Personal Values and Shopping Behavior: a Structural Equation Test of the Rvs in China

Citation:

Zhengyuan Wang and C.P. Rao (1995) ,"Personal Values and Shopping Behavior: a Structural Equation Test of the Rvs in China", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 373-380.

Advances in Consumer Research Volume 22, 1995      Pages 373-380

PERSONAL VALUES AND SHOPPING BEHAVIOR: A STRUCTURAL EQUATION TEST OF THE RVS IN CHINA

Zhengyuan Wang, University of Arkansas

C.P. Rao, Old Dominion University

INTRODUCTION

In recent years, one of the most dynamic areas of research in social science disciplines has been the measurement and functions of personal values (Kamakura and Mazzon 1991). Despite considerable empirical research on functions of values (Pitts and Woodside 1984), the measurement issue remains unresolved. There is disagreement among researchers concerning the conceptualization of values and the validity and reliability of various measures of values (e.g., Beatty et al. 1985; Miethe 1985; Munson 1984; Novak and MacEvoy 1990). Most previous research relies on exploratory factor analysis to assess the underlying dimensionality of values (e.g., Feather and Peay 1975; Maloney and Katz 1976; Munson and McQuarrie 1988; Valette-Florence and Jolibert 1990; Vinson et al. 1977). Many researchers regard values as consisting of independent dimensions which can be measured by single items (e.g., Rokeach 1973; Schopphoven 1991).

However, exploratory factor analysis and the use of single-item measures have shown to be inadequate in construct validation (Gerbing and Anderson 1988; Nunnally 1978). The present study attempts to overcome some of the weaknesses in previous research. Using confirmatory factor analysis via LISREL (J÷reskog and S÷rbom 1988), it examined the dimensionality of the most widely used value instrument, the Rokeach Value Survey (RVS), in the context of Chinese culture. Although developed in Western countries, the Rokeach Value Survey has shown to be largely consistent with the Chinese Value Survey (Bond 1988; Ralston et al. 1992).

THEORETICAL BACKGROUND

Personal Values

Rokeach (1973) argues that values are separately organized into relatively enduring hierarchical structures of terminal and instrumental values. Terminal values are defined as a person's beliefs concerning desirable end-states of existence (ends) while instrumental values refer to a person's beliefs regarding desirable modes of conduct (means to achieve the ends). The Rokeach Value Survey (RVS) was developed by selecting 18 terminal and 18 instrumental values from a larger pool of several hundred values descriptors (Rokeach 1973). Earlier studies have provided some empirical evidence in support of the structure of terminal versus instrumental values (e.g., Feather and Peay 1975; Vinson et al. 1977).

The original version of the RVS consisted of 18 instrumental values and 18 terminal values that were rank ordered in terms of their importance as guiding principles of the respondent's life. The major limitations of the original RVS include: (1) subjects are forced to rank one value at the expense of another which may actually be equally important to them (Alwin and Krosnick 1985), (2) the presentation of the 36 value items may exceed the respondent's ability to accurately process information and thus distort the ranking procedure (Miller 1956), (3) the ranking nature of the data precludes the use of a wide variety of useful statistical analysis techniques that might otherwise be used (Rankin and Grube 1980).

To overcome these problems, researchers have modified the original RVS to yield an interval measure of value importance (Miethe 1985; Moore 1975; Munson and McIntyre 1979; Rankin and Grube 1980). Instead of ranking the values, respondents are required to assess the importance of each value item on a seven-point rating scale. Such a rating approach imposes fewer constraints on the data. Since the goal of value survey is to identify people's underlying value dimensions, the rating approach would seem to be preferred. The present study employed such a modified RVS. Since terminal and instrumental values are conceptually distinct constructs, the current study focused on instrumental values.

Structure of Instrumental Values

According to Rokeach (1973), both terminal and instrumental values are basically independent. In other words, each of the 36 value items represents a distinct personal value. Other researchers disagree with Rokeach and have sought to achieve a more parsimonious value structure (e.g., Beatty et al. 1985; McQuarrie and Langmeyer 1985; Munson and McQuarrie 1988; Prakash and Munson 1985). These researchers have noted that many of the Rokeach value items appeared to be largely irrelevant to consumption behavior. They argue that some value items can be dropped to make the measures of values more relevant to consumption. For example, Munson and McQuarrie (1988) developed their Value Instrumentality Inventory by retaining 24 value items out of the original 36 RVS items. One alternative measure of values, the List of Values (LOV), was developed by Kahle (1983) by modifying Rokeach's 18 terminal values into a smaller set of 9 values. In testing the causal model of value-attitude-behavior hierarchy, Homer and Kahle (1988) further reduced the nine values into three latent constructs: (1) individual internal values, (2) inter-personal internal values, and (3) external values.

A review of earlier studies that factor analyzed Rokeach's instrumental values suggests an internal versus external structure of instrumental values (Crosby et al. 1990; Feather and Peay 1975; Rokeach 1973; Vinson et al. 1977). Internal dimensions of Rokeach's instrumental values contain such value items as being ambitious, imaginative, independent, intellectual, etc. External dimensions reflect social conformity such as being forgiving, honest, polite or concern for others. The hypothesized structure of instrumental values is shown in Table 1. As noted by Crosby et al (1990), theoretical and empirical evidence of such internal and external dimensions can be found in a number of previous studies (e.g., Alwin and Jackson 1982; Lefcourt 1966; McGuire 1974).

The above theoretical arguments and empirical findings lead to the following two rival hypotheses:

H1: Rokeach instrumental values are independent with one another.

H2: Rokeach instrumental values contain two latent dimensions. One is internal and the other external.

Nomological Network of Instrumental Values

Previous research on RVS has called for further research that will examine the nomological network linking value dimensions to consumer behavior (e.g., Crosby et al. 1990). For a construct to have a real validity, it must be shown to be useful as an explanatory device. In other words, the construct must have nomological validity. The present study used two shopping behavior related constructs, shopping involvement and price sensitivity, to assess the nomological validity of instrumental values.

Earlier studies, both theoretical and empirical, have suggested linkages between instrumental values and shopping involvement and price sensitivity (e.g., Carman 1977; Darden et al. 1979; Petit et al. 1985). Figure 1 presents the hypothesized nomological network of instrumental values. Chinese consumers with externally focused values (conformity dimension) are likely to be more involved in shopping than those with internally focused values (self-direction dimension). In China, the traditional doctrine of Confucianism, which is the foundation of the Chinese value system (Kluckhohn and Strodbeck 1961), teaches the values of diligence and frugality (Yau 1988). In addition, the propaganda of the Chinese Communist Party, China's ruling party over the past four decades, has also been extolling the virtue of thrifty and persuading people to resist the so-called "bourgeois" way of life. Capitalism has often been portrayed in China as encouraging a hedonic life style that is self-indulgent and wasteful. As a result, it is socially desirable to save money and be a meticulous shopper in China. Since consumers with internally focused values (self-direction dimension) are less prone to normative compliance, they are expected to be less involved in shopping and less sensitive to price.

TABLE 1

HYPOTHESIZED STRUCTURE OF INSTRUMENTAL VALUES

FIGURE 1

NOMOLOGICAL NETWORK OF INSTRUMENTAL VALUES

TABLE 2

CHI-SQUARE AND GOODNESS OF FIT VALUES FOR ROKEACH INSTRUMENTAL VALUE MODELS

On the basis of the previous discussions, the following hypotheses are derived:

H3: There is a positive relationship between conformity dimension of instrumental values and the two constructs of shopping involvement and price sensitivity.

H4: There is a negative relationship between self-direction dimension of instrumental values and the two constructs of shopping involvement and price sensitivity.

METHOD

Research Design

The data used to test the research hypotheses resulted from personal interviews with pre-designed questionnaires. Data collection was undertaken respectively in two major Chinese cities: Beijing and Shanghai. No attempt was made to select the sample on a purely random basis. Rather, care was exercised in collecting a representative cross-section of the target population, which is defined in this study as the urban population in these two cities. Altogether 196 usable questionnaires were obtained in which 101 came from Beijing and 95 obtained from Shanghai. The demographic characteristics of the sample were similar to those of the target population.

The questionnaire was translated into Chinese by a doctoral student from China who was studying business administration at a Western university. The translated questionnaire was then back translated into English by another Chinese doctoral student. This process resulted in a number of modifications that were necessary to make the Chinese version of the questionnaire as less ambiguous as possible.

Operational Measures

Personal values measures. The construct of personal values was measured by asking respondents how important each of the 18 Rokeach instrumental values was to them on a seven-point importance rating scale. As will be discussed later, Rokeach instrumental values were found to comprise two latent dimensions: self-direction dimension and conformity dimension. The self-direction dimension consists of the following four items: (1) ambitious, (2) imaginative, (3) intellectual, and (4) logical while the conformity dimension consists of six items: (1) forgiving, (2) helpful, (3) honest, (4) cheerful, (5) responsible, and (6) self-controlled.

Shopping involvement and price sensitivity. Shopping involvement is defined as the self-relevance of shopping activities to the individual. It was measured on a 7-point Likert scale using four items adapted from the purchasing involvement scale developed by Slama and Tashchian (1985). They were: (1) "Before going shopping, I sit down and make out a complete shopping list;" (2) "I consult with friends and/or experts before making a major purchase;" (3) "I am an impulse buyer;" and (4) "I am a meticulous shopper." Note that item three above was reverse worded. Price sensitivity was measured with three items on a 7-point Likert scale: (1) "I like to bargain for most of the things I buy;" (2) "I shop a lot for specials;" and (3) "A store's own brand is usually a better buy than a nationally advertised brand."

RESULTS

Results of Testing Hypotheses 1 and 2

Hypotheses 1 and 2 are rival hypotheses regarding the structure of Rokeach instrumental values. The efforts to test H₁ and H₂ followed the recommended procedures for confirmatory factor analysis (Gerbing and Anderson 1988). Since H₁ states that Rokeach instrumental values are independent with one another, i.e., no factors underlie the observed Rokeach instrumental values, it constitutes the null or baseline model. Three hypothesized or comparison models were tested against the null model. To evaluate these competing models, chi-square differences and two incremental fit indices, namely, Bentler & Bonett's (1980) normed index, D₁, and Bentler & Bonett's (1980) non-normed index, r₂, were calculated. Table 2 presents values of chi-square and other goodness of fit indices for competing models of instrumental values.

As shown in Table 2, both the null model and one-factor model have very poor fit indices. The two- and three-factor models, on the other hand, have substantially improved fit indices. The hypothesized two-factor model, which corresponds to H₂, is shown in Table 1. Crosby et al. (1990) suggest that the external dimension of the two-factor model may contain two sub-dimensions: conformity and virtuousness. Conformity contains four items: (1) cheerful, (2) clean, (3) obedient, and (4) polite, while virtuousness consists of six items: (1) forgiving, (2) helpful, (3) honest, (4) loving, (5) responsible, and (6) self-controlled. So the three-factor model tested in the current study includes the following three factors: self-direction, conformity, and virtuousness.

Fit indices shown in Table 2 indicate that the three-factor model has a slightly better fit than the two-factor model. However, the inter-factor correlation between virtuousness and conformity, f₃₂, is 0.88, indicating a lack of discriminant validity between conformity and virtuousness. In the two-factor model, the inter-factor correlation, f₂₁, is -0.16, while the average factor loading is 0.71. These results provide support for H₂. It is also worth noting here that exploratory factor analysis using the criterion of eigenvalues greater than one also suggests a two-factor solution. The first two factors account for 57% of the total variance. Hence, it appears that Rokeach instrumental values contain two latent dimensions: internal and external values.

TABLE 3

MEASUREMENT MODEL RESULTS FOR INSTRUMENTAL VALUES

Results of Testing Hypotheses 3 and 4

Hypotheses 3 and 4 relate to the nomological network of instrumental values. As shown in Figure 1, two shopping behavior related constructs, shopping involvement and price sensitivity, were used to assess the nomological validity of instrumental values. Before a structural model assessing nomological validity is conducted, measurement models associated with instrumental values and shopping behaviors must first be examined (Anderson and Gerbing 1988). Following conventional scale development procedures (Nunnally 1978; Gerbing and Anderson 1988), the present study examined unidimensionality, reliability, convergent validity, and discriminant validity associated with each of the constructs shown in Figure 1.

Measurement Model for Instrumental Values. As recommended by Gerbing and Anderson (1988), confirmatory factor analysis was used to assess unidimensionality of instrumental values. As shown in Table 2, the initial two-factor confirmatory measurement model involving all 18 Rokeach instrumental value items resulted in a model fit that was not quite satisfactory (^c2=279.88, df=134, p=0.000; GFI=0.869, AGFI=0.833, RMSR=0.074). An examination of corrected item-to-total correlations as well as modification indices suggested the deletion of eight items. The practice of dropping some value items is consistent with previous research that attempted to modify existing measures of values (e.g., Kahle 1983; Munson and McQuarrie 1988).

As indicated in Table 3, a measurement model of the remaining ten items loading onto the two dimensions of instrumental values (self-direction and conformity) resulted in a model of acceptable fit (^c2=51.40, df=34, p=0.028; GFI=0.953, AGFI=0.924, RMSR=0.035). These results seem to indicate that unidimensionality has been achieved.

Since the establishment of acceptable unidimensionality alone is not sufficient to ensure the usefulness of a scale, the reliability of the scale should then be assessed (Gerbing and Anderson 1988). Both Cronbach's (1951) coefficient alpha and J÷reskog's (1971) construct reliability are computed in this study. Both the coefficient alpha and construct reliability are 0.83 for self-direction dimension and 0.87 for conformity dimension. These results indicate that both the four-item self-direction measure and the six-item conformity measure seem to be quite reliable.

After unidimensionality and reliability have been acceptably established, the convergent and discriminant validity should be assessed (Anderson and Gerbing 1988). Two approaches were used to assess convergent validity. One approach recommended by Anderson and Gerbing (1988) is to examine whether each indicator's pattern coefficient (loading) from the measurement model is statistically significant. Another approach is recommended by Fornell and Larcker (1981). According to this approach, the average variance extracted for each construct should be more than 0.5 in order to achieve convergent validity. As shown in Table 3, all the t tests for l's are statistically significant and the average variance extracted for self-direction and conformity are 0.56 and 0.53 respectively. These results suggest that convergent validity for both measures is achieved.

TABLE 4

MEASUREMENT MODEL RESULTS FOR SHOPPING BEHAVIOR SCALES

There are also two approaches that can be used to assess discriminant validity. Anderson and Gerbing (1988) suggest that discriminant validity can be assessed by constraining the phi value for a pair of constructs to unity and then estimating the resulting measurement model. In the current study, the ^c2 value for the constrained model is 148.02 with 35 degrees of freedom. Because the ^c2 difference between the constrained and nonconstrained models is 96.62, which for 1 degree of freedom is highly significant, discriminant validity seems to be indicated. Fornell and Larcker (1981) present another approach to assess discriminant validity. According to them, for any pair of constructs to demonstrate discriminant validity, the average variance extracted for each construct should be greater than the squared structural link (derived by LISREL) between the pair. As shown in Table 3, the average variances extracted for self-direction and for conformity are far greater than ^f21², indicating that discriminant validity is achieved.

Measurement Model for Shopping Involvement and Price Sensitivity. Following the same procedures as those for the measurement model of instrumental values, unidimensionality, reliability, convergent and discriminant validity were assessed. As shown in Table 4, the measurement model for shopping involvement and price sensitivity has a very good fit, indicating that unidimensionality is achieved.

Both coefficient alpha and construct reliability are 0.89 for shopping involvement and 0.90 for price sensitivity. These results indicate that both measures seem to be highly reliable. As shown in Table 4, all the t tests for l's are statistically significant and the average variances extracted for both measures are greater than 0.5, suggesting that convergent validity is achieved. The ^c2 value for the constrained model is 71.61 with 14 degrees of freedom. Because the ^c2 difference between the constrained and nonconstrained models is 57.87, which for 1 degree of freedom is highly significant, discriminant validity seems to be indicated. In addition, the average variances extracted for shopping involvement and for price sensitivity, as shown in Table 4, are far greater than ^f21², indicating also that discriminant validity is achieved.

Structural Equation Model for the Nomological Network. Figure 1 presents the hypothesized nomological network of instrumental values. Table 5 displays the results of the full structural equation model that simultaneously models measurement and structural relations. As summarized in Table 5, all the goodness of fit indices demonstrate that the data provide a good fit for the overall model. The gamma coefficients linking two dimensions of instrumental values and two shopping behavior related constructs are in the expected direction and statistically significant at the 0.05 level for two of the four paths, namely, g₁₁ and g₁₂. The path linking self-direction and price sensitivity (g₂₁) is significant if tested with one-tail t test at the 0.10 level. These results provide some support to the hypothesized nomological network implied in H₃ and H₄.

DISCUSSION AND CONCLUSIONS

Research on personal values has been gaining momentum during the past few years. Although considerable empirical research has examined various functions of values, the issue of how to measure values accurately remains unresolved. Following the "two-step" approach recommended by Anderson and Gerbing (1988), the present study assessed unidimensionality, reliability, convergent validity, discriminant validity, and nomological validity of the Rokeach instrumental values.

TABLE 5

STRUCTURAL EQUATION MODEL RESULTS FOR THE NOMOLOGICAL NETWORK

Results suggest that the 18 Rokeach instrumental values are not independent with one another. Rather, there seem to be two latent dimensions measured by the RVS. One dimension identified in the current study was a four-item measure of self-direction and the other dimension was a six-item measure of conformity. Results of confirmatory factor analysis via LISREL provided support for the unidimensionality, reliability, convergent and discriminant validity associated with these two dimensions. Results of the full structural equation model that simultaneously models measurement and structural relations lend some support to the nomological network of instrumental values. Specifically, conformity dimension was found to be linked positively to shopping involvement while self-direction dimension had a negative relationship with shopping involvement. Although the relationships between price sensitivity and the two dimensions of instrumental values were not statistically significant at the 0.05 level, they were in the hypothesized direction.

Results of the study indicate an need for researchers to examine measures of personal values in a more rigorous manner. Structural equations with latent variables, as demonstrated in the current study, provide an appropriate way of assessing multi-item measurement scales. We believe research on personal values will benefit from further development of more valid and comprehensive value instruments.

Several limitations that could be explored in future studies are worth noting. First, the present study was based on a relatively small and nonprobability sample in China. Hence, caution must be exercised in generalizing the findings beyond the sample used. Second, a modified RVS instrument was used in the study that requires rating rather than ranking. Such scales have been found to be more susceptible to response styles and social desirability effects (Alwin and Krosnick 1985; Kamakura and Mazzon 1991; Rokeach 1973). Finally, this study has investigated only the Rokeach instrumental values. It would be interesting for future research to examine the Rokeach terminal values and other measures of personal values such as Kahle's (1983) List of Values (LOV).

REFERENCES

Alwin, Duane F. and Jon A. Krosnick (1985), "The Measurement of Values in Surveys: A Comparison of Ratings and Rankings," Public Opinion Quarterly, 49 (Winter), 535-552.

Alwin, Duane F. and D.J. Jackson (1982), "The Statistical Analysis of Kohn's Measures of Parental Values," in K.G. J÷reskog and H. Wold, eds., Systems Under Indirect Observation: Causality, Structure and Prediction. Amsterdam: North-Holland, 197-223.

Anderson, James C. and David W. Gerbing (1988), "Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach," Psychological Bulletin, 103 (3), 411-423.

Beatty, Sharon E., Lynn R. Kahle, Pamela Homer, and Shekhar Misra (1985), "Alternative Measurement Approaches to Consumer Values: The List of Values and the Rokeach Value Survey," Psychology and Marketing, 2 (3), 181-200.

Bentler, P.M. and Douglas G. Bonett (1980), "Significance Test and Goodness of Fit in the Analysis of Covariance Structures," Psychological Bulletin, 88(3), 588-606.

Bond, Michael H. (1988), "Finding Universal Dimensions of Individual Variation in Multicultural Studies of Values: The Rokeach and Chinese Value Surveys," Journal of Personality and Social Psychology, 55 (6), 1009-1015.

Carman, J.M. (1977), "Values and Consumption Patterns: A Closed Loop," in H.K. Hunt, ed., Advances in Consumer Research, 5, 403-407.

Cronbach, Lee J. (1951), "Coefficient Alpha and the Internal Structure of Tests," Psychometrika, 16 (September), 297-334.

Crosby, Lawrence A., Mary Jo Bitner, James D. Gill (1990), "Organizational Structure of Values," Journal of Business Research, 20, 123-134.

Darden, W.R., O. Erdem, D.K. Darden, and T. Powell (1979), "Consumer Values and Shopping Orientations," in Proceedings of the Southwestern Marketing Association, San Antonio, TX.

Feather, N.T. and E.R. Peay (1975), "The Structure of Terminal and Instrumental Values: Dimensions and Clusters," Australian Journal of Psychology, 27, 151-164.

Fornell, Claes and David F. Larcker (1981), "Evaluating Structural Equation Models with Unobservable Variables and Measurement Error," Journal of Marketing Research, 18 (February), 39-50.

Gerbing, David W. and James C. Anderson (1988), "An Updated Paradigm for Scale Development Incorporating Unidimensionality and Its Assessment," Journal of Marketing Research, 25 (May), 186-192.

Grunert, Susanne C. and Gerhard Scherhorn (1990), "Consumer Values in West Germany: Underlying Dimensions and Cross-Cultural Comparison with North America," Journal of Business Research, 20, 97-107.

Homer, Pamela M. and Lynn R. Kahle (1988), "A Structural Equation Test of the Value-Attitude-Behavior Hierarchy," Journal of Personality and Social Psychology, 54 (4), 638-646.

J÷reskog, Karl G. (1971), "Statistical Analysis of Sets of Congeneric Tests," Psychometrika, 36 (June), 109-133.

Joreskog, Karl G. and Dag S÷rbom (1988), LISREL VII: Analysis of Linear Structural Relationships by Maximum Likelihood. Mooresville, IN: Scientific Software.

Kahle, Lynn R. (1983), Social Values and Social Change: Adaptation to Life in America, New York: Praeger.

Kamakura, Wagner A. and Jose A. Mazzon (1991), "Value Segmentation: A Model for the Measurement of Values and Value Systems," Journal of Consumer Research, 18 (September), 208-218.

Kluckhohn, F.R. and Strodbeck, F.L. (1961), Variations in Value Orientation, IL: Row, Paterson and Co.

Lefcourt, H.M. (1966), "Internal Versus External Control of Reinforcement: A Review," Psychological Bulletin, 65, 206-220.

Maloney, J. and G.M. Katz (1976), "Value Structures and Orientations to Social Institutions," Journal of Psychology, 93, 203-211.

McGuire, William J. (1974), "Psychological Motives and Communication Gratification," in J.G. Blumler and C. Katz, eds., The Use of Mass Communications. Beverly Hills, CA: Sage Publications, 167-196.

McQuarrie, Edward F. and Daniel Langmeyer (1985), "Using Values to Measure Attitudes Toward Discontinuous Innovations," Psychology & Marketing, 2 (Winter), 239-252.

Miethe, Terance D. (1985), "The Validity and Reliability of Value Measurements," Journal of Psychology, 119 (5), 441-453.

Miller, George A. (1956), "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information," Psychology Review 63 (March), 81-97.

Moore, Michael (1975), "Rating vs. Ranking in the Rokeach Value Survey: An Israeli Comparison," European Journal of Social Psychology, 5 (3), 405-408.

Munson, J. Michael and Shelby H. McIntyre (1979), "Developing Practical Procedures for the Measurement of Personal Values in Cross-Cultural Marketing," Journal of Marketing Research, 16 (February), 55-60.

Munson, J. Michael (1984), "Personal Values: Considerations on Their Measurement and Application to Five Areas of Research Inquiry," in Pitts, Robert E. Jr. and Arch G. Woodside ed., Personal Values and Consumer Psychology, Mass.: Lexington Books.

Munson, J. Michael and Edward F. McQuarrie (1988), "Shortening the Rokeach Value Survey for Use in Consumer Research, Advances in Consumer Research, 15, 381-386.

Novak, Thomas P. and Bruce MacEvoy (1990), "On Comparing Alternative Segmentation Schemes: The List of Values (LOV) and Values and Life Styles (VALS)," Journal of Consumer Research, 17 (June), 105-109.

Nunnally, Jum C. (1978), Psychometric Theory. New York: McGraw-Hill Book Company.

Petit, Kathy L., Sonja L. Sawa, Ghazi H. Sawa (1985), "Frugality: A Cross National Moderator of the Price-Quality Relationship," Psychology and Marketing, 2, 253-266.

Pitts, Robert E., Jr. and Arch G. Woodside, eds. (1984), Personal Values and Consumer Psychology, M.A.: Lexington Books.

Prakash, Ved and J. Michael Munson (1985), "Values, Expectations from the Marketing System and Product Expectations," Psychology & Marketing, 2 (Winter), 279-296.

Ralston, D.A., D.J. Gustafson, P.M. Elsass, F. Cheung, R.H. Terpstra (1992), "Eastern Values: A Comparison of Managers in the United States, Hong Kong, and the People's Republic of China," Journal of Applied Psychology, 77(5), 664-671.

Rankin, William L. and Joel W. Grube (1980), "A Comparison of Ranking and Rating Procedures for Value System Measurement," European Journal of Social Psychology, 10 (3), 233-246.

Rokeach, Milton (1973), The Nature of Human Values. New York: Free Press.

Schopphoven, Iris (1991), "Values and Consumption Patterns: A Comparison between Rural and Urban Consumers in Western Germany," European Journal of Marketing, 25 (12), 20-35.

Slama, Mark E. and Armen Tashchian (1985), "Selected Socio-Economic and Demographic Characteristics Associated with Purchasing Involvement," Journal of Marketing, 49 (Winter), 72-82.

Tucker, L.R. and C. Lewis (1973), "A Reliability Coefficient for Maximum Likelihood Factor Analysis," Psychometrika, 38, 1-10.

Valette-Florence, P. and A. Jolibert (1990), "Social Values, A.I.O., and Consumption Patterns: Exploratory Findings," Journal of Business Research, 20, 109-122.

Vinson, D.E., J.D. Scott, and L.M. Lamont (1977), "The Role of Personal Values in Marketing and Consumer Behavior," Journal of Marketing, 41, 44-50.

Yau, Oliver H.M. (1988), "Chinese Cultural Values: Their Dimensions and Marketing Implications," European Journal of Marketing, 22 (5), 44-57.

----------------------------------------