Change of economic structure and social trends directly or indirectly affect dietary habits. More and more consumers expect to access quick, hygienic, delicious, and inexpensive food (Chen, 1997). In recent years, all-you-can-eat buffet has become a prevailing pattern in the diet market of Taiwan. From a common saying on the street, one can recognize the general attitude toward all-you-can-eat buffet: "There are three kinds of treasure (Pao) in Taiwan: health insurance (Chian Pao), labor insurance (Lao Pao), and $199 all-you-can-eat (Chih Tao Pao)." The all-you-can-eat buffet service not only allows one to enjoy a varied and abundant diet but to avoid the inconvenience of ordering and waiting. Although all-you-can-eat buffet is relatively popular in Taiwanese adolescents' diet market (Ai, 2001), the relationship of their buffet preferences and psychological correlates is largely unknown. However, since Taiwanese adolescents' buffet preference is nationwide (Lu, 1997; Su, 2000), it can be investigated from a cultural psychology perspective. The present research focused on the prominent characteristics of adolescents' self-construal in a collectivist society and examined whether they were associated with their buffet preference.
CHINESE SELF-CONSTRUAL AND DIET PREFERENCE
Taiwan is deeply influenced by Chinese culture (Yang, 1992). In general, Chinese and Western cultures exhibit extremely different self-construals that affect individuals' cognition, emotion, and motives and thereby determine their behavior. Western culture stresses independence and autonomy--the development of one's distinct characteristics and self-actualization, whereas the highly collectivist Chinese culture tends to value the connectedness and interdependence between individuals and other people (e.g., Kim, 1997; Markus & Kitayama, 1991; Voronov & Singer, 2002). Collectivists' interdependent self-construal suggests that individuals' behavior, opinions, and judgments are contingent upon interpersonal interaction; in other words, collectivists are quite flexible and contingency-oriented (Triandis, Bontempo, Villareal, Asai, & Lucca, 1988; Yang, 1993).
Based upon differential self-construal across cultures, national behavior and consuming preferences are associated with that self-construal (Lee & Green, 1991). Roth (1995) proposed that consumers' preferences would be affected by their social value system. In an interdependent self-construal culture, whether product or service meets social function is critically important. In the context of diet consumption, an all-you-can-eat buffet offers high flexibility in its diverse choices (Ai, 2001; Su, 2000). If the characteristics of buffet consumption are congruent with Chinese self-construal, one might further assume that their diet preferences would be associated with self-construal.
Contingent Orientation and Buffet Preference
In terms of Chinese interdependent self-construal, contingent orientation has been widely demonstrated and recognized as a crucial determinant of their overt behavior (Kim, 1997; Markus & Kitayama, 1991). Contingency-orientation in Chinese self-construal refers to flexibility in making personal decisions.
Adolescents generally eat at a buffet with others (Ai, 2001). Since buffet-style restaurants offer multiple options, those in collective consumption are more likely to find what they prefer. Further, seeking interpersonal harmony and avoiding interpersonal disagreement and conflict are highly valued in Chinese culture (Kim & Markus, 1999; Yang, 1992). Buffets not only are more likely to satisfy diverse preferences, but coincide with Chinese cultural norms. Therefore, when other conditions (such as price, environment, service quality) are equivalent, Chinese adolescents with higher contingent orientation are more likely to prefer all-you-can-eat restaurants. Contingency-orientation in Chinese interdependent self-construal may highlight the advantages of buffet restaurants over others, such as set-menu or single-cuisine. It was predicted that Taiwanese adolescents' contingent orientation is associated with their preferences for all-you-can-eat buffet.
Constrained Autonomy and Buffet Preference
Self-construals are quite different in Eastern and Western cultures which affects an individual's development of autonomy. Western culture emphasizes the expression of one's inner needs and rights (Markus & Kitayama, 1991). However, in Chinese culture (Yang, 1992), a self-construal tends to regard the interdependence between individuals and their significant others as an ultimate goal. In order to satisfy significant others' expectations, the Chinese need for autonomy tends to yield to a social-orientation (Triandis et al., 1988; Triandis, 1989). Those in a collectivist culture (e.g., Taiwan in the present study) tend to restrain or hide their real selves in order to play roles that are accepted by significant others, family, and society (Yang, 1993). Thus, collectivists' autonomy is obviously constrained compared to that of individualist cultures in which personal autonomy is highly encouraged and respected (Hsu, 1981).
Further, the literature has consistently noted that the development and practice of adolescent autonomy is likely to be constrained in major life activities such as education, friendship, marriage, and occupation, especially by parents (e.g., Hsu, 1981; Yang, 1992, 1993). In such circumstances, Chinese adolescents may seek compensation for the lack of individual independence in other areas. Thus, the freedom of choice offered by buffet consumption can help adolescents achieve some sense of autonomy. Being free from the limitations of diet choices in single-cuisine or set-menu restaurants (Ai, 2001; Lu, 1997; Su, 2000), buffet restaurants may be more likely to attract Chinese adolescents with constrained autonomy. Thus one could assume that, if Chinese adolescents possess less autonomy in major life events, they will be more likely to seek compensatory satisfaction through buffet consumption compared to those who feel greater autonomy. Therefore, it was predicted that Taiwanese adolescents' perceived autonomy is related to their buffet preference.
Generally, this research employed two quantitative studies. By conducting a panel survey, the first study aimed to determine whether the prominent characteristics in Chinese self-construal (contingent orientation and constrained autonomy) were both associated with their preferences for all-you-can-eat buffet. The second study investigated the effect of perceived autonomy on buffet preference.
Participants and Design
The adolescent population was stratified into three demographic areas: Northern, Central, and Southern Taiwan. In a longitudinal survey, participants were asked to complete a series of questionnaires about their contingency and autonomy in major life events as well as their preferences for all-you-can-eat buffet. Among the 530 Taiwanese adolescents in the formal sample, 286 were males (54%) and 244 were females (46%).
Analysis of this study was conducted with the cross-lagged panel design (Cook & Campbell, 1979). According to this framework (see Figure 1), "variable A" is contingency or autonomy whereas "variable B" is buffet preference. Subtitled numbers signify the time of measurement: "one" denotes the pretest and "two" denotes the posttest. Test-retest reliability coefficients were calculated by the correlation of [A.sub.1] and [A.sub.2] as well as [B.sub.1] and [B.sub.2]. The correlations of [A.sub.1][B.sub.1] and [A.sub.2][B.sub.2] are defined as the synchronous coefficients, which are used to determine if the relationship of test variables is stable and consistent over time. The correlations of [A.sub.1][B.sub.2] and [B.sub.1][A.sub.2] are defined as the panel coefficients and may provide information about the mutual predictability of test variables.
[FIGURE 1 OMITTED]
Contingent orientation. Adolescents' contingent orientation was measured by the respondents' evaluations while making decisions on major life events. Items were produced according to the development task theory (Havighurst, 1972) and the Seven Vectors Model (Chickering & Reisser, 1993). Development tasks for adolescents were selected as the initial items which denote their major life events. An initial sample of 202 participants were recruited for scale development. Those items which failed to increase internal consistency and whose item-scale correlations were not significant were excluded from the pool. The formal scale consisted of 15 items. Participants reported their contingency when making decisions on these major life events on a 6point scale from least contingently to very contingently. The higher the score, the greater the contingency on major life events. The internal consistency (Cronbach's Alpha) of the formal scale was .93, and the test-retest reliability was .69 over a half-year period.
Perceived autonomy. Items on the autonomy scale were the same as those on the contingency scale, which also consisted of major life events for adolescents. Respondents were asked to rate their autonomy on a 6-point scale from totally constrained to totally autonomous. Higher scores represent greater autonomy when participants make decisions on major life events. The internal consistency (Cronbach's Alpha) was .95 and the test-retest reliability was .93.
Buffet preference. Preference for all-you-can-eat buffet was measured by relative preference in the choice alternatives of diet consumption. Relative preference for buffet over single-cuisine and set-menu restaurants was measured on a non-graded scale with two end points: "All-you-can-eat buffet restaurants are more attractive" and "Single-cuisine or set-menu restaurants are more attractive." Participants were asked to rate their relative preferences under other consuming conditions (price, atmosphere, transportation, service quality, and decoration styles) using the two alternatives. Ten occasions were selected from a preliminary pretest (N = 202) as the most frequent diet-consuming conditions, such as birthday celebration, peer get-together, victory celebration, banquet for appreciating teacher, date, family meeting, during travel, and festival celebration. Respondents' evaluations on each occasion were later quantified by measuring the distance from the lower end of the scale to the respondents' mark and were standardized on a 100-point scale (Chernev, 2001). Thus 1 signified "strong preference for single-cuisine dish or set-menu" and 100 signified "strong preference for all-you-can-eat buffet." Participants' average scores of relative preferences across 10 occasions were computed for subsequent analysis.
According to the data of normality testing and the means of measures in this study (see Table 1), respondents' contingency was toward the lower end on a 6-point scale, whereas their autonomy was toward the higher end at both pretest and posttest. The results not only indicated that participants' contingent orientation and constrained autonomy were very high but also showed that they are prominent characteristics among Taiwanese adolescents. Participants' buffet preferences over the other diet consumptions were significantly higher than the midpoint (50) of a 100-point scale in the pretest and posttest data, t(529) = 14.17 at pretest, p < .001; and t(529) = 16.82 at posttest, p < .001, respectively. The findings revealed that participants exhibited greater buffet preference over the others.
To test the hypotheses, participants' ratings on contingency and autonomy served as predictor variables, and their buffet preference was treated as a criterion variable. Multiple regression analysis was conducted in the pretest and posttest data respectively (see Table 2).
As seen in Table 2, overall model testings at both pretest and posttest were significant, Adjusted [R.sup.2.sub.pretest] = .46, F(2, 527) = 227.09, p < .001; and Adjusted [R.sup.2.sub.posttest] = .19, F(2, 527) = 62.93, p < .001, respectively. The regression coefficients at pretest revealed that contingency was positively correlated with the buffet preference, [beta] = 0.60, t(528) = 18.48, p < .01, whereas autonomy was negatively correlated with the buffet preference, [beta] = -0.23, t(528) = -7.02, p < .01. In the posttest data, the regression coefficients of contingency ([beta] = 0.37) and autonomy ([beta] = -0.22) also exhibited a similar pattern with the findings found at pretest, t(528) = 9.30 for contingency, p < .01; and t(528) = -5.62 for autonomy, p < .01, respectively. Together these findings indicated that both contingency and autonomy could serve as the effective predictors of participants' buffet preferences.
The cross-lagged panel analysis (Cook & Campbell, 1979) was conducted to further test the mutually predictive relationships between the characteristics of Chinese self-construal (contingent orientation and constrained autonomy) and participants' buffet preferences. Results of the cross-lagged panel analysis were presented with the correlation matrixes (see Table 3).
With regard to the relationship between contingency and buffet preference, the synchronous coefficients in the pretest and posttest data were both significant, [r.sub.12] = .64, p < .01; and [r.sub.34] = .38, p < .01, respectively. These findings indicated that Chinese participants' contingent orientation was positively correlated with their buffet preferences, which also replicated the previous findings observed in Multiple Regression Analysis. Moreover, the correlations of contingency and buffet preference were stable and consistent over a half-year period.
The panel coefficients of the present analysis were significant. Specifically, participants' contingency at pretest was positively correlated with their buffet preferences at posttest, [r.sub.14] = .57, p < .01, and their buffet preferences at pretest were also positively correlated with contingency at posttest, [r.sub.23] = .38, p < .01. The results indicated that participants' prior contingent orientation could predict their subsequent buffet preference.
With regard to the cross-lagged panel analysis of autonomy and buffet preference, the synchronous coefficients were significant, [r.sub.12] = .34, p < .01; and [r.sub.34] = -.25, p < .01, respectively. The findings indicated that participants' autonomy was negatively associated with the buffet preference, and that the correlation of autonomy and the buffet preference was stable and consistent over a half-year period, and showed a consistent pattern with the findings observed in multiple regression analysis.
The panel coefficients of autonomy and the buffet preference showed that they could be predicted by each other. Specifically, participants' prior autonomy was negatively correlated with their subsequent buffet preference, [r.sub.14] = -.32, p < .01, and their prior buffet preference was also negatively correlated with the subsequent perceived autonomy, [r.sub.23] = -.28, p < .01.
In general, the results in multiple regression analyses and the cross-lagged panel analyses support the predictions. In a collectivist country with high interdependent self-construal, adolescents' contingent orientation and constrained autonomy are both closely associated with their buffet preferences.
The aim of this experimental study was to examine the impact of adolescents' perceived autonomy on their buffet preferences. The findings could be compared to the results of the first study to achieve more reliable conclusions. However, contingency is a trait-like construct, which is not appropriate for experimental manipulation. Thus, the second study only examined the effect of adolescents' perceived autonomy on their buffet preference.
Participants and Design
There were 288 college students who participated in this experiment. Perceived autonomy was manipulated between subjects. Participants were randomly assigned to one of the perceived autonomy conditions (low vs. moderate vs. high) with block-random method in groups of three. In the high level of perceived autonomy condition, there were 52 males and 44 females; 42 males and 54 females were in the moderate level of perceived autonomy condition. There were 46 males and 50 females in the low level of perceived autonomy condition.
Participants' perceived autonomy was manipulated with the emotional-event recollection technique (Leith & Baumeister, 1996), which was designed to enable them to experience various levels of autonomy through the recall of major life events. The emotional-event recollection was disguised by a self-reflection study. Participants were informed that the main purpose of the research was "to explore the relationship between self-reflection and decision-making." Before starting to recall past life events, participants received a booklet describing self-reflection as the "ability to re-experience personal past events with significant meaning here and now." To increase involvement in this task, participants were further told that "people with better self-reflection are found to be better parents, lovers, couples and managers, and they tend to learn lessons from experiences which enable them to avoid making the same mistakes."
Next, each participant was presented with instructions regarding one of the perceived autonomy conditions. At the high level of perceived autonomy condition, participants were asked to recall two major life events in which they felt the strongest sense of control and sufficient autonomy. Respondents who were assigned to the low level of perceived autonomy condition were asked to recall two major life events in which they felt the weakest sense of control and autonomy. For the condition of the moderate level of perceived autonomy, they were asked to recall one major life event with a high sense of control and another with a low sense of control.
Each participant was given enough time to become involved in the recollection procedure. Yoked-control method was employed to control the duration of participants' recollections for each session. After participants reported that they were capable of reentrance in the recalled life events, they were asked to rate their affective states and sense of control on a 7-point scale. Finally, each respondent was asked to imagine going for a meal in the present mood--which diet style was preferable.
Relative preferences for all-you-can-eat buffet over single-cuisine and set-menu restaurants were measured on a non-graded scale with two end points: "All-you-can-eat buffet is more attractive" and "Single-cuisine dish or set-menu restaurant is more attractive." Participants were asked to rate the relative preferences without considering other conditions related to diet preference, such as price, atmosphere, transportation, and service quality. Respondents' evaluations were later quantified by measuring the distance from the lower end of the scale to the respondents' mark and then standardized on a 100-point scale with I signifying "strong preference for single-cuisine dish or set-menu restaurant" and 100 signifying "strong preference for all-you-can-eat buffet."
Manipulation check was employed before the formal analysis. Respondents' sense of control and their positive affective states were used to check the manipulation of perceived autonomy (see Table 4).
For participants' sense of control, one-way ANOVA showed that the sense of control in the three autonomy groups was significantly different, F(2, 285) = 26.46, p < .001. Further trend analysis indicated that individuals with a higher level of perceived autonomy had a higher sense of control, F(1, 285) = 52.16, p < .001. With regard to participants' affective states, positive affective states (happiness and joy) should be positively correlated with a higher level of perceived autonomy. The linear trend effect of perceived autonomy on positive affects was significant, F(1, 285) = 15.89, p < .001, and this finding indicated that the higher the level of perceived autonomy, the higher the scores on positive affective states. Total findings in the manipulation check indicated that the manipulation of perceived autonomy was satisfactory.
ANOVA was then conducted to determine if the perceived levels of autonomy affect the preference for all-you-can-eat buffet. Results showed that the main effect of perceived autonomy was significant, F(2, 285) = 25.58, p < .001. Subsequent trend analysis indicated that the linear effect was significant, F(1, 285) = 51.06, p < .001, and this finding revealed that individuals with higher perceived autonomy had a higher preference for all-you-can-eat buffet than the other styles. Further comparisons indicated that the buffet preference of the low autonomy group ([M.sub.Low] = 51.88) was significantly higher than the moderate autonomy group ([M.sub.Moderate] = 37.08), t(285) = 3.84, p < .001, whereas the buffet preference of the moderate autonomy group was significantly higher than the high autonomy group ([M.sub.High] = 24.33), t(285) = 3.31, p < .001. Generally, the data from the experimental study suggests that the buffet preference was negatively associated with participants' perceived levels of autonomy, which was also consistent with the results obtained in the first study. The autonomy data in the two quantitative studies also showed that Chinese consumers' perceived autonomy is closely related to their preference for all-you-can-eat buffet.
The research employed a panel survey and a laboratory experiment to explore the determinants of Chinese adolescents' buffet preferences based on the characteristics of collectivists' interdependent self-construal, namely, contingent orientation and constrained autonomy. The results of Study 1 showed that contingent orientation and constrained autonomy are prominent characteristics among Taiwanese adolescents, and that they preferred buffet over single-cuisine or set-menu restaurants.
For adolescents' contingent orientation, the results of Study 1 showed it to be positively correlated with buffet preference and also served as a significant predictor of this diet preference. The findings of the cross-lagged panel analysis echoed the relationship between participants' contingency and their buffet preferences over a half-year period. As noted above, collectivists' self-construal exhibits situation-specific and contingent orientation in making decisions (Roth, 1995). All-you-can-eat buffet provides abundant variety of dishes as a primary appeal (Ai, 2001; Su, 2000). "Multiple and flexible choices" in buffet consumption coincide with the characteristic of Chinese contingent orientation. Therefore, Chinese adolescents who were more contingency oriented exhibited higher preferences for buffet consumption.
With regard to autonomy, for Chinese adolescents' in major life events it is vulnerable (Yang, 1992, 1993). This is because satisfying significant others' expectations and concern for interdependence are highly valued in a culture of high collectivism (Triandis, 1989). Buffet, which offers sufficient autonomy in choosing dishes and provides relief from the constraints of single-cuisine or set-menu meals, might provide Chinese adolescents with a convenient way to compensate for their limited autonomy in major life events. The outcomes of Study 1 seem to support this view. Findings indicated that participants whose autonomy was lower in major life events exhibited greater preferences for buffet. Moreover, the results of the cross-lagged panel analysis revealed that participants' prior autonomy was negatively correlated with their subsequent buffet preferences. Research findings suggest that due to Chinese adolescents' constrained autonomy in major life events, they might adopt diet consumption as an alternative way to fill their need for autonomy-seeking. Chinese adolescents, who have lower autonomy in major life events, may be more likely to obtain replacement satisfaction through buffet consumption.
The results of Study 2 also supported the view that adolescents' perceived autonomy would affect their preferences for all-you-can-eat buffet. It indicated that the higher the perceived level of autonomy, the stronger the preference for all-you-can-eat buffet over the other choices. Since the manipulation of participants' perceived autonomy in a laboratory experiment was temporary and the measurement of buffet preference was immediate, it is appropriate to propose that Chinese adolescents' "state-like" buffet preferences might be influenced by the immediate consciousness of autonomy. According to the notion of "affect-as-information" (Schaller & Cialdini, 1990; Zillmann, 1988), individuals' moods or affective states tend to be involved in decision making at that moment. Those who are experiencing negative affective states would take action or make decisions that would compensate for or alter negative moods. Thus, lower perceived autonomy may result in negative affective states and then affect participants' diet preferences. As a result, Chinese adolescents with lower autonomy tend to use buffet consumption to alter negative moods.
Limitations and Future Directions
The present research employed only a particular cohort in Taiwan which may limit generalizability of the results. Future research could employ varied cohorts to expand validity of the findings.
Both collectivist and individualist self-construal may coexist in a society and further affect national consuming preferences (Triandis, 1989; Triandis, McCusker, & Hui, 1990). Chinese tend to be collectivists and their consuming habit in all-you-can-eat restaurants is collective consumption. Based on the characteristics of "relationship-orientation" and "favor-orientation" in Chinese culture (Huang, 1987; King, 1991; Yang, 1992), one can attempt to determine if adolescents with high "favor-orientation" will be more likely to prefer all-you-can-eat buffet, particularly for a group dinner party.
With regard to the issue of cross-cultural comparisons, findings of this research indicated that both contingent orientation and constrained autonomy in Chinese adolescents were associated with their buffet preferences. As to adolescents in highly individualistic Western cultures, the diverse choices in all-you-can-eat buffet may be less attractive, because those cultures value the uniqueness of personal expression (Yang, 1992). Moreover, individuation and autonomy is highly encouraged in an individualist culture (Kim, 1997; Lee & Green, 1991). Individualists' autonomy is not likely to be constrained as it in a collectivist culture such as Taiwan. Therefore, the relationship of autonomy with buffet preference may not be observed in an individualist diet market.
Culture, a concept crucial to the understanding of consumer behavior, determines the overall priorities consumers attach to different activities and products, and also determines the success or failure of specific products and services. The present research demonstrated that Chinese adolescents' contingent orientation is closely related to their diet preferences. For marketers who want to target collectivist cultures with high interdependent self-construal (e.g, Mainland China, Taiwan, Japan, Turkey, and Thailand; Triandis, 1989), "multiple choices" and "a flexible combination of diverse cuisines" may resonate with those cultures.
In terms of constrained autonomy among the Chinese, the results in both studies consistently indicated that set-menu and single-cuisine dishes were less attractive than all-you-can-eat buffet when other consuming conditions are equivalent. Buffet consumption may serve as an alternative choice to compensate for adolescents' constrained autonomy in Chinese culture. Thus, "sufficient autonomy" and "freedom of choice" seem to be crucial criteria for Chinese adolescents' diet consumption. Marketers who wish to enter the Chinese diet market might do well to monitor the dynamic change in adolescents' autonomy development in order to determine if all-you-can-eat buffet would be an alternative compensation for their constrained autonomy.
In conclusion, adolescents' prominent characteristics of self-construal may be related to their diet consumption. Variations across cultures indicate that each culture is unique, with its own value system, conventions, and regulations (see Clark, 1990, for a similar viewpoint). Therefore, if what we eat is who we are, researchers who wish to explore adolescents' diet consumption and its psychological correlates should take their self-construal or national character into consideration.
This research was supported by a grant from the National Science Council of the Republic of China. The author acknowledges the assistance of Jing-Yu Hsu and Yu-Ping Lin in collecting and coding the data, and also thanks ChaoChin Yang for her comments on the proposal of this research.
Ai, M. T. (2001). A qualitative study on marketing strategies for buffet services at international tourism hotels in Taipei [in Chinese]. Unpublished master's dissertation, Shih Hsin University, Taipei, Taiwan.
Chen, Y. (1997). The enhancement of service quality in the food and beverage industry [in Chinese]. Technological and Vocational Education Bimonthly, 39, 49-54.
Chernev, A. (2001). The impact of common features on consumer preferences: A case of confirmatory reasoning. Journal of Consumer Research, 27, 475-488.
Chickering, A. W., & Reisser, L. (1993). Education and identity (2nd ed.). San Francisco, CA: Jossey-Bass.
Clark, T. (1990). International marketing and national character: A review and proposal for an integrative theory. Journal of Marketing, 54, 66-79.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis issues for field settings. Boston, MA: Houghton Mifflin.
Havighurst, R. J. (1972). Developmental tasks and education (3rd ed.). New York: David McKay.
Hsu, F. L. K. (1981). American and Chinese: Passage to differences (3rd ed.). Honolulu, HI: University of Hawaii Press.
Huang, K. K. (1987). Face and favor: The Chinese power game. American Journal of Sociology, 92, 944-974.
Kim, U. (1997). Asian collectivism: An indigenous perspective. In H. S. R. Kao & D. Sinha (Eds.), Asian perspectives on psychology (pp. 147-163). New Delhi, India: Sage.
Kim, H., & Markus, H. R. (1999). Deviance or uniqueness, harmony or conformity? A cultural analysis. Journal of Personality and Social Psychology, 77, 785-800.
King, A. Y. C. (1991). Kuan-his and network building: A sociological interpretation. Deadalus, 120, 63-84.
Lee, C., & Green, R. T. (1991). Cross-cultural examination of the Fishbein behavioral intentions model. Journal of International Business Studies, 22, 289-305.
Leith, K. P., & Baumeister, R. F. (1996). Why do bad moods increase self-defeating behavior? Emotion, risk-taking, and self-regulation, Journal of Personality and Social Psychology, 71, 1250-1267.
Lu, C. Y. (1997). A study of consumer's eating behavior in buffet [in Chinese]. Unpublished master's dissertation, National Taiwan Normal University, Taipei, Taiwan.
Markus, H., & Kitayama, S. (1991). Culture and self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224-253.
Roth, M. S. (1995). Effects of global market conditions on brand image customization and brand performance. Journal of Advertising, 24, 55-75.
Schaller, M., & Cialdini, R. B. (1990). Happiness, sadness, and helping. In T. E. Higgins & R. M. Sorrentino (Eds.), Handbook of motivation and cognition (pp. 265-296). New York: Guilford.
Su, F. J. (2000). Introduction to food and beverage management [in Chinese]. Taipei, Taiwan: Jing-Hwei Yang.
Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96, 506-522.
Triandis, H. C., McCusker, C., & Hui, H. C. (1990). Multimethod probes of individualism and collectivism. Journal of Personality and Social Psychology, 59, 1006-1020.
Triandis, H. C., Bontempo, R., Villareal, M. J., Asai, M., & Lucca, N. (1988). Individualism vs. collectivism: Cross-cultural perspectives on self-in-group relationship. Journal of Personality and Social Psychology, 54, 323-338.
Voronov, M., & Singer, J. A. (2002). The myth of individualism-collectivism: A critical review. The Journal of Social Psychology, 142, 461-80.
Yang, K. S. (1992). Can traditional and modern values coexist? In Proceedings of the Conference on Chinese Perspectives on Values (pp. 117-158). Taipei, Taiwan: Center for Sinological Studies.
Yang, K. S. (1993). Chinese social orientation: An integrative analysis. In W. L. Y. C. Cheng, F. M. C. Cheung & C. N. Chen (Eds.), Psychotherapy for the Chinese (pp. 19-56). Hong Kong: Chinese University of Hong Kong.
Zillmann, D. (1988). Mood management: Using entertainment to full advantage. In L. Donohow & H.E. Sypher (Eds.), Communication, social cognition, and affect (pp. 147-171). Hillsdale, NJ: Erlbaum.
Requests for reprints should be sent to Wen-Bin Chiou, General Education Center, National Kaohsiung Hospitality College, No. 1, Sung-Ho Rd., Kaohsiung 812, Taiwan, ROC. E-mail: email@example.com
Descriptive Statistics of the Measures in Study One
Variable M SD Skewness Test
Contingency 4.47 1.21 .22 .10 *
Autonomy 2.30 1.10 .57 .14 *
Buffet preference 67.56 28.45 .35 .26 *
Contingency 4.20 1.24 .23 .10 *
Autonomy 2.34 1.17 .43 .10 *
Buffet preference 65.23 20.84 .41 .11 *
Note. Contingency and autonomy were measured on a 6-point
scale. The rating for buffet preference was transformed
into a 100-point scale. Normality was examined with the
Kolmogorov-Smirnov test (N = 530).
* p < .05.
Regression of Buffet Preference on Contingency and Autonomy
in Study One
Variable B SE B [beta] t Adjusted
Contingency 14.14 0.77 0.60 ** 18.48 .46 **
Autonomy -5.98 0.84 -0.23 ** -7.02
Contingency 6.14 0.66 0.37 ** 9.30 .19 **
Autonomy -3.94 0.70 -0.22 ** -5.62
Note. Criterion variable was buffet preference. Multiple
regression analyses were conducted with the simultaneous
model (N = 530).
** p < .01.
Correlation Matrixes of the Cross-Lagged Panel Analysis in Study One
Variable 1 2 3 4
Cross-lagged panel analysis of contingency and buffet preference
1. Contingency at pretest --
2. Buffet preference at pretest .64 (a) --
3. Contingency at posttest .69 (b) .38 (c) --
4. Buffet preference at posttest .57 (c) .84 (b) .38 (a) --
Cross-lagged panel analysis of autonomy and buffet preference
1. Autonomy at pretest --
2. Buffet preference at, pretest -.34 (a) --
3. Autonomy at posttest .93 (b) -.28 (c) --
4. Buffet preference at posttest -.32 (c) .84 (b) -.25 (a) --
Note. N = 530. All correlations were significant (p < .01).
(a) The synchronous coefficients.
(b) The coefficients of test-retest reliability.
(c) The panel coefficients.
Means and Standard Deviations of the Measures in Study Two
Perceived Autonomy Group
Variable Low Moderate High
Sense of control
M 3.42 4.63 5.42
SD 1.86 2.23 1.51
Positive affective states
M 3.12 3.46 4.75
SD 1.50 1.62 1.78
M 51.88 37.08 24.33
SD 29.32 27.29 23.12
Note. Each group consisted of 96 participants. Sense of control
and positive affective states were both rated on a 7-point scale.
Buffet preference was rated on a non-graded scale and later
transformed into a 100-point scale.