovereign is required to rule justly; otherwise, his subordinates would not serve loyally. (human-heartedness, love, benovolence, These are called the wu lun (five cardinal namely: those RESEARCH PROBLEM AND RELATED RESEARCH / 19 As seen above, in the Confucian paradigm man is regarded as a relational being. Each individual is required to follow li in order to maintain a smooth relationship with others. However, relations alone do not totally define the man. Confucians also recognize great importance of chi, which can be conceptualized as wisdom in a moral sense. In the Confucian perspective, the individual is more than a role-player mechanically performing the role-related behavior prescribed by li; he is an active and reflective entity. Chi provides the individual self the capacity to adhere to what is right according to yi, which is the principle of conferring the propriety on the behavior pattern in question. In this sense, Confucianism not only emphasizes interpersonal relationships, it also stresses personal integrity and righteousness. This aspect of Confucianism justifies rebellion against tyrannical authorities even though the consequence is disobedience to one's superior. In Confucianism, chi is embodied in the Confucian conception of jen, which is the natural desire for accomplishing li in the spirit of yi. Jen was to be identified with both the totality of all virtues and the essence of all virtues. It is the most fundamental for making a man a moral being. According to Confucianism, if one is in the practice of jen, he would seek to fulfill others in order to fulfill himself, as well as to fulfill himself in order to fulfill others ( *L K, S^^lSiJtAl- It is clearly showed that helping others to fulfill themselves is a corollary of one's self-realization. In other words, to involve the other in one's self-realization is not only altruistic; it is required for one's own self-development. Therefore, being the natural desire to consider the others in one's self-realization, jen is the basis that a man relates to others and RESEARCH PROBLEM AND RELATED RESEARCH / 20 to become himself. As a summary, the essential aspect of Confucianism is the construction of a harmonious society according to li, yi, chi and jen. In the practice of these virtues, an individual is encouraged to maintain hierarchial order and harmonious interpersonal relationships in a relatively stable and permanent social structure. Such Confucian socialization teaches Chinese to accept hierarchy in which people in higher positions are given more privileges and authority to control resources. Consensual acceptance of this differential access to resources prevents conflict and competitiveness which would directly jeopardize the effectiveness of group effort and the harmonj' of the agricultural society. To maintain such interpersonal harmony, Chinese psychologically tend to act in accordance with social expectations or social norms, rather than according to the individual's needs and goals. They are more concerned about cooperativeness, social conformity, harmony maintenance, social acceptance and avoidance of conflict (Yang, 1981). As such, they would be able to function as an integral part of the social network. e. Empirical Findings on Chinese Cultural Characteristics The eco-cultural demands and Confucian philosophy just discussed are two major historical genesis of Chinese cultural characteristics. This section review some empirical findings in the study of these cultural characteristics. Among these studies, most of them are cross-cultural in nature, mainly contrasting Chinese and American cultures. Before reviewing the relevant literature, one point should be clarified. For ideological and political reasons, almost no empirical research RESEARCH PROBLEM AND RELATED RESEARCH / 21 was conducted in Mainland China in the last two decades. Because of numerous restrictions of the flow of information from Mainland China, most of the work to be discussed in this section used samples of Chinese people drawn from Taiwan and Hong Kong as subjects or respondents. Special caution should be taken when the results are generalized from one Chinese society to the other. As mentioned, collectivism and hierarchical social structure are two fundamental aspects of Chinese culture; however, there are only a few empirical studies in the literature examining these two cultural dimensions. In their comparative study on authoritarianism, Singh, Huang and Thompson (1962) provided information about the collective nature of Chinese culture. Compared with the local university students at a large American universit}', the Indian and the Chinese foreign students are found to be more society-oriented. They gave high ratings to group participation and showing more concern for other people on the Edwards' Personal Preference Scale (Edwards, 1954). More direct evidence was collected by Hofstede (1980) in his survey of work-related values. Working for a large multinational organization, Hofstede and his associates conducted a survey of 100,000 employees in 40 nations around the world. Three were Chinese nations: Hong Kong, Singapore and Taiwan. Using factor anatysis to simplify the mass of data, Hofstede derived four dimensions of cultural variation along which his nations could be located. Two of these dimenisons are of special importance for the present discussion. The first dimension is collectivism-individualism. In the Hofstede's conception, RESEARCH PROBLEM AND RELATED RESEARCH / 22 collectivism-individualism refers to the degree of interdependence that a society maintains among individuals. Individualism stands for a preference for a loosely knit social structure in which individuals are supposed to take care of themselves and their immediate families only. Its extreme, collectivism, refers to a preference for a tightly knit social framework in which individuals can expect their relatives or other in-groups to look after and support them in exchange for unquestioning loyalty. The second cultural dimension is power distance which Hofstede defined as the extent to which the members of a society accept that power in institutions and organizations is distributed unequally. People in large power distance societies accept a hierarchical order in which everybody has a defined power and resources which needs no further justification. People in small power distance societies seek equalization of power and demand justification for power inequalities. These two dimensions are important for the present discussion, because Hofstede's study provided, the first piece of direct empirical evidence on the collectivism and the hierarchical social structure of Chinese culture. Compared with the other western countries, all three Chinese samples were located in similar positions on these two dimensions. That is, they all scored considerably lower in individualism and moderately high in power distance. In contrast, American and Canadian employees obtained the top scores in individualism and moderately low in power distance. Similarity in position among Chinese societies was also obtained in a more recent research study (Chong, Cragin & Scherling, 1983). Replicating Hofstede's study, RESEARCH PROBLEM AND RELATED RESEARCH / 23 Chong and his associates administered a similar value survey to a wide variety of managerial employees in Mainland China. Comparing their results with those of Hofstede (1980), they found a close correspondence among all four Chinese samples on the dimensions of collectivism-individualism and power distance. These two studies reveal an important point: collectivism and hierarchical social structure, as two fundamental Chinese values, have remained intact among modern Chinese societies. To maintain group integrity and interpersonal harmony, Chinese people are required to cooperate with each other and try to avoid competition and conflict with their in-group members. Consequently, the Chinese shift their position on an issue toward the majority position or toward the authority's position more frequently than people in more individualistic cultures like those in North America. Surprisingly, only very few empirical studies have examined this aspect of Chinese behavior. For example, Li, Cheung and Kau (1979), who presented their subjects with Madsen's game board (1971), found that schoolchildren in Hong Kong and Taiwan continued to cooperate even when the reward structure was changed to reinforce competitiveness. Similarly, Cook and Chi (1984) found American-Chinese boys to be more cooperative than American boys on their modified Madsen's game board. Authoritarianism and social conformity are two Chinese cultural characteristics which have received relative^ more attention from social scientists. Numerous empirical studies have showed that Chinese people exhibit more authoritarianism and social conformity. Huang and Harris (1973) found that Chinese college RESEARCH PROBLEM AND RELATED RESEARCH / 24 students from Taiwan imitated a college professor to a greater extent than did American counterparts. Along the same line, Chu (1966) also found that Chinese subjects from Taiwan were more affected by the mass media than were American subjects in a similar study performed earlier by Janis and Field (1959). Using different populations (i.e., Chinese secondary school students in Hong Kong versus similar British students), Lai (1981) obtained similar results. The first set of Chinese data on authoritarianism was collected by Singh, Huang and Thompson (1962) in their comparative study of Taiwan Chinese, American and Indian university students. The Chinese scored slightly lower than Indians, but much higher than their American counterparts on authoritarianism as measured by Sanford and Older's Authoritarian Scale (1950). In another study, Meade and Whittaker (1967) gave the English version of the California Fascism Scale (F Scale) (Adorno et. al., 1950) to Chinese university students from Hong Kong and five other groups of students from India, Brazil, Arabia, Rhodesia and the United States. Indian and Rhodesians showed the highest level of authoritarianism. Chinese came next, followed by Arabian, Bralizian and the American students. Relevant data were also reported by Domino and Mo (1987). In this study, a set of 701 stories generated by 80 Beijing Chinese and 80 American children was content analyzed. The results showed that Chinese stories clearly reflected social orientation, greater concern with authority, containing more affective elements and fewer instances of physical aggression than the stories generated by American children. Conclusively, the foregoing discussion and empirical findings revealed that Chinese RESEARCH PROBLEM AND RELATED RESEARCH / 25 culture is a typical collective culture with its emphasis on harmony, inter-dependence, loyalty to the group and submission to authority. With such cultural characteristics, it is reasonable to speculate that in China people are socialized to be more FD, external LOC and certainty-oriented than in North America where the culture stresses individualism and competitiveness. f. Relationships among field dependence-independence, locus of control and uncertainty orientation The other interesting field of research on FD-I, LOC and UNCER is their inter-relationships. Conceptually, FD-I, LOC and UNCERT are three similar constructs; they should be correlated. FD-I and LOC have similar conceptual frameworks; both describe individuals on the basis of whether they will rely on their own internally or externally derived frame of reference to guide their behaviors (Lefcourt, 1976, p.11; Witkin & Berry, 1975, p.9). Also, as discussed previously, both have similar hypothesized soico-cultural antecedents. Those cultures which place positive values on social conformity and submission to authority are speculated to socialize individuals with FD and external LOC orientations, while those which emphasize individualism and self-reliance are expected to socialize individuals with FI and internal LOC orientations. There is also empricial evidence which suggests that a number of psycho-social phenomena associated with LOC orientation bear a similar relationship with FD-I orientation. These psycho-social phenomena include: conformity, creativity and achievement behavior (e.g., Phares, 1976; Witkin & Goodenough, 1977), reliance on one's own reinforcement history as opposed to others' norms (e.g., Deever, 1967), the RESEARCH PROBLEM AND RELATED RESEARCH / 26 response to autonomy in reaction time task (e.g., Bachrach, Huesmann & Peterson, 1977; Caring, 1970; Mas, 1984; Schleifer, 1972). Some researchers (e.g., Bloomberg & Soneson, 1976; Guthrie, 1985; Lefcourt & Telegdi, 1971; Panda, 1971) have also reported that the combination of LOC and FD-I will yield greater precision in predicting cognitive activity, verbal productivity and moral judgement than using just one of them. A number of studies (Chance & Goldstein, 1971; Feather, 1967; Lefcourt & Telegdi, 1971; Mclntire & Dreyer, 1973; Rotter, 1966) have indicated no direct relationship between the LOC and FD-I constructs despite the logical soundness. Among these studies, Rotter's I-E Scale which has been criticized for its unidimensionality was usually employed as the instrument for assessing LOC. Therefore, it is fruitful to investigate further the relationship between FD-I and LOC constructs by employing other multi-dimensional LOC measures. Conceptually, the construct UNCERT should also correlate with the constructs LOC and FD-I. As defined, certainty-oriented individuals are dominated by a cognitive need to maintain clarity by avoiding confusion. Such persons are more comfortable with those situations with a high degree of certainty. External referents, such as social norms, traditional beliefs or standards defined by authority (e.g., family or the law), are very powerful in providing this kind of certainty to certainty-oriented individuals. Logically, certainty-oriented individuals should rely more on well-established external referents. Interestingly, all of these behavioral characteristics are also hypothesized to be associated with FD and external LOC individuals (see Lefcourt, 1982; Witkin & Goodenough, 1981 for RESEARCH PROBLEM AND RELATED RESEARCH / 27 review). On the other hand, uncertainty-oriented persons are more likely to attain clarity by discovering the unknown. Logically, they are more rational, analytical, and challenge-seeking and less likely be persuaded by external information. Again, these behavioral characteristics of uncertainty-oriented persons are similar to those of FI and internal LOC individuals. As showed in the foregoing discussion, it seems that FD-I, LOC and UNCERT are three related constructs, at least at the conceptual level. 2. Proposi t ion 2a - 2c: F ie ld dependent, external controlled and certainty-oriented individuals are more conventional i n mora l judgement than field independent, internal controlled and uncertainty-oriented individuals a. Conventional versus autonomous moral judgement Adapted from Kohlberg's three levels of moral development (1976), MJ is classified into two categories in this study. (Kohlberg's three levels of moral development are: the post-conventional level, the conventional level, and the preconventional level. These levels of moral development are presented in an ascending order; higher levels of moral development represent greater maturity in MJ.) Similar to Kohlberg's conventional level of moral development, conventional MJ refers to those who maintain, support and justify the expectations of the family, group or nation as their own individual's moral standards. On the other hand, autonomous MJ, which is analogous to Kohlberg's postconventional level of moral development, refers to those who have their own subjective values and RESEARCH PROBLEM AND RELATED RESEARCH / 28 expectations that may or may not correspond to expectations of the family, group or nation. Although in the Kohlberg's perspective, a higher level of moral development represents more maturity in MJ, ho such value judgement should be applied to the present classification. Here, differences in MJ reflect only different orientations individuals have in dealing with moral issues. There is intended no implication that autonomous MJ is inherently better than conventional MJ; each orientation can be adaptive in a particular eco-cultural context. In addition, the possible motives for developing conventional MJ are manifold, including the desires to pursue collective goals, to avoid conflict, to be correct, to be dependent on authority, to ingratiate oneself with the source of influence and so forth. Thus, it would be short-sighted simply to predict that people with conventional MJ are more dependent and passive than those with autonomous MJ. b. Relationships among field dependence-independence, locus of control and moral judgement As the theoretical definitions impfy, FI and internal LOC individuals are those who rely on an internally derived frame of reference to guide their behaviors, while FD and external LOC individuals rely more on an externally frame of reference (Lefcourt, 1976, p . l l ; Berry, 1975, p.9). Logically, FD and external LOC individuals are more likely to have conventional MJ which, to a certain extent, requires reliance on some powerful external sources of information \u00E2\u0080\u0094 such as norms defined by the authority. However, at present, research in this area is sparse and there is no conclusive evidence to support this relationship empirically. RESEARCH PROBLEM AND RELATED RESEARCH / 29 Several lines of research seem to provide evidence of the relationship among FD-I, LOC and MJ. In investigating social conformity as a function of FD-I and LOC, a number of studies reported that FI and internal LOC individuals are relatively autonomous and independent from conformity influence than FD and external LOC individuals (c.f. Lefcourt, 1976, 1982; Witkin & Goodenough, 1977, 1981 for review). These studies seems to support that FD and external LOC persons tend to have conventional MJ which, to some extent, requires conformity to authority. To date, only a few studies have examined directly the relationships among FD-I, LOC and MJ. Only three studies have attempted to relate MJ to FD-I; Kohlberg's Test of Moral Development was the common measure used to assess the subjects' moral maturity. All of these studies reported that FI are associated with more mature MJ and FD with less mature MJ (Caring, 1971; Guthrie, 1985; Schleifer, 1972). Although a positive correlation between FD-I and MJ was obtained in these three studies, more research is definitely needed to explore such a relationship. Cross cultural studies would be required to examine whether such a relationship exists in other cultures. The singular relationship between MJ and LOC has been explored more thoroughly. Among such studies, four reported moderate positive correlations between LOC and MJ (Adams-Webber, 1969; Bachrach, Huesman & Peterson, 1977; Guthrie, 1985; Penk, 1969) and three showed no correlation (Alker & Poppen, 1973; Guttman, Bar-Zohar & Statter, 1981; Janzen & Boersma, 1976). Again, Kohberg's Test of Moral Development was the most common measure used in these studies. Regarding the complex relationship among FD-I, LOC and MJ, Arbuthnot (1971), using the Kohlberg's moral judgement interview, the RESEARCH PROBLEM AND RELATED RESEARCH / 30 Rotter's I-E Scale, the Rod-and-Frame Test, and the Guilford-Zimmerman M-F Scale as a measure of relative masculinity, found that MJ was not associated with LOC and could only predict FD-I in masculine men. Females, however, were not included in Arbuthnot's study. On the other hand, Bloomberg and Soneson (1976) found that among female college students, those with post-conventional MJ are also both field independent and internally controlled. Additionally, the- results indicated that neither of these variables alone adequately predicted levels of MJ; rather, a combination of both was required. However, males were not included in Bloomberg and Soneson's sample. Guthrie (1985) suggested one possible explanation for the lack of unequivocal evidence in the study of the relationship between LOC and MJ. She suggested that lack of a standardized method for evaluating LOC may lead to difference between samples as to which subjects are classified as 'internal' and which are classified as 'external'. This factor becomes quite critical when the results of various studies are compared. This problem will affect studies of LOC and MJ until a common method for evaluating LOC is developed. The other possible explanation for inconclusive evidence in this area may be that there is, indeed, no direct causal relationship between LOC and MJ. Perhaps, there are important mediators in between LOC and MJ, which have not been included in previous studies. Further investigations are necessary to resolve this issue. RESEARCH PROBLEM AND RELATED RESEARCH / 31 c. Relationship between uncertainty orientation and moral judgement It is expected that certainty-orientation, which is characterized by a cognitive need to maintain clarity by avoiding confusion, will correlate negatively with autonomous MJ. Autonomous MJ, defined as the establishment of one's own value system and expectations which may or may not correspond to the values of one's family and nation, provide uncertainty. Thus an individual who is oriented toward certainty will likely avoid autonomous thinking. Such individuals may avoid confusion in the moral realm by following convention and stressing duty as defined by such authorities as the family or the law. Consequently, certainty-oriented persons tend to have conventional MJ which, to a certain extent, requires reliance upon authority; such reliance enables the certainty-oriented individual to maintain clarity with respect to moral issues. Following the same logic, uncertainty orientation is hypothesized to be associated with autonomous MJ. 3. Proposition 3: Chinese have less learned helplessness experience than North Americans Seligman and his associates (Overmier & Seligman, 1967; Seligman, 1975) have proposed that L H occurs when organisms learn that their responses are independent of desired outcomes, and that they are therefore helpless to initiate a change in their later circumstances. These researchers have stressed that a perceived lack of control over circumstances is the crucial factor in producing LH. One could argue that the underlying assumption of L H is the perception that an RESEARCH PROBLEM AND RELATED RESEARCH / 32 individual, as a self-reliant creature, can control his environment. An individual who perceives that he can control his environment will have stronger negative psychological responses when he encounters uncontrollable experiences than one who perceives that he cannot control his environment. It is reasonable to speculate that the former group of individuals is more susceptible to LH. Expanding this argument to the cultural level, it can be hypothesized that people in a particular culture may be more likely to experience L H than those in other cultures due to their different perceptions about themselves imposed by their cultures. As- the foregoing discussion revealed, Chinese culture is a typical collective culture with positive values on harmony, interdependence and conformity to group norms. With these cultural characteristics, Chinese people are used to coping with most of their life events collectively, especially in consort with their ingroup members. Unlike North Americans, Chinese people do not perceive themselves as totally self-reliant individuals; rather, they are interdependent members of groups. Such intensive self-ingroup relationship and collective coping make it easier for Chinese people to cope with the uncontrollable experiences in their everyday lives. Following this line of argument, Chinese people, with an emphasis on harmony and collective coping, are expected to be less susceptible to L H than North Americans whose culture stresses individualism and competitiveness. RESEARCH PROBLEM AND RELATED RESEARCH / 33 4. Proposi t ion 4: Individuals wi th conventional moral judgement are less susceptible to learned helplessness than those wi th autonomous moral judgement The derivation of Proposition 4 is completely based on logical deduction of the first three propositions. Once the first three propositions are confirmed empirically, Proposition 4 can be deduced. At present, no empirical evidence has been obtained to directly support this proposition. C. CHAPTER SUMMARY This literature review addressed the logical reasoning and empirical evidence relating to each hypothesis derived from the suggested model. As the review indicated, these hypotheses are based more on logical reasoning than on empirical evidence. At present, available research data in the related areas are so sparse and insufficiently clear that it is not yet possible to support any hypotheses about the inter-relationships among FD-I, LOC, UNCERT, MJ and LH. Therefore, much more empirical work should be done to modify and to confirm the suggested model. In the next chapters, an empirical study is designed and presented as a test of Proposition 1 in the model. Note that the other propositions are not tested in the present study. For a master thesis, testing all the propositions is not managable. Also, a relatively cultural-free moral judgement instrument has not yet be developed for testing Proposition 2. II. NATURE OF THE STUDY A. RESEARCH OBJECTIVES The present study is a cross cultural study comparing North Americans and Hong Kong Chinese on FD-I, LOC and UNCERT. The purpose of this study is two-fold: to investigate American-Chinese differences on FD-I, LOC and UNCERT and to examine cultural diferences in the inter-relationships among these cognitive constructs. The study was designed as an empirical test of Proposition 1 in the model elaborated in the previous chapter. Besides being the first step in confirmation of the suggested model, this study may have some contributions to filling the gaps in research on FD-I, LOC and UNCERT. Although there is increasing interest in cross cultural studies, very few investigations can be found in the literature dealing with comparisons between North American and Chinese cultures on FD-I, LOC or UNCERT. The literature review in Chapter I indicated that only one study has examined the cultural difference on UNCERT and only two studies have investigated the influence of Chinese culture on FD-I \u00E2\u0080\u0094 more precisely, the influence of Hong Kong culture on FD-I. Even though the cross cultural differences in LOC has been explored more thoroughly, it is difficult to compare studies in this area because of sample variations in age, cultural background and in the LOC instruments used. Among these five studies, three (Chan, 1981; Hsieh et. al., 1969; Tsui, 1977) have used Hong Kong students as their subjects. However, their results were inconclusive regarding the LOC orientation of Hong Kong Chinese. In summary, the number 34 NATURE OF THE STUDY / 35 of FD-I, LOC and UNCERT studies involving Chinese are so few and so inconclusive that it is not yet possible to describe the expression of FD-I, LOC or UNCERT in Chinese culture. As a result, little information is available about how the Chinese who, as a whole, occupy a quarter of the world population, function in these cognitive domains. Do Chinese function differently from North Americans in these domains? The present study addresses this question. The second question regarding FD-I, LOC and UNCERT constructs relates to their inter-relationships. The present study is not only the first to examine the inter-relationships among these cognitive constructs, but also the first investigation to address the comparability of these relationship across cultures. As the foregoing discussion shows, these three constructs are inter-related conceptually. They have similar patterns of correlation with certain personality and cognitive phenomena. A combination of FD-I and LOC has also been reported to yield a greater precision in predicting some behavioral characteristics. However, most of the past studies have showed no direct relationship between FD-I and LOC. Such results may be due to methodological flaws. Most of these studies have employed Rotter's I-E Scale, which has been criticized for its unidimensionality, to assess subjects' LOC orientation. Thus, it is necessary to explore further the relationship of FD-I and LOC constructs by using multidimensional instruments. Levenson's IPC Scale (1972), which measures the LOC construct in three dimensions, was used in the present study to explore the relationship of FD-I and LOC. The other advantage of employing Levenson's IPC Scale is that it should provide more detail about domains in which North Americans and Hong Kong Chinese differ in their control perceptions. NATURE OF THE STUDY / 36 Practically, this study may also have some implications for acculturation and for achieving constructive communication between North American and Chinese cultures. In this era of free trade and high technology, the active interchanges of human resources, merchandise and information between different nations has increased. When different cultures encounter each other, it is not uncommon for one try to assimilate or reject the other. Through the processes of conflicts and resolutions, constructive communciation between different cultures develops. Information about how the people in the other culture function in certain cognitive domains would promote communciation and mutual understanding between different cultures. In addition, this kind of information would help to establish a broader views and more open attitudes toward the other culture without being pre-occupied with biases or impeded by rivalry. This is especially valuable in a multicultural country such as Canada. It is also with this concern that the present study was undertaken. B. RESEARCH QUESTIONS The general objective of this study is to investigate American-Chinese differences in FD-I, LOC and UNCERT, as well as the inter-relationships among them. In pursuit of this objective, four specific research questions were formulated: 1. Are Hong Kong Chinese more FD than Canadians? 2. Are Hong Kong Chinese more external LOC than Canadians? 3. Are Hong Kong Chinese more certainty-oriented than Canadians? 4. How does the pattern of inter-correlations among FD-I, LOC and UNCERT differ between Hong Kong Chinese and Canadians? NATURE OF THE STUDY / 37 C. CHAPTER SUMMARY This chapter introduced the objective and the specific research questions of the study. A cross-cultural study comparing North Americans and Hong Kong Chinese was posited as a test of Proposition 1 in the suggested model (see Chapter I). Other theoretical and practical implications of the research questions were also discussed. This study has potential for increasing the understanding of how Chinese function in these cognitive domains and the inter-relationships among these constructs. Also, it may stimulate further cross-cultural studies in these areas and therefore increase the likelihood of communication between different cultures. III. M E T H O D A . SUBJECTS & SAMPLING PROCEDURES The subjects used in this study were drawn from two groups of College students: 1. Canadians taking an introductory psychology course at Langara Community College in Vancouver, Canada, and 2. Hong Kong Chinese in Businesss Studies at Caritas Vocational College in Hong Kong On the initial design, two groups of Chinese subjects (i.e., Hong Kong Chinese and Mainland Chinese) were planned to determine whether there is any differential cognitive functioning among Chinese in different social and cultural contexts. However, due to delays in acquiring approval from corresponding universities in Mainland China, only the Hong Kong group of Chinese subjects was ultimately selected. The two sample groups were selected as convenience samples; recognizing that they may not be precisely representative of their corresponding populations, results should be interpreted cautiously. Subject Selection Criteria. Students were selected on the basis of their cultural background and their consent to participate in the study. A student was recruited if he/she met all the following criteria: 1. Must have lived in the homeland for minimum 10 years. 2. Must be a native speaker of the language of his/her culture. In other words, Hong Kong Chinese speak Cantonese in their home communication whereas 38 METHOD / 39 Canadians speak English in their home communication. 3. Must have been educated in the respective culture. 4. Must have attended class on the day that the test was administered. 5. Must have agreed to participate in the study. (The consent form is presented in Appendix A.) Of the students approached (39 Chinese students and 58 Canadian students), all the Chinese students were recruited as they all met the selection criteria. In the Canadian group, 41 students were eligible. Of the ineligible students, 11 were classified into other cultural groups according to the first three criteria, while the cultural background of the remainders could not be identified because these students supplied no demographic information. B. INSTRUMENTS Instruments employed in this study included: a sentence interpretation test as a projective measure of need for uncertainty; a \"general attitude survey\", which consisted of Levenson's IPC Scale and Byrne and Lamberth's measure of authoritarianism; a figure test which was employed to assess the subjects' FD-I orientation and a demographic questionnaire. All instruments are presented in Appendix B. Descriptions of the instruments are given as follows. 1. F ie ld Dependence-Independence Instrument METHOD / 40 To assess subject's FD-I, the Group Embedded Figure Test (GEFT) (Witkin, Oltman, Raskin & Karp, 1971) was employed. The GEFT was selected because it is one of the common FD-I measures used in cross-cultural studies. The GEFT was designed to allow group testing of the original individually administered Embedded Figure Test (EFT). It contains 18 complex figures, 17 of which were taken from the EFT while the other one was originally designed for the GEFT. The test contains three sections: the first contains 7 very simple items and is basically for practice, and the second and the third sections, each of which contains 9 more difficult figures. Subjects are required to trace a simple form (which is given on the back page) on the corresponding complex figure in a certain time limit. The score is the total number of simple forms correctly traced in the second and third sections combined. Omitted items are scored as incorrect. The items in the first section are not included in the total score. The more correct simple forms that the subject traces, the more FI he/she is. According to the GEFT manual (Witkin et. al., 1971, p.27-28), it indicates that GEFT is a very reliable test. The split-half reliablility coefficient is .82 for both males (N = 80) and females (N = 97) at an eastern liberal art college. In addition, a high correlation has been obtained between scores on GEFT and EFT, with a coorrelation of .82 for male undergraduates and .63 for female undergraduates. The combined evidence suggests that the GEFT is a useful test for assessing FD-I orientation. METHOD / 41 2. Locus of Contro l Instrument Levenson's IPC Scale (1972) was chosen as the measure of LOC in the present study because of its multidimensional nature. For the purpose of cross-cultural study, multidimensional instruments are more desirable than unidimensional instruments, because they offer a more detailed picture about dimensions in which cultural groups may differ. Also, it was judged that Levenson's IPC Scale would be easier to translate into Chinese language than other common multidimensional LOC instruments, (e.g., Lefcourt's MMCS, 1979). Levenson's IPC Scale was designed to measure the I-E concept more accurately, through utilizing three dimensions of control \u00E2\u0080\u0094 belief in personal control (I subscale), powerful others (P subscale), and chance (C subscale). The I subscale is scored in the internal direction, whereas the P and C subscales are scored in the external direction. Thus, a high score on the I subscale means higher internality, and a high score on the P and C subscales means higher externality. The I subscale measures the extent to which people believe that they have control over their own lives; the P subscale deals with control by powerful others; and the C subscale is concerned with perceptions of chance control. Each subscale consists of 8 items on a 6-point Likert format ranging from strongly disagree to strongly agree (see Appendix B). To obtain the score of each subscale, the subject's response to each item is added to a constant of 24 in order to eliminate negative values. Thus, the range on each subscale is from 0 -48. A word of caution about interpretation is necessary. High scores on each subscale reflect only a tendency to believe in that particular locus of control, METHOD / 42 with no implication for beliefs about the other dimensions. It is possible then, for a person to obtain a high score on the I subscale yet also obtain low scores on the P or C subscale. Various reliability indices showed that Levenson's IPC Scale is a quite reliable instrument. Split-half reliabilities (Spearman-Brown) was: r = .62, .66 and .64 for the I, P and C subscales respectively. Test-retest reliability for a one-week period were in the .60 - .79 range (Levenson, 1973). Even though the overall internal consistency estimates are only moderately high, Levenson (1981) noted that this is expected since the IPC items were selected from a wide variety of situations. For a student sample (n=152), Kuder-Richardson reliability estimates were .64 for the I subscale, .77 for the P subscale and .78 for the C subscale (Levenson, 1974). Wallston, Wallston and DeVellis (1978) reported similar estimates for their adult sample (n=115). In this study, the alpha reliablilities for the three subscales were: .51, .72, .73 respectively. For a hosipitalized psychiatric sample, Levenson (1973) obtained a Kuder-Richardson's reliability indiex of .67, .82, .79 for the I, P and C subscales respectively. The above evidence suggest that the internal structure of the I subscale is relatively weak than that of the other two subscales. In regard of validity, the P and C subscales were usually correlated significantly with each other (in the range of .41 - .60) and they were usually unrelated to the I subscale (ranging form -.25 to .19). METHOD / 43 3. Assessment of Uncertainty Orientation A composite measure of uncertainty orientation consists of a projective measure of need for uncertainty (Frederick, Sorrentino & Hewitt, 1987), and Byrne and Lamberth's measure of authoritarianism (1971). To develop measures from which to infer uncertainty orientation, Sorrentino and his associates followed procedures similar to those used in the development of measures for assessing achievement motives. In the achievement area, projective test of achievement (McClelland, Atkinson, Clark & Lowell, 1958) was used to infer the motive to succeed, and Self-Report Test Anxiety Scale (Mandler & Sarason, 1952) was used to infer the motive to avoid failure. Similarly, Frederick, Sorrentino and Hewitt (1987) developed a projective measure of need for uncertainty (n Uncertainty) from which to infer need for uncertainty, and used Byrne and Lamberth's measure of authoritarianism (1971) to infer need for certainty. The n Uncertainty test contains four ambiguous sentences. Subjects are required to make up their own story elicited by each of these sentences. The scorer then inteprets the subjects' story in regard of expressions of need for uncertainty. The sentences were presented in the following order: (1) Two people are working in a laboratory on a piece of equipment. (2) A person is sitting, wondering about what may happen. (3) A young person is standing: A vague operation scene is in the background. (4) An older person is talking to a younger person. To obtain higher interrater reliability, the original scoring system has been slightly modified METHOD / 44 in this study (see Frederick, Sorrentino & Hewitt, 1987 for the original scoring system). Stories which contain n uncertainty imagery were scored 2, doubtful stories were scored 1, and stories that show no n uncertainty imagery were scored 0. In addition, those stories which contain n uncertainty imagery were scored for the presence of 5 subcategories (there are 10 subcategories according to the original system), 1 for each. These subcategories are: need, instrumental activity, anticipatory goal, obstacles and nuturant press (see Frederick, Sorrentino & Hewitt, 1987, for detailed scoring instructions). This yielded possible scores ranging from 0 to 7 for each story. The total score of this projective test was obtained by summing up the score of each story. A high score on this measure indicates that the subject has higher need for uncertainty. Byrne and Lamberth's authoritarianism measure (F Scale) was used to infer need for certainty. Based on research findings, Sorrentino and his associates presumed that authoritarianism parallels need for certainty. Past studies indicated that compared with low authoritarians, high authoritarians were found to be more intolerant of ambiguity (Kirscht & Dillehay, 1967), have less experience with novel or uncertainty situations (Kelman & Barclay, 1963), and show lower levels of integrative complexity for uncertain situations (Schroder, Driver & Streufert, 1967). It appears that the phenomena of need for certainty and authoritarianism have some common characteristics. Byrne and Lamberth's measure of authoritarianism consists of 22 items measured on a 6-point Likert format, ranging from strongly disagree to strongly agree. Four original items (i.e., item 16, 17, 21 & 22) were excluded from the present METHOD / 4 5 investigation because some terminologies in these items may have different meanings to Chinese people, such as the term \"religion\". The total score was obtained by summing up the subject's response to each item. The higher score that the subject obtains on this scale, the more authoritarian he/she is. Untilizing two measures to infer the opposite elements of uncertainty orientation is exactly analogous to the formulation of a composite measure of achievement motivation whereby two uncorrelated measures, n Ach and test anxiety measures, independently assess the opposite tendencies. To caculate the composite score for uncertainty orientation, F scores and n Uncertainty scores were first transformed into 2 scores, and then the standardized F scores were multiplied by a constant -1 in order to reverse the direction of scoring. Finally, the composite score was obtained by summing up the standardized F scores and the standardized n Uncertainty scores. The higher composite score that the subject obtains, the more uncertainty-oriented he/she is. 4. Demographic Questionnaire This questionnaire was designed to collect subjects' personal information like gender, age, cultural background, educational level and their parents' socioeconomic status (SES) which all were found to be confounded with the development of LOC and FD-I orientations. It is believed that such demographic variables would also have an influence on the individual's UNCERT. The information collected from this demographic questionnaire would assist to explain the cultural differences, if any, in LOC, FD-I and UNCERT. METHOD / 46 For testing Chinese subjects, all questionnaires were translated into Chinese language by three bilinguals, two of them have no knowledge in psychology and have no idea about the purpose of the present investigation. Differences in translation were finally resolved by the third bilingual. The wordings of some items of the original instruments were slightly modified to produce smooth and natural-sounding statements in the Chinese language. The Chinese version of all questionnaires was then submitted to a pretest with 12 Chinese 11-graders. Minimal changes were required after the pretest. The final copy of these questionnaires is given in Appendix C. C. PROCEDURE In one group testing session, the subjects from both cultural groups were given the questionnaires from which FD-I, LOC and UNCERT were assessed. Hong Kong subjects were tested by one of their teachers in Hong Kong while Canadian subjects were tested by the researcher and her supervisor in Vancouver, Canada. All the subjects were told that they were participating in a study of the relationship between visual perception and different opinions on certain personal and social issues. The Sentence Interpretation Test from which n Uncertainty was assessed were administered first, following standard procedures perscribed for projective measures of motives (Atkinson, 1958, Appendix 3). The subjects were required to generate their own stories on the basis of the corresponding sentence in five minutes. These stories were then scored by two teams of scorers. For the Chinese group, all the stories were scored by two bilingual graduate students and the stories generated by the Canadian students METHOD / 47 were scored by one bilingual student and one Canadian graduate student. Differences in scores were resolved through the discussion between the scorers on each team. According to the standardized procedures (Witkin et. al., 1971, p.27-28), the GEFT was administered next, followed by the \"General Attitude Survey\" which embraced Levenson's IPC Scale and Byrne and Lamberth's measure of authoritarianism. For the \"General Attitude Survey\", the subjects were required to indicate their opinions on certain personal and social issues by selecting a response on a 6-point Likert format, ranging from strongly disagree to strongly agree. No time limit was set for completing this survey; the subjects were asked to finish the survey at their own pace. After that, the subjects were asked to fill out a demographic questionnaire which was designed to collect their personal information like age, gender, SES, cultural background and educational level. Finally, the subjects were given a detailed debriefing message which explained the real purpose of the study. The possible negative feelings that the subjects might have were also concerned. The debriefing message is given in Appendix D. Total testing time was approximately one hour and ten minutes. D. CHAPTER SUMMARY Presented in this chapter were the criteria of selecting the subjects and the descriptions of all the questionnaires used in the study. The procedure of administering the study was also depicted. The following chapter outlines the demographic characteristics of the subjects and their cultural background. The METHOD / 48 cultural differences in the FD-I, LOC and UNCERT would be investigated using 2 the combination of multivariate Hotelling's T and univariate Student's t-test. The relationships among FD-I, LOC and UNCERT would be examined by Pearson correlation, multiple regression analysis and canonical correlation analysis. Finally, such relationships would explore further by computing internal consistency estimates. When interpreting the results, it should be recalled that the I subscale was scored in the internal direction whereas the P and C subscales were scored in the external direction; a high score on F Scale indicates more authoritarian whereas a high score on n Uncertainty measure indicates higher need for uncertainty. The significance level was assigned to be .05 in all the analyses. IV. RESULTS A. DEMOGRAPHIC CHARACTERISTICS OF SUBJECTS Demographic characteristics of the subjects described below are: cultural background, age, gender, place of birth, educational level, and SES. These data were collected from the Demographic Questionnaire (see Appendix B). 1. Cu l tu ra l Background The subject in this study was selected according to cultural background, utilizing three criteria: 1. Minimum 10-year-residence in the respective culture; 2. Cantonese/English as the primary language spoken by the subject and his/her family members at home. 3. Previous education must have been completed in the respective home country/city. Table 1 summarizes these characteristics. Length of Residence. The average length of residence was 17.54 for Hong Kong students and 21.45 for Canadian students; the mean was higher for Canadian students mainly because they are somewhat older than Hong Kong students. The minimum length of residence was 10 years for the Hong Kong group and 13 years for the Canadian group. Language Spoken by The Subjects and Their Family Members. As showed in Table 1, Cantonese was the only language spoken by Hong Kong students and 49 RESULTS / 50 Table 1. Cultural Background of the Subjects Cultural Variables Hong Kong (n = 39) Vancouver (n=41) LENGTH OF RESIDENCE (Years) Mean S.D. Min. Value Max. Value No. of Responses Obtained LANGUAGES SPOKEN BY THE SUBJECTS (n) Cantonese English Speaking Two Different Languages No. of Responses Obtained LANGUAGES SPOKEN BY THE FAMILY MEMBERS (n) Cantonese English Mandarin Other Language Speaking Two Different Lauguages No. of Responses Obtained CULTURE IN WHICH ELEMENTARY SCHOOL WAS RECEIVED (n) Hong Kong Canada No. of Responses Obtained CULTURE IN WHICH SECONDARY EDU. WAS RECEIVED (n) Hong Kong Canada Did Not Completed No. of Responses Obtained 17.54 2.35 10 21 39 39 39 38 1 39 38 38 38 38 21.45 6.58 13 48 40 34 6 40 30 2 8 40 41 41 40 1 41 RESULTS / 51 their family members, except for one student whose family spoke Mandarin. Most Canadian students' families (n=30) spoke English as their primary language. Among Canadian students who were biliguals, five other languages were spoken: Cantonese, Korean, French, Italian or Spanish. It reflects the multicultural nature of Canada. Culture In Which Education Was Received. All the subjects completed their elementary and high school education in their respective \"home\" cultures. 2. Gender, Age and Educat ional Background As showed in Table 2, a total of 80 subjects were tested; 39 (6 males and 32 females) were Chinese whereas 41 (27 males and 12 females) were Canadians. The average age of Hong Kong students was 19.16, approximately three years younger than Canadian students whose average age was 22.03. Age and gender were tested for significance of cultural difference: both variables were found to differ significantly. In regard of place of birth, 32 of 39 Chinese students and 33 of 41 Canadian students were born in their respective home country/city. Among those who were born in other countries, one Chinese student was born in Singapore, three Canadian students were born in Korea, and one Canadian student was born in Philippines and the other one was born in France. With respect to the highest educational level completed, the data indicated that the two groups of students had a similar educational backgrounds. Most (39 Chinese students and 32 Canadian students) were high school graduates. RESULTS / 52 Table 2. Demographic Characteristics of the Subjects Hong Kong Vancouver (n = 39) (n = 41) GENDER a Males 6 27 Females 32 12 No. of Responses Obtained 38 39 A G E b Mean 19.16 22.03 S.D. 1.13 5.98 No. of Responses Obtained 38 38 PLACE OF BIRTH (n) Hong Kong 32 2 China 6 1 Canada 0 33 Other Country 1 5 No. of Responses Obtained 39 41 EDUCATIONAL LE V E L (n) Elementary School Graduation 1 High School Graduation 39 32 College/Technical School Graduation 2 Some University Training 6 No. of Responses Obtained 39 41 a. X 2 ( l ) = 2.03, p<.001 b. t (75) = -3.03, p<.005 3. Socioeconomic Status Table 3 presents the SES of subjects' parents, which was measured by three indices: highest educational level, employment status and the nature of their occupations. The parents of the two cultural groups were found to be significantly different in their highest educational level (Father: t(75) = -5.26, p< RESULTS / 53 Table 3. Socioeconomic Status by Cultural Group Hong Kong Vancouver Father Mother Father Mother SES Indix & Level % % % EDUCATIONAL L E V E L Less Than Elementary School Grad. 36.8 41.0 10.3 5.1 Elementary School Graduation 23.7 20.5 7.7 5.1 Some High School Education 13.2 15.4 12.8 20.5 High School Graduation 15.8 17.9 23.1 25.6 Some College/Technical School Edu. 2.6 2.6 2.6 12.8 College/Technical School Graduation 2.6 2.6 5.1 5.1 Some University Training 5.3 2.6 7.7 University Degree 7.7 10.3 Some Postgraduate Training 5.1 Postgraduate Degree 23.1 7.7 No. of Responses Obtained 38 39 39 39 EMPLOYMENT Self Employed 23.1 7.7 42.5 15.0 Full Time 69.2 23.0 40.0 40.0 Part Time 10.3 20.0 Retired 7.7 10.0 7.5 Student 2.5 Homemaker 59.0 15.0 Other 7.5 No. of Responses Obtained 39 39 40 40 OCCUPATION Homemaker/Retired 8.1 64.9 10.5 26.3 Labourer 5.4 21.6 10.5 7.9 Farm, Farm Manager 2.6 2.6 Craftsman 8.1 2.7 7.9 2.6 Clerical 13.5 2.6 18.4 Proprietor, Owner of a Small Business 5.3 5.3 2.6 Service 13.5 8.1 5.3 7.9 Operative 8.1 10.5 Sales 5.4 5.3 5.3 Technician 2.7 2.6 Manager, Administrator, Govt. Official 8.1 5.3 2.6 Owner of a Larger Business 18.4 2.7 15.8 7.9 Professional I 2.7 2.6 13.2 Professional II 10.5 2.6 Not Classified Elsewhere 2.6 2.6 No. of Responses Obtained 38 37 38 38 RESULTS / 54 .001; Mother: t(76) = -6.14, p<.001) and in their employment status (Father: X 2(5) = 8.407, p<.05; Mother: X 2(5)= 18.249, p<.005). With 36.8% of the Chinese fathers and 41% of the Chinese mothers reported to have less than elementary school graduation, it made the Hong Kong group a relatively less educated sample than the Canadian group where most of the subjects' parents had graduated from high school (69.3% of the Canadian fathers and 69.2% of the Canadian mothers). Although the Chinese fathers were less educated, more of them were employed (91.3% of the Chinese fathers versus 82.5% of the Canadian fathers). On the other hand, the data showed that a large proportion of Canadian mothers (75%) were employed than Chinese mothers (41%). Most of the Chinese mothers (59%) were housewives. This is probably a result of more rigid sex stereotpying in Hong Kong society. The socioeconomic index of occupations used in this study was adapted from Blishen and his associates' socioeconomic index of occupations in Canada (Blishen, Caroll & Moore, 1987). As this index was originally developed for occupations in Canada, it may not be a precise guage of \"natural\" SES occupation variations in Hong Kong society. However, due to the fact that the nature of occupations varies substantially across societies, it would be difficult to derive a common index for comparing the SES of respondents in two societies which differ as much as Hong Kong and Canadian. Blishen's index was employed because it is one of the recognized SES indices of occupations in the literature. The present index was adapted from Blishen's system by J. D. Willms (personal communication, April, 1989). He coded the SES scores reported in Blishen, Caroll and Moore's study (1987) for 510 occupations in Canada, and classified each RESULTS / 55 occupation into 15 categories (see Appendix E). The SES score listed for each catergory is the average SES score for all occupations classified into that category. Then, the categories were ranked according to their average SES scores. To increase representativeness for occupations in Hong Kong society, one more category, 'owner of a larger business', was added to Willms' categories. The results showed that there was no significant difference between the two cultural groups in father's occupation (t(72) = -.71, p>.05) However, mother's occupation did differ significantly (t(72) = -4.36, p<.001). More Canadian mothers than Chinese mothers were employed in high-ranking occupations. (Note that the category of 'Not Classified Elsewhere' was excluded from the present statistical analysis.) B. MAIN ANALYSES 1. Confounding Effects Before turning to the main analyses, an investigation of the effects of confounding variables is presented. The three demographic variables which appeared in the literature as having some influence on the development of FD-I, LOC and UNCERT are gender, age and SES (see Lefcourt, 1982; Witkin & Goodenough, 1981). Occupations of subjects' parents were used in this study to indicate SES. Table 4 presents the correlation of these three confounding variables with the five criterion measures (i.e., the Group Embedded Figure Test (GEFT), Levenson's I, P, and C subscales, and the composite measure of uncertainty orientation, ZCOMP). Only a few correlations were found to be RESULTS / 56 Table 4. Correlation of Gender, Age & ! SES with Criterion Measures GEFT I P C ZCOMP a GENDER*3 Hong Kong .2157 (.097) -.2563 (.060) -.1365 (.207) .2253 (.087) -.0151 (.465) Vancouver .1367 (.203) .4299* - (.003) .0062 (.485) .1209 (.232) -.0662 (.345) AGE Hong Kong -.2178 (.094) -.0587 (.363) .0022 (.495) -.1474 (.189) -.1538 (.185) Vancouver -.2205 (.092) .1645 (.162) -.2758* (.047) -.3185* (.026) .44461 (.003) MOTHER'S OCCUPATION Hong Kong .0640 (.353) -.0577 (.367) -.0756 (.328) .0682 (.344) -.0742 (.336) Vancouver .3062* (.033) -.1789 (.145) .1107 (.257) .1091 (.260) -.1894 (.131) FATHER'S OCCUPATION Hong Kong .0524 (.379) .2528 (.066) .0370 (.414) .0977 (.283) -.0115 (.474) Vancouver .1592 (.173) -.2839\" (.044) .2931* (.039) .0544 (.375) -.0426 (.401) a. ZCOMP: the composite score for UNCERT b. Coding for gender: M=l, F=2 significance levels are presented in parentheses * significant atc\u00C2\u00A3=.05 significant at the .05 level using a one-tailed test. Gender was not significantly correlated with the criterion measures, except for one coefficient which attained statistical significance. Inconsistent with the past findings, which showed that RESULTS / 57 females tend to be more external, Vancouver female students demonstrated high internal scores. Age was negatively associated with most of the criterion measures in both groups; three of these coefficients were significant. Among the correlations with occupation, only three in the Canadian group were significant. The magnitude of these few significant correlations, ranging from .4446 to -.3185, were so moderately low as to lack practical significance. This suggests that the influence of the three confounding variables on the criterion measures is so trivial that statistical control for them is not necessary in the subsequent analyses. The other feature of examining confounding effects on the criterion measures is comparing the \"confounding correlations\" between the two cultural groups. Only a few \"confounding correlations\" appear to be significant. They are: Gender-I, Father's Occupattion-I and Age-ZCOMP correlations. Such a few potential significant correlations seem to have little influence on the criterion measures. Future studies should explore further confounding effects between different cultural groups. 2. Cul tu ra l Differences i n F ie ld Dependence-Independence, locus of control and uncertainty-orientation As discussed in the literature review, Hong Kong students are expected to be more FD, external LOC and certainty-oriented than their Canadian counterparts. A summary of results for group differences of FD-I, LOC and UNCERT is displayed in Table 5. Since multiple criterion measures were involved, Hotelling's 2 T test was first performed to test for differences between the two cultural groups on the composite set of measures, using the centroids of the criterion RESULTS / 58 measures as the basis of comparison. Then, one-tailed Student's t tests were computed to evaluate group differences on the individual criterion measures. This procedure follows an approach suggested by Hummel and Sligo (1970). In their study, they compared three analysis of variance approaches analyzing group differences in multivariate normal data. These approaches are: completely 2 univariate Student's t-test, completely multivariate Hotelling's T approach, and 2 the combination of univariate Student's t-test and multivariate Hotelling's T procedure. The researchers recommended that the combination approach is more desirable to analyze the multivariate data than the other approaches. The use of the univariate approach with multivariate data should be discouraged due to its inflated Type I error. The completely multivariate procedure is also not a desirable approach because of its extreme conservativeness which makes it difficult to observe possible group differences on individual criterion measures. As expected, the two cultural groups were found to differ significantly on the 2 centroid of the criterion measures (T (5, 74) = 16.39, p<.05). Univariate t-tests showed that it was clearly significant group differences on the P and C subscales (t(78) = 2.88, p<.005; t(78) = 3.33, p<.005). Hong Kong students were found to be more external in the domains of 'Powerful Others' and 'Chance'. Even though no significant difference was reported on the composite measure of uncertainty orientation, closer examination of the component scores for uncertaintj7 orientation revealed that the two cultural groups were significantly different on both authoritarianism and n Uncertainty measures (t(78) = 4.42, p<.001; t(78) = 2.14, p<.05). Hong Kong students were more authoritarian than Canadian students. However, the group difference for n Uncertainty was in the opposite RESULTS / 59 Table 5. Means, Standard Deviation and Significance Tests of Cultural Group Differences Hong Kong Vancouver Hotelling T Mean S.D. Mean S.D. Lower C.I. Upper C.I. t a GEFT 13.51 35.80 26.49 27.18 -0.20 4.30 3.68 5.29 4.33 1.23 12.63 35.52 22.93 23.20 0.21 5.22 4.52 5.74 6.15 1.50 4.64 3.52 7.90 8.18 0.67 2.89 2.96 0.78 0.21 1.46 0.82 0.23,. 2.88* 3.33 -1.31 I P C ZCOMP a. Chinese students was assigned to be group 1; Canadian students was assigned to be group 2 Degree of Freedom for the t-tests =78 * significant atc\u00C2\u00A3\u00C2\u00BB=.05 direction. That is, Hong Kong students, unexpectedly, were found to have higher need for uncertainty. 3. Correlations among F ie ld Dependence-Independence, locus of control and uncertainty-orientation To examine whether FD-I, LOC and UNCERT were inter-related, Pearson correlations were computed among them. The summary of these results is presented in Table 6, where correlation coefficients for Hong Kong students are reported in the upper triangle of the matrix and those for Canadian students are in the lower triangle. Although there were some significant correlations, their typical magnitude is not so high as to challenge seriously the proposition of relative independence among the constructs. Given that only two significant RESULTS / 60 Table 6. Intercorrelation Among FD-I, LOC & UNCERT GEFT I P C F N U N a GEFT 1.000 .147 .062 -.008 -.268 -.008 (.187) (.355) (.481) (.050) (.482) I -.385'' 1.000 .338* -.059 .163 -.286* (.006) (.018) (.361) (.161) (.043) P .088 -.398* 1.000 .158 .051 -.002 (.293) (.005) (.168) (.380) (.415) C -.086 -.133 .445* 1.000 .316* -.122 (.297) (.203) (.002) (.025) (.236) F -.133 .366'' .173 .188 1.000 -.177 (.204) (.009) (.139) (.119) (.148) NUN .030 .060 -.301* -.324* -.011 1.000 (.427) (.354) (.028) (.019) (.473) a. NUN: the n Uncertainty measure significance levels are presented in parentheses * significant atcL=.05 correlations involving GEFT were found and none of them explained more than 14.8% of shared variance, it seems that FD-I is unrelated to LOC and UNCERT. This proposition was tested further by conducting standard multiple regression analyses on GEFT using the Levenson's subscales and the component measures for UNCERT as two sets of predictor variables. Results of standard multiple regression analyses are summarized in Table 7. Neither LOC nor UNCERT was showed to be a good predictor of FD-I. Considering the relationship between LOC and UNCERT, weak and irregular RESULTS / 61 Table 7. Summary of Multiple Regression Analyses on GEFT Predictors R R 2 MS MS F-Ratio p reg res Hong Kong C, I, P 0.15 0.02 5.07 19.62 0.26 .850 NUN, F 0.27 0.07 18.52 13.79 1.34 .275 Vancouver C, I, P 0.41 0.17 60.91 24.56 2.48 .076 NUN, F 0.14 0.02 -0.03 5.31 0.36 .702 associations were found, as reported in Table 6. Significant correlations included: positive C-F correlation (10% of common variance) and negative I-NUN correlation (8.2% of common variance) for Hong Kong students; positive I-F correlation (13.4% of common variance), negative P-NUN correlation (9% of common variance) and negative C-NUN correlation (10.5% of common variance) for Canadian students. It appears that the two cultural groups have different patterns of correlation. Since the I-E control and UNCERT scores collected from the subjects were derived from Levenson's three subscales and two component measures for UNCERT, canonical analysis would provide a more pertinent statistical analysis for the pattern of relationship under study. In addition, canonical analysis provides a method of determining the relative importance of the separate subtests within each instrument. Canonical analysis was performed using the BMDP6M program (Dixon, 1983), with Levenson's I, P, and C subscales as the first set of variables and the UNCERT component measures as the second set of variables. RESULTS / 62 Since one of the requirements for canonical analysis is that none of the variables in each set should be too highly correlated with, or near linear combinations of, others in the set, an investigation of the correlation between/among the subtests in each set is necessary. Intercorrelations between/among the subtests are displayed in Table 6. Consistent with the conceptual definition of the UNCERT construct, which was assessed by two uncorrelated measures (i.e., the F Scale and the n Uncertainty measure), the magnitude of F-NUN correlation was found to be very low and insignificant in both groups. Considering the association among I, P and C subscales, all the significant correlations were consistently weak; the highest correlation explained only 19.8% of variance. Again, the two cultural groups seem to have different patterns of realtionship among I, P and C. For Canadian students, although the correlations were consistently weak, the magnitude of P-C correlation was larger than the other correlations, thereby suggesting that P and C may have more overlapping variance than either P and I or C and I. This was expected since both P and C subscales were designed to measure different aspects of externality. A different pattern of association was found for Hong Kong students. Only one significant correlation, P-I correlation, was noted and it was unexpectedly positive in direction. The remaining correlations were insignificant. The correlation matrix indicated that relationship between/among subtests in each set were low enough to allow the canonical analysis to proceed. (Different patterns of relationship among I, P and C subscales are discussed further in the \"Additional Analyses\" section.) Two canonical correlations were generated and evaluated for significance in each group. For Hong Kong students, the first canonical correlation was .45 (20% of RESULTS /. 63 2 variance); the second was .14 (2% of variance). Neither was significant (X (6) = 8, 2 p>.2; X (2) = .65, p>.7). For Canadian students, the first canonical correlation was .55 (30% of variance); the second was .28 (8.2% of variance). ' Only the 2 first canonical correlation was significant (X (6) = 16.46, p<.02) and useful for interpretation of the relationship between the two sets of variables. The second 2 canonical correlation did not attain statistical significance (X (2) = 3.22, p>.l). Therefore, a significant relationship between the two sets of variables explained by the first canonical correlation was found in the Canadian group, but not in the Hong Kong group. The following discussion, therefore, focuses mainly on an interpretation for Canadian students. Analysis relevant to the first (and only significant) canonical correlation for Canadian students is summarized in Table 8, which shows correlations (loadings) between the variables and the respective canonical variates, squared multiple correlations of each variable in the set with all variables in the other set (i.e., 2 multiple R ) and canonical correlations. Correlations between variables and variates in excess of .3 are eligible for interpratation in the present study. As seen in the table, all variable-variate correlations exceeded .3 level. It is evident that those Canadian subjects who were internal in their personal life (.48) and external in the domains of 'Powerful Others' (.54) and 'Chance' (.59) also tended to be more authoritarian (.87) and have low need for uncertainty (-.50). In regard of the relative importance of each variable in the canonical relationship, the F Scale, with a variable-variate correlation of .87 and a 2 2 significant multiple R (R =.25, p<.05), contributed more strongly than the n RESULTS / 64 Table 8. Summar3' of Canonical Correlation Analysis Variables Within-Set Variable-Canonical Correlation Standardized Coefficient Multiple R LOC SET ** I 0.48 0.82 0.13\u00E2\u0080\u009E P 0.54 0.68 0.12*:s C 0.59 0.41 0.14 UNCERT SET * F 0.87 0.87 0.25 NUN . -.50 -.49 0.14 Canonical R^ 0.55 Canonical R 0.30 p<.05 p<.l Uncertainty measure to the canonical equation. Among the LOC variables, I, P and C subscales made approximately equal contribution to the equation. C. ADDITIONAL ANALYSES Different patterns of inter-correlation of I, P and C subscales were found between the two cultural groups. As mentioned before, Levenson's IPC Scale was developed to tap the I-E control in three different dimensions. The I subscale is scored in the internal direction whereas the P and C subscales are scored in the external direction. Therefore, the I subscale should relate negatively to the P and C subscales, while the P and C subscales should associate positively with each RESULTS / 65 other. This general pattern of correlation among I, P and C was found for Canadian students; however, a different pattern of correlation was observed for Hong Kong students. For the Hong Kong group, the P and I subscales were unexpectedly related in a positive direction while the other two correlations were very low and insignificant. This pattern suggested that the three subscales, which were originally designed for the North American population, might not assess the I-E construct in the same way for Hong Kong students. To investigate further the functioning of the IPC Scale, internal consistency was computed using the LERTAP program (Nelson, 1974). Internal consistency is an estimate of the extent to which the subjects' responses are consistent across the items within a test (or subtest). When the responses are consistent, items are interpreted as measuring the same construct, and the test (or subtest) is said to have strong internal consistency. The LERTAP program also provides an internal consistency estimate for the composite test. In order to compute the internal consistency estimates for the composite test, all the subtests should be scored in the same direction. Thus, the scoring direction of the P and C subscales was reversed for the present analysis. Hoyt's coefficient and Cronbach's alpha are used to estimate internal consistency in the LERTAP program. A summary of internal consistency estimates of the IPC Scale is displayed in Table 9. For Canadian subjects, given that the IPC items were selected from a wide variety of situations, the data showed that except for the I subscale, the overall internal consistencies of Levenson's subscales were moderately high. Cronbach's alpha reliability of .60 further suggested that Levenson's Scale is a quite reliable RESULTS / 66. Table 9. Summary of Item Statistics by I, P and C subscales Subscales Mean S.D. Hoyt Cronbach HONG KONG I P C 35.79 29.51 28.85 3.68 5.29 4.37 .48 .69 .53 The Composite Test 94.5 7.54 .51 -.10 VANCOUVER I P C 35.51 33.07 32.80 4.52 5.74 6.15 .51 .63 .69 The Composite Test 101.39 12.34 .76 .60 composite instrument for Canadian subjects. On the other hand, the subscale Hoyt's coefficients of .48 and .53 indicated that the items of the I and C subscales were not homogeneous for Hong Kong subjects. In other words, besides assessing the I-E construct, these items might also assess some other constructs or traits. This finding at least partially explained why for the Hong Kong group, low and unexpected correlations were found among the I, P and C subscales. One effect of having heterogeneous I and C items was that the Hoyt's coefficient for the composite test was low even though internal consistency of the P subscale remained moderately high for Hong Kong subjects. This, accompanied by the Cronbach's alpha of -.10, indicated that Levenson's Scale is an unreliable instrument for Hong Kong subjects. RESULTS / 67 D. CHAPTER SUMMARY This chapter described the demographic characteristics of the subjects, including cultural background, age, gender, place of birth, educational level and their parents' SES. Potential confounding effects of age, gender and SES were also investigated. Pearson correlations showed that the influence of these confounding 2 variables was trivial. A combination of multivariate Hotelling's T and univariate Student's t test was employed to examine cultural differences in FD-I, LOC and UNCERT. Significant differences were found in two of the Levenson's subscales (i.e., the P and C subscales) and in the two separate component measures of UNCERT. However, the direction of the group difference for the n Uncertainty measure opposite to that predicted. Interrelationships among FD-I, LOC and UNCERT were analyzed using standard multiple regresssion and canonical analyses. The results showed FD-I was unrelated to LOC and UNCERT, for both cultural groups. With respect to the relationship between LOC and UNCERT, a moderate association (30% of overlapping variance) was found between LOC and UNCERT for the Canadian sample, but not for the Hong Kong group. Finally, subscale internal consistencies were evaluated for the I, P and C subscales. Different patterns of estimates emerged between the two cultural groups: for Canadian subjects, internal consistencies were moderately high, except RESULTS / 68 for the I subscale. For Hong Kong subjects, results showed that the I and C items might assess more than one construct. V. DISCUSSION The primary objective of the present study was to investigate cultural differences in FD-I, LOC and UNCERT, and the differences in the inter-relationships among these cognitive constructs. The results partially supported the propositions of lb and lc: the cultural differences in LOC and UNCERT. Proposition la, in which cultural differences in FD-I was speculated, was hot confirmed. Results concerning the interrelationships among these cognitive domains suggested a moderate association between LOC and UNCERT for Western subjects, but not for Chinese subjects. No other significant correlations among these constructs was obtained in this study. The confounding effects of age, gender and SES were also assessed; their influence was trivial and deemed not to have contaminated the current results. A. CULTURAL DIFFERENCES IN FIELD DEPENDENCE-INDEPENDENCE, LOCUS OF CONTROL & UNCERTAINTY-ORIENTATION As discussed in the literature review, it was speculated that Hong Kong students would be more FD than their Canadian counterparts. Result of this study did not support this expectation; no cultural difference was found for FD-I. Students from both cultural groups obtained similar score levels on the GEFT. One probable explanation of this result is that the GEFT is not a good instrument for differentiating college students in the FD-I domain. The test may be too simple for college students who have little difficulty in finishing the whole test within the time limit. To demonstrate whether the current result is due to the 69 DISCUSSION / 70 inappropriate use of the GEFT for the present samples, more extensive adult norms are needed for the GEFT. A strong cultural difference was reported on the two separate component measures of UNCERT, but not on the composite score. Consistent with the theoretical expectation, and with Lau's (1985) earlier findings, Hong Kong students scored higher on the F Scale. In other words, Hong Kong students were more authoritarian than Canadian students. On the other hand, in conflict with Lau's finding, Hong Kong students scored higher on the n Uncertainty measure, displaying a higher need for uncertainty. The composite score showed no cultural difference. The question of why inconsistent findings were obtained in two similar studies remains unanswered. More empirical work should be done to explore the cultural dynamics of UNCERT. Pitfalls of projective techniques may at least partially explain the inconsistent findings. One limitation of projective techniques is the lack of objectivity in scoring. The n Uncertainty measure, like other projective tests, depends extensively on the scorers to evaluate and integrate the meaning of the responses as signs of the traits being studied \u00E2\u0080\u0094 need for uncertainty, in this instance. Different scorers' interpretations of the responses may not be comparable. In the present case, the n Uncertainty measure was scored by two teams of scorers. The first team, consisting of two bilingual graduate students, scored the stories generated by the Chinese students; the second team, composed of one bilingual graduate student and one Canadian graduate student, scored the Canadian students' stories. Even though differences in interpretation of the DISCUSSION / 71 responses were resolved through discussion between scorers on each team, variations in interpretation may still have occured between two teams of scorers. Therefore, variation in interpretation of the responses, not just the cultural influence per se, may explain the current results. The proposition of a cultural difference in LOC was partially supported in this study. Statistical differences were observed on the P and C subscales, but not on the I subscale. As expected, Hong Kong students obtained higher scores on the P and C subscales than their Canadian counterparts. Before a firm conclusion can be drawn to state that Hong Kong students are more external in the domains of 'Powerful Other' and 'Chance', the comparability of Levenson's IPC Scale for the two cultural groups of subjects should be established. Careful translation of the English IPC Scale into Chinese does not necessarily guarantee that all items have the same psychological meaning for the two groups of subjects. Because of different cultural values and perceptions, it is possible that some of the IPC items may not have the same meaning for Chinese subjects as it does for those from North American cultures. In this study, equivalency in operationalization was tested by comparing the internal structure of the IPC Scale for the two cultural groups of subjects. Internal consistency estimates suggests that the internal structure of the IPC Scale is not congruent for the two culture groups. The IPC Scale, originally developed in North American culture, appears to function more effectively for Canadian subjects than for Hong Kong subjects. Given that the IPC items were selected from a wide varietj' of situations, the overall internal consistency of the DISCUSSION / 72 IPC Scale is moderately high for Canadian subjects. As the previous findings showed (see Chapter III), the internal structure of the I subscale is relatively weaker than that of the other two subscales. Results showed that the IPC items, especially the P and C items, are homogeneous for Canadian subjects; they all were designed to assess the I-E construct in their specific sub-domains. On the other hand, when the scale was applied to Hong Kong subjects, the data showed that the P subscale is the only subscale which seems to have similar psychological meaning for the two cultural groups of subjects. For the I subscale, its internal structure is still weak. Accompanied by the past evidence, low internal cosistency of the I subscale for both cultural groups clearly suggests that this subscale needs to be revised in order to improve the quality of the IPC Scale. Internal consistency analysis of the C subscale indicated that, for Hong Kong subjects, other constructs or traits which are not operationally defined as a part of the LOC construct may contaminate the 'Chance' domain. This finding raises questions regarding the construct validity of the C subscale as an indicator of the I-E control for Chinese subjects. Different internal consistency estimates of the C subscale were obtained for the two cultural groups. It implies that subjects from the two cultures may perceive the C items differently, and the C items seems to be more appropriate for Canadian subjects than for Hong Kong subjects. Therefore, to measure the I-E control more accurate in the context of Chinese culture, it would be advantageous to examine the Chinese concept of chance. Further work in this direction would be useful to develop a more valid LOC instrument for the Chinese population. DISCUSSION / 73 Until there is more empirical work investigating what the IPC Scale really measures in Chinese cultures, extreme caution should be taken in interpreting cross-cultural comparisons with the IPC Scale, including the present study. B. RELATIONSHIPS AMONG FIELD DEPENDENCE-INDEPENDENCE, LOCUS OF CONTROL & UNCERTAINTY ORIENTATION Consistent with past findings (see Chapter I), no association betweeen FD-I and LOC was reported in this study, among either Canadian subjects or Hong Kong subjects. It seems that FD-I and LOC are two separate psychological constructs. However, the methodological problems inherent in the instruments make it difficult to interpret the obtained association unambiguously, especially for the Hong Kong sample. Further research should be done on the validation of instruments. An moderate association were found between LOC and UNCERT for Canadian subjects, but not for Hong Kong subjects. This study provides tentative evidence that UNCERT and LOC are two related constructs, at least for Canadian subjects. The lack of correlation for Hong Kong subjects may be due to the invalidity of the IPC Scale in Chinese culture. The correlation might have been much higher, if the IPC Scale had been more valid and more reliable for Hong Kong subjects. (The internal consistency estimates and the Cronbach's alpha for the scale were low in the Hong Kong group). The relationship between UNCERT and FD-I was also examined in this study. Interestingly, no association between UNCERT and FD-I was obtained, for either Canadian or Hong Kong subjects. At this stage, no sufficient evidence has been obtained to allow clear interpretation DISCUSSION / 74 of these associations. Independent replications of the current results are encouraged. C. IMPLICATIONS FOR CROSS-CULTURAL RESEARCH STUDIES The main implications for further cross-cultural research arise largely out of the limitations in the present study. A shortcoming of the present study is the use of Hong Kong Chinese as a group of Chinese subjects. Being a highly industrialized British colony for a long time, Hong Kong is a melting pot of Western and Eastern cultures. Many of its values, social norms and expectations are a blend of both Chinese and Western traditions. Therefore, it is reasonable to expect that Hong Kong Chinese may have different social values and perceptions as compared with the Chinese in the other societies, like Mainland China and Taiwan. For a better understanding of the dynamics and the effects of Chinese culture, it would be highly desirable to extend the present study to more 'typical' Chinese societies and to investigate how Chinese in different social and cultural contexts function in cognitive domains. Perhaps Chinese poeple who settle into differing Western sub-cultures adapt by developing differential patterns of cognitive abilities, much as they have been showed to develop differential socio-economic patterns of adaptation in Jamaica and Guyana (Patterson, 1975). Equally important is the question of the representativeness of the college samples. In making cross-cultural comparisons, the samples upon which the comparisons DISCUSSION / 75 are based must be comparable, such as in educational level, age, and family background, so that the observed cognitive differences cannot be attributed simply to these confounding variables. College/University students become a natural choice not only because of their availability, but also because of their comparability (many potential confounding variables are naturally matched by using college/university students in different cultures). However, the use of college students always poses another question: the external validity of the findings. A sample of college/university students in many countries usually represents a group of middle-class people who are relatively young and well-educated. Certainly, this group of people does not represent the general population, which is composed of persons from all walks of life. As a result, findings obtained from college students do not readily generalize to others. To increase the generalizibility of the current results, further research should be performed by using different samples of subjects. The lack of equivalency of measuring instruments used in different cultures is one of the major methodological problems in cross-cultural research. The present study further highlights the importance of evaluating cross-cultural construct validity for measuring instruments. In most previous cross-cultural studies, researchers have neglected the importance of cross-cultural construct validity. They have assumed that once the instruments have been accurately translated, a direct cross-cultural comparison of test scores can be made. By estimating internal consistency, the current results indicated that translation does not guarantee that the instruments have the same psychological meaning to subjects from different cultural backgrounds. A psychological test which lacks cross-cultural construct DISCUSSION / 76 validity is in effect two separate measures, one for each culture being studied. A direct cross-cultural comparison of these test scores could therefore be quite likely misleading. Only by establishing cross-cultural construct validity, we can confidently make meaningful comparisons across cultures. The other area that is in need of empirical attention is the interscorer reliability for the n Uncertainty measure. As mentioned before, the interpretation of the stories on the n Uncertainty measure largely depends on scorers' subjective judgements. The variation in interpretation may lead to a great difference in test scores. Because of this, the obtained findings on UNCERT may be attributed to the low interscorer reliability. In order to interpret results clearly, interscorer reliability must be demonstrated in future studies. D. CONCLUSION This study was exploratorj' and speculative in nature. Independent replications are needed before any firm conclusion can be drawn from the current results. The discussion in this chapter is some suggestive directions for future research. Despite its limitations, the present study represents the first step of a potentially exciting line of research. Further cross-cultural studies in this area should be encouraged, especially in such multicultural countries as Canada and probably Australia. Such studies could be most relevant and valuable for fostering intercultural communciation, understanding and acceptance. R E F E R E N C E S Adams-Webber, J.B. (1969). Generalized expectancies concerning locus of control reinforcements and the perception of moral sanctions. British Journal of Social & Clinical Psychology, _8, 340-343. Alker, H.A. & Poppen, P.J. (1973). Personality and ideology in university students. Journal of Persoality, 11, 653-671. Amir, Y. (1972) International and intro-ethnic comparisons of intellectual functions in Israeli and middle Eastern populations. Abstract Guide of the 20th International Congress of Psychology, 175-176. Arbuthnot, J. B. (1971). Field independence & maturity of moral judgement, critical distinctive feature analysis, & perceived locus of control. Unpublished doctoral dissertation, Cornell University. Atkinson, J.W. (Ed.) (1958). Motives in Fantasy, Action, and Society. Princeton, N.J.: Van Nostrand. Bachrach, R., Huesmann, L.R., & Peterson, R.A. (1977). The relation between locus of control and the development of moral judgement. Child Development, 48, 1340-1353. Baron, S. (1971). Devleopment and validation of a TAT type projective test for use among Bantu-speaking people. South Africa: National Institute for Personnel Research. Bax, J.C. (1966). Internal-external control and field dependence. Unpublished honor's thesis, University of Waterloo. Berry, J.W. (1966). Temne and Eskimo perceptual skills. International Journal of Psychology, _1, 207-229 Birnbaum, A.S.G. (1975) Social correlates from field articulation in adolescents: The effect of upon productivity in the presence of others. Dissertation Abstracts International, 36, 1959-1960. Blishen, B.R., Caroll, W.K. & Moore, C. (1987). The 1987 socioeconomic index for occupations in Canada. Canadian Review of Sociology & Anthropologhy, 24, 465-488. Bloomberg, M. & Soneson, S. (1976). The effects of locus of control and field independence-dependence on moral reasoning, Journal of Genetic Pshchology, 128, 59-66. Bond, M.H. & Tornatsky, L.G. (1973). Locus of control in students from Japan and the United States: Dimensions and levels of response. Psychologia, 16, 209-213. 77 / 78 Byrne, D., & Lamberth, J. (1971). The effect of erotic stimuli on sex arousal, evaluative response, and subsequent behavior. In Technical reports of the Commission on Obscenity and Pornography (vol. 8, pp. 41-48). Washington, DC: U. S. Government Printing Office Caring, L. (1971). The relation of cognitive styles, sex, and intelligence to moral judgement in children. Dissertation Abstract International, 31, 7568. Chance, J., & Goldstein, A. (1967). Locus of control and performance on embedded figures. Paper presented at the meeting of the Midwestern Psychological Association, Chicago. Chan, C. (1981). A study of acculturative stress and internal-external control on Hong Kong foriegn students at the University of Waterloo. Unpublished honors thesis, University of Waterloo. Cherlin, A., & Bourque, L.B. (1974). Dimensionality and reliability of Rotters I-E Scale. Sociometry, 37, 565-582. Chiu, L.H. (1986). Locus of control in intellectual situations in American and Chinese school children. International Journal of Psychology, 21, 167-176. Chong, L.E., Cragin, J.P., & Scherling, S.A. (1983). Manager work-related values in a Chinese corporation. Paper presented to the annual meeting of the Academy of International Business. San Francisco, U.S.A., April. Chu, G.C. (1966). Culture, personality & persuasibility. Sociometry, 29, 167-174. Cook, H., & Chi, C. (1984). Cooperative behavior and locus of control among American and Chinese-American boys. Journal of Psychology, 118, 169-177. Crandall, V.C., Katkovsky, W. & Crandall, V.J. (1965). Children's belief in their control of reinforcement in intellectual academic achievement behaviors. Child Development, 36, 91-109. Dawson, J.L.M. (1970) Psycholgical research in Hong Kong. International Journal of Psychology, _5, 63-70. Dawson, J.L.M., Young, B.M., & Choi, P.P.C. (1974). Developmental influence in pictorial depth perception among Hong Kong Chinese children. Journal of Cross Cultural Psychology, _5, 322. Deever, S. (1967) Ratings of task-oriented expectancy for success as a funciton of internal control and field independence. Doctoral dissertation, Unversity of Florida, Ann Arbor, Mich.: University Microfilms, No. 68-9470. Dixon, W.I. (Ed.) (1983). BMDP Statistical Software. Berkeley: University of California Press. / 79 Domino, G. & Mo, T.H. (1987). A comparative analysis of social values of Chinese and American children. Journal of Cross Cultural Psychology, 18, 58-77. Dyal, J.A. (1984). Cross cultural research with locus of control construct. In H.M. Lefcourt (Ed.) Research with locus of Control Construct(pp. 209-306). Vol. 3, New York: Academic Press. Edwards, A.L. (1954). Manual of Edwards Personal Preference Schedule. N.Y.: The Psychological Cooperation. Feather, N.T. (1967). Some personality correlates of external control. Australian Journal of Psychology, 19, 253-260. Fineley, G.E., Solla, J. & Cowan, P.A. (1977). Field dependence, egocentrism, and conservation in young children, Journal of Genetic Psychology, 131, 155-156. Frederick, J.E., Sorrentino, R.M. & Hewitt, E.C. (1987). Need for Uncertainty Manual: Revised Edition. Research Bulletin #618. University of Western Ontario, Canada. Fung, Y.L. (1948). A Short History of Chinese Philosophy. N.Y.: MacMillan. Gurin, P., Gurin, G., Lao, R.C, & Beattie, M. (1969). Internal-external control in the motivational dynamics of Negro youth. Journal of Social Issues, 25, 29-53. Guthrie, K.H. (1985). Locus of control and field independence-dependence as factors in the development of moral judgement. Journal of Genetic Psychology, 146, 13-18 Guttman, J., Bar-Zohar, Y., & Statter, K. (1981). Locus of control and moral judgement: A cross cultural study in Israel. Journal of Moral Education, 10, 186-191. Hofstede, G. (1980). Cultural Consequences: International Differneces in Work-Related Values. London & Beverly Hills: Sage. Holley, M. (1972) Field dependence, sophistication-of-body-concept, and social distance selection. Dissertation Abstracts International, 33, 296. Hsieh, T.Y.T., Skybut, J., & Lotsof, E.J. (1969). Internal vs. external control and ethnic membership: A cross cultural comparison. Journal of Counselling & Clinical Psychology, 33, 122-134. Hsu, F.L.K. (1970). American and Chinese. Garden City, N.Y.: National History. Hsu, F.L.K. (1981). American and Chinese: Passages to Difference. Honolulu: University of Hawii Press. / 80 Huang, L.C., & Harris, M.B. (1973). Conformity in Chinese and Americans, a field experiment. Journal of Cross-Cultural Psychology, _4, 427-434. Hummel, T.J. & Sligo, J.R. (1971). Empirical comparison of univariate & multivariate analysis of variance procedures. Psychological Bulletin, 76, 49-57. Janis, I.L., & Field, P. (1959). Sex difference and persoanality factors related to persuasibility. In C.I., Hovland & I.L., Janis (Eds.), Personality & Persuasibility (pp. 55-68). New Haven: Yale University Press. Janzen, H.L. & Boersma, F.L. (1976). Locus of control and its relationship to moral development. Alberta Journal of Educational Resources, 22, 237-274. Joe, V.C., & Jahn, J.C. (1973). Factor structure of the Rotter IE scale. Journal of Clinical Psychology, 29, 66-68. Kelman, K.C. & Barclay, J. (1963). The F scale as a measure of breadth and perspective. Journal of Abnormal & Social Psychology , 67, 608-615. Kirscht, J. & Dillehay, R. (1967). Dimensions of authoritarianism. Lexington: University of Kentucky Press. Kohlberg, L. (1976). Moral stages and moralization: The cognitive developmental approach. In T. Lickona (Ed.), Moral development & behavior: Theory, research & social issues. New York: Holt, Rinehart & Winston. Lai, P.Y. (1981). Personality in Chinese and English secondary students. Unpublished master's thesis. University of London. Lao, R.C. (1977). Levenson's IPC scale: A comparison of Chinese and American students. Journal of Cross Cultural Psychology, _9, 113-122. Lau, B. (1985). Preliminary test: Cultural effects on learned helplessness as a funciton of uncertainty orientation. Unpublished honors thesis, University of Western Ontario. Lefcourt, H.M. (1976). Locus of control: Current trends in theory & research. Hillsdale, N.J.: Lawrence Erlbaum Associates. Lefcourt, H.M. (1979). Locus of control for specific goals. In L.C. Perlmutter, & R.A. Monty (Eds.) Choice and Perceived Control. Hillsdale, N.J.: Lawrence Erlbaum Associates. Lefcourt, H.M. (1982). Locus of Control: Current Trends in Theor}' and Research. (2nd ed.). N.J.: Lawrence Erlbaum Associates. Lefcourt, H.M., & Siegel, J.M. (1970). Reaction time performance as a funciton of field dependence and autonomy in test administration. Journal of Abnormal Psychology, 76, 475-481. / 81 Lefcourt, H.M. & Telegdi, M.S. (1971). Perceived locus of control and field dependence as predictors of cognitive activity. Journal of Consulting & Clinical Psychology, 37, 53-56. Levenson, H. (1972). Distinctions within the concept of internal-external control: Development of a new scale. Proceedings of the 80th Annual Convention of the American Psychological Association, 261-262. Levenson, H. (1973). Multidimensional locus of control in psychiatric patients. Journal of Consulting & Clinical Psychology, 41, 397-404. Levenson, H. (1974). Activism and powerful others: Distinctions within the concept of internal-external control. Journal of Personality Assessment, 38, 377-383. Levenson, H. & Miller, J. (1976). Multidimensional locus of control in sociopolitical activity of conservative and liberal idiologies. Journal of Personality & Social Psychology, 33, 199-208. Levenson, H. (1981). Differentiating among internality, powerful others, and chance. In H.M., Lefcourt (Ed.) Research with the Locus of Control Construct, Vol. I. N.Y.: Academic Press, Inc. Li, M.C., Cheung, S.F. & Kau, S.M. (1979). Competitive and cooperative behavior of Chinese children in Taiwan and Hong Kong. Acta Psychologica Taiwanica, 21, 27-33. Madsen, M.C. (1971). Developmental and cross-cultural differences in the cooperative and competitive behavior of young children. Journal of Cross-Cultural Psychology, 2, 365-371. Mandler, G., & Sarason, S.B. (1952). A study of anxiety and learning. Journal of Abnormal & Social Psychology, 47, 166-173. Mas, CA. (1984). Locus of control and attained level of moral reasoning as factors affecting the moral reasoning level of female high school students. Dissertation Abstracts International, 45, 2454. McClelland, D.C., Atkinson, J.W., Clark, R.A., & Lowell, E.L. (1958). A scoring manual for the achievement motive. In J.W. Atkinson (Ed.), Motives in Fantasy, Action & Society, (pp.179-205). Princeton, N.Y.: Van Nostand. Mclntire, W. G. & Dreyer, A.S. (1973). Relationship of cognitive style to locus of control. Perceptual & Motor Skill, 37, 553-554. Meade, R.D., & Whittaker, J.D. (1967). A cross-cultural study of authoritarianism. Journal of Social Psychology, 72, 3-7. Messick, S., & French, J.W. (1975). Dimensions of cognitive closure. Multivariate Behavorial Research, 10, 13-16. / 82 Mirels, H.L. (1970). Dimensions and internal versus external control. Journal of Consulting & Clinical Psychology, 34, 220-228. Needham, J. & Ling, W. (1956). Science and Civilization in China: History of Scientific Thought. Cambridge University Press. Nelson, L.R. (1974). Guide to LERTAP: Use & Interpretation. Dunedin, N.I.: University of Otago Press. Okonji, M.O. (1969). Differential effects of rural and urban upbringing on the development of cognitive styles. International Journal of Psychology, _4, 293-305. Overnrier, J.B. & Seligman, M.E.P. (1967). Effects of inescapable shock on subsequent escape and avoidance learning. Journal of Comparative & Physiological Psychology, 63, 23-33. Panda, K. (1971). Effects of social reinforcement, locus of control, and cognitive style on concept learning among retarded children. Paper presented at the meeting of the American Educational Research Association, New York City. Patterson, O. (1975). Context and choice in ethnic allegiance: A theoretical framework & Caribbean case study. In N. Glazer & D.P. Moynihan (Eds.) Ethnicity: Theory & Experience. Cambridge, Mass.: Havard University Press. Penk, W. (1969). Age changes and correlates of Internal-external locus of control scale. Psychological Report, 25, 856. Phares, E.J. (1976). Locus of Control in Personality. N.J.: General Learning Press. Raynor, J.O. & Zubek. J.M. (1984). Information value and functioning reinterpretation of cognitive development and theory in terms of uncertainty orientation. Unpublished manscript. University of Buffalo. Rotter, J.B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General & Applied, 80, 1-28 Rotter, J.B. (1975). Some problems and misconceptions related to the construct of internal versus external control of reinforcement. Journal of Consulting & Clinical Psychology, 43, 56-67. Schneider, J.M., & Parsons, CA. (1970). Categories on the locus of control scale and cross-cultural compariosns in Denmark and the United States. Journal of Cross-Cultural Psychology, _1> 131-138. Schrader, H.M., Driver, M.J., & Streufert, S. (1967). Human Information Processing . N.Y. : Holt & Winston. / 83 Seligman, M.E.P. (1975). Helplessness: On depression, development and death. San Francisco: Freeman. Schleifer, M. (1972). Moral judgements of children. Dissertation Abstract International, 33, 452. Singh, P.N., Huang, S., & Thompson, G.C. (1962). A comparative study of selected attitudes, values, and personality characteristics of American, Chinese, and Indian students. Journal of Social Psychology, 57, 123-132. Sorrentino, R.M., Short, J.A.N., & Raynor, J.O. (1984). Uncertainty orientation: Implications for affective and cognitive views of achievement behavior. Journal of Personality & Social Psychology, 46, 189-206. Tsui, C.L.C. (1977). Culture and control orientation: A study of internal-external locus of control in Chinese and American-Chinese women. (Doctoral dissertation, University of California, Berkeley). International Dissertation Abstracts, 39, 770A. Wallston, K.A., Wallston, B.S., & DeVellis, R. (1978). Development of multidimensional health locus of control (MHLC) Scales. Health Education Monographs, 0, 160-170. Witkin, H.A. & Berry, J.W. (1975). Psychological differentiation in cross cultural perspective. Journal of Cross Cultural Psychology, _6, 4-67. Witkin, H.A., Dyk, R.B., Faterson, H.F., Goodenough, D.R., & Karp, S.A. (1962). Psychological Differentiation: Studies of Development. New York: Wiley. Witkin, H. A. & Goodenough, D.R. (1977). Field dependence and interpersonal behavior. Psychological Bulletin, 84, 661-689. Witkin, H.A., & Goodenough, D.R. (1981). Cognitive Styles: Essence and Origins. Psychological Issues: Monograph 51. New York: International Universities Press, Inc. Witkin, H.A., Oltman, P.K., Raskin, E. & Karp, S.A. (1971). Manual for Embedded Figure Tests. Palto Alto: Consulting Psychologists Press, Inc. Witkin, H.A., Price-Williams, D. Bertini, M. Christiansen, B. Oltman, P.K., Ramirez, M., & Ven Meel, J. (1974). Social conformity and psychological differentiation. International Journal of Psycholgoy, _9_, 11-29. Wright, A.F. (1962). Values, roles and personalties. In A.F., Wright & D. Twitchett (Eds.). Confucian Personalities (p. 323). Stanford: Stanford University Press. / 84 Yang, K.S. (1981). The formation and change of Chinese personality: a cultural-ecological perspective. Acta Psychologica Taiwanica, 23, 39-56. A P P E N D I X A . C O N S E N T F O R M 85 / 86 Bonita Lau University of British Columbia, Dept. of Educational Psychology & Special Education, Faculty of Education, 2125 Main Mall, University Campus, Vancouver, B.C., Canada, V6T 1Z5. Re: Consent Form Dear Participant, I am seeking your participation in a study being carried out by myself and Dr. R. Conry of the Department of Educational Psychology & Special Education at the University of British Columbia, Canada. The research investigates the relationship between visual perception and different opinions about certain personal and social issues. Your participation will involve answering three questionnaires and a figure test; it will take about one hour of your time. The information we gather will be helpful to researchers working in the area of visual perception. In addition, you may discover some interesting things about yourself as you complete the questionnaires. Your assistance in this study would be greatly appreciated. However, your participation is totally voluntary. You may withdraw at any time or refuse to answer any specific question that you object to. All information will be kept confidential; it is to be used for research only. You are not required to write your name on any of the forms. Furthermore, all the obtained data will be destroyed when the study is completed. Please detach the consent form from this letter and complete the consent form. REMEMBER TO SUBMIT THE CONSENT FORM WITH THE ENCLOSED QUESTIONNAIRES. If you have any questions, please feel free to contact Bonita Lau or Dr. R. Conry at UBC, Canada (Telephone: the mailing address is shown as the above). Thank you very much for your assistance. Yours sincerely, (Bonita LAU) / 87 Consent Form I (consent/do not consent) to participate in the research investigated by Bonita Lau and Dr. R. Conry at the University of British Columbia, Canada. I am aware that about one hour will be required to finish the study, and that the questionnaires will be returned anonymously to the University of British Columbia, Canada for scoring: I understand that the results of this study will be kept confidential. Also, I realize that my participation \" in this study is totally voluntary and it may be terminated at any time I like. As a participant, I will complete the enclosed questionnaires and submit them with this consent form. (signature) (Date) / 88 x/es\"/ AWfAj MALL, UNttfsGyir campus. M *h /I A. A fit 4 \" ^ t % kk t % t fC%-^ ^ Ifat%\-%-MX^MW