The Use of Social Psychological and Applied Value Research For the Measurement of Cultural Differences Among Consumers

ABSTRACT - The paper presents a two-dimensional value system centred around modern traditional and individual versus socially orientation. It is shown how the dimensions are derived and how they relate to other comprehensive value systems such as VALS, RISK, CCA, and the like. A simple AKompas@ value chart description of the findings is introduced and it is shown how different the values discriminate better than traditional sociodemographic criteria, and the possible application of the findings across countries and over time is discussed.

Citation:

Camilla P. Christiansen and Flemming Hansen (2001) ,"The Use of Social Psychological and Applied Value Research For the Measurement of Cultural Differences Among Consumers", in AP - Asia Pacific Advances in Consumer Research Volume 4, eds. Paula M. Tidwell and Thomas E. Muller, Provo, UT : Association for Consumer Research, Pages: 5-22.

Asia Pacific Advances in Consumer Research Volume 4, 2001      Pages 5-22

THE USE OF SOCIAL PSYCHOLOGICAL AND APPLIED VALUE RESEARCH FOR THE MEASUREMENT OF CULTURAL DIFFERENCES AMONG CONSUMERS

Camilla P. Christiansen, Copenhagen Business School, Denmark

Flemming Hansen, Copenhagen Business School, Denmark

ABSTRACT -

The paper presents a two-dimensional value system centred around modern traditional and individual versus socially orientation. It is shown how the dimensions are derived and how they relate to other comprehensive value systems such as VALS, RISK, CCA, and the like. A simple "Kompas" value chart description of the findings is introduced and it is shown how different the values discriminate better than traditional sociodemographic criteria, and the possible application of the findings across countries and over time is discussed.

Following this the Schwartz Value Domains are introduced. The stability of the Value Domains between more than 40 countries are stressed and the nature of the Domains are explained. Finally it is suggested that the Kompas and in the Scwartz system, and it is suggested that the Kompas dimensions may be interpreted in terms of the Schwartz Value Dimensions.

The comparison focuses on three issues: (1) the degree f similarity of the two instruments as regards their underlying structure, (2) the potential of the two instruments for classifying respondents in a similar way, and (3) the predictive power of the two instruments for a number of marketing variables.

VALUE RESEARCH IN SOCIAL PSYCHOLOGY

The number of definitions and conceptualisations of values are legion. The smallest common denominator would be to interpret values as those patterns by which individuals orientate themselves in and adapt to their environment. These patterns were described by Tolman (1951) as basic conceptions about life which underlie an individual’s behaviour. Values are both self-centred and social-centred in the sense that they are at the crossroads between the individual and the society: As general orientation standards, they include external, social-centred aspects insofar as they are effective as guiding principles established by the social environment. They also include internal, self-centred aspects when they are standards internalised and accepted by the individual. Rokeach, who promoted the idea of values as influencing all human life domains, viewed a value as a centrally held, enduring belief which guides actions and judgements across specific situations and beyond immediate goals to more ultimate end states of existence (Rokeach, 1968), p. 160). These beliefs are regarded as the cognitive-emotive representations and transformations of needs. Values thus provide the goals towards which behaviour is motivated, their immediate function is to guide human action in daily situations (Rokeach, 1973).

Generally speaking, five main features of values can be outlined, which are considered as constituting the common background for social science research on values (Schwartz & Bilsky, 1987). Values are: (1) concepts or beliefs (2) about desirable end states or behaviours (3) that transcend specific situations, (4) guide the selection or evaluation of persons, behaviour and events, and (5) are ordered by relative importance.

Besides these formal characteristics of values, their meaningful content can be defined as cognitive-emotive representations of three types of universal human requirements: Biologically based needs of the organism, social interactional claims for interpersonal co-ordination, and social institutional demands for group welfare and survival (cf., Kluckhohn, 1951; Maslow, 1959). Hence, values serve both individualistic and collectivist interests as well as a mixture of these (cf., Hofstede & Bond, 1984; Triandis, 1985).

The abstractness of the value concept and its main features makes it especially attractive for cross-cultural and multicultural research. Cultural differences in specific behaviours have been suggested to be more easily explained by referring to the more abstract levels of the cognitive-emotive hierarchy (Grunert, Grunert, & Beatty, 1989).

The assessment of value systems in psychology has occurred to a minor extent by projective techniques, content analysis, therapeutic dialogue, participant observation, and pictorial stimuli (Grunert-Beckmann & Askegaard, 1997; Johnson & Bolstad, 1973; Pittel & Mendelsohn, 1966). Much more often, so-called paper-and-pencil tests have been used because they combine several advantages. In general, they can be used with larger groups or representative samples, allow the comparison of groups and cohorts over time, can be linked with other measures, and simplify the control of validity and reliability of the measurement process.

Many scales have been developed for measuring respondents’ value systems; more than 20 are well known. One of these is the Study of Values (Vernon & Allport, 1931) which has been used in several 100 empirical studies. Another instrument is the Rokeach Value Survey RVS (Rokeach, 1973), where respondents are asked to rank 36 values, 18 instrumental and 18 terminal values. In marketing, Values and Lifestyes VALS (Mitchell, 1983) has been used frequently, which employs value-related statements to be rated on Likert-type scales, as well as the List of Values LOV, which has been developed at the Survey Research Center of the University of Michigan (Kahle, 1983). The latter has mainly been used in a consumer behaviour context (e.g., (Homer & Kahle, 1988), both in national and cross-cultural studies (Grunert & Muller, 1996; Grunert & Scherhorn, 1990; Kahle, Beatty, & Homer, 1989). Other methods include free associations (GSrling, Lindberg, Montgomery, & Waara, 1985; Montgomery, Drottz, GSrling, Persson, & Waara, 1985), which led to more than 150 different values; paired comparison procedures of shorter versions of the RVS (Munson & McQuarrie, 1988); and value lists with a dichotomous categorisation of agreement vs. disagreement (RaffTe & Wiedmann, 1987).

FIGURE 1

More recently, the Schwartz Value Inventory SVI (Schwartz, 1992) has been introduced to consumer behaviour research by Grunert and Juhl (1995) and Puohiniemi (1995). This inventory has been developed on the basis of several decades of theoretical and empirical work in psychology research on values. Its theoretical starting point is to view values as representing motivations, because they are criteria used by individuals to select and justify actions, and to evaluate people, the self, and events. These criteria are supposed to differ in content as well as in their structural relations to each other, an assumption that hinges on the above mentioned formal characteristics and meaningful contents of values. Ten motivational domains of values are suggested, postulating that these domains are culturally universal in their content and structure. These domains reflect either an individualistic interest dimension, or a collectivist, or both, and they can be grouped into four higher-order dimensions:

1. Openness to change

$ Self-direction (SDI, individualistic)

$ Stimulation (STI, individualistic)

$ Hedonism (HED, individualistic)

2. Self-enhancement

$ Achievement (ACH, individualistic)

$ Power (POW, individualistic)

3. Conservation

$ Security (SEC, individualistic and collectivist)

$ Conformity (CON, collectivist)

$ Tradition (TRA, collectivist)

4. Self-transcendence

$ Benevolence (BEN, collectivist)

$ Universalism (UNI, individualistic and collectivist)

Since individuals experience values as relevant when making choices that have implications for the expression or attainment of their goals, they will also experience conflicts when it comes to choices that have mutually contradictory implications for the realisation of some of their values. The pursuit of personal success (achievement) is likely to interfere with the adherence to the preservation and enhancement of the welfare of other people (benevolence), thus forcing individuals to make choices among competing value categories. Hence, in Figure 1, this structure of value priorities is illustrated as follows: adjacent value types are most compatible, whereas an increasing distance around the circular order indicates a decreasing compatibility, and therefore an increasing conflict. Value types that emerge in opposing directions from the origin should be in greatest conflict. The total value structure can be summarised in terms of a two-dimensional structure composed of four higher-order value dimensions that are basic and bipolar.

APPLIED VALUE RESEARCH

Commercial value research has resulted in a number of measurement instruments such as VALS, RISK, Minerva, Kompas and CCA. Based on a large number of questions often related to life-style issues, activities, interests, and opinions, these instruments are usually reduced to representing two basic and bi-polar dimensions of and underlying value structure (see Table 1; Hansen, 199).

"Kompas" by Gallup A/S Denmark is an example for this type of value measurement instrument. In its present form, Gallup Kompas was developed in Denmark in the mid 90’ies in connection with the construction of the Danish target group media index "Index Danmark/Gallup," which is presently the dominating media planning system in Denmark, and which is now being introduced in countries such as Poland, Israel and Sweden.

"Index/Danmark" is based on a large media/marketing database including values, opinions, and interests. Approximately 1500 items appear in the basic questionnaire. Also included are questions on brands, possessions of durables, product use, shopping habits, and the like. Data collection for "Index/Danmark" is based upon more than 50.000 annual CATI interviews. A subset of the respondents from that sample (approx. 18.000) are completing a self-administered questionnaire including the 1500 measures, this sample is again divided into 4 quarters, each of which are also answering a subset of an additional total of 4000 brand and product questions.

"Kompas" supplements "Index/Danmark." Its purpose is to illustrate the distribution of basic values in the population. Any model is a simplification of the reality, intended to offer understanding of basic tendencies, similarities, and differences. This is also the "raison d’Otre" and the limitation of the "Kompas" model. "Kompas" gives a very valuable insight and overview, but it is not believed to fit all situations. There will be markets and cases where it is appropriate to produce more specific models as supplements to the basic "Kompas" model.

The system is based on a two-step factor-analytical procedure resulting in two basic dimensions, which are constructed on the basis of 22 questions. The dimensions reflect "modernism" versus "traditionalism" and "social" versus "individual" orientation. Each respondent is given a score on each of the two dimensions.

TABLE 1

BASIC DIMENSIONS IN COMMERCIAL VALUE MEASUREMENT INSTRUMENTS

In the "Kompas" questionnaire, values are measured on 6-point scales, interests (in various behaviours, products, etc.) on 5-point scales, and activities on a pre-coded scale ranging from daily, more than 5-6 times a week, 3-4 times a week, once or twice a week, less frequent to never.

During the process of developing the "Kompas" dimensions, analyses were carried out on attitudes, interests, and activities separately, and on the three groups of measures combined. It soon became evident that the solutions that emerged were quite similar regardless of the battery used, and, when the three batteries were combined, the value questions were by far the most discriminating items, thus explaining most variance. Consequently it was decided to derive the solution based upon the value items alone. In this process, the initial total number of 120 values was reduced to 80, later to 60, by eliminating obvious duplicates, items that could be misinterpreted, and items with low discriminatory power. With these 60 items two dimensions were found that explained a large proportion of the total variance.

For the purpose of simplifying future analysis of the data it was desirable to reduce the number of variables from the initial 60 ones. This was done by concentrating on those items relating most strongly to the two first factors in the initial solution and by focusing on those with the highest discriminatory power. Thus items were excluded where respondents tended all to agree or to disagree. This process led to a 22-item battery based upon which the two dimensions can be identified. This battery has so far been kept unchanged in the Danish questionnaire from 1994 to 1998. In other countries, where "Kompas" is used, modifications were made. The presentation here is based primarily on results from 1997 data from Denmark

At this point it should be mentioned, that a further reduction has been made in the number of questions necessary to identify the dimensions. Here a total of eight questions have been selected. This is done to make the instrument applicable in ad-hoc projects withut having to overload the questionnaire with too many value items. The eight items are selected from the 22 ones with four relating to each of the two dimensions: Two strongly positively loaded, two negatively loaded for each of them. This reduced battery, it turned out, reproduces the grouping of the respondents based upon the larger questionnaire very well. Actually, in 1997, 98 per cent of the respondents were grouped identically when the reduced battery was used.

In the construction of the model, one of the most important decisions to be made is to choose the number of groups the population is divided into. There is always a difficult balance between realismBas many groups as possibleBand simplicityBas few groups as possible. This is done by disregarding the more or less neutral centre and by dividing the remaining respondents into eight pie slice segments. Each respondent has a score on each of the two dimensions on a number indicating the segment where s/he belongs to. For each of the segments rich and colourful descriptions and pictorial representations were developed.

$Modern

$Modern individually oriented

$Individually oriented

$Traditional individually oriented

$Traditional

$Traditional socially oriented

$Socially oriented

$Modern socially oriented

Gallup Kompas is primarily a tool for positioning brands, products, and consumers in a diagram like the one for male users of Harley Davidson deodorant in Denmark in Figure 2. Here the colour and the sizes of the pie slices signify the relative importance of the particular segments. On the average 8% of the target group use the brand. However, among the modern, individually oriented consumers 23.8% do so, and among the traditional, socially oriented consumers only 2% do so.

DIFFERENCES BETWEEN SVI AND "KOMPAS"

The two instruments that have been described in the preceding sections differ in several ways:

1. The SVI works directly with value items such as "social justice," "authority" and "wealth" that are grouped into the ten motivational domains. The "Kompas" dimensions are based on day-to-day opinion statements like, e.g., "I am against too many immigrants in my community."

2. The 57 value items used for identifying respondents’ location in the SVI mapping seem to require a certain level of intellectual capacity of respondents, since they are measuring rather abstract terms with an anchoring technique. The "Kompas" statements are presented to respondents as opinions about a number of issues.

3. The SVI items were derived through a thorough study of psychological value theories, and there are both theoretical and empirical grounds for claiming that they cover values central and relevant to describing social character. The "Kompas" instrument is like other commercial value systems basically derived from empirical evidence.

4. The majority of the experiences gained with the SVI are based upon data collected from predominantly professional samples, such as students, college teachers and other professionals, only a few studies have been carried out with the general population (Beckmann, Christensen, & Kilbourne, in press). The "Kompas" Value Sysem is based upon probability samples of the general population with controlled weighting for non-response.

In spite of this each of the two approaches seem to divide individuals along two dimensions with the two bipolar dimensions appearing very similar. It is the purpose of the study reported in this paper here to shed further light on the nature of the relationship between the two approaches.

FIGURE 2

MALE USERS OF HARLEY DAVIDSON DEODORANT

THE STUDY

It has recently been suggested that the values measured by the SVI and "Kompas" resemble each other, thus leading to a structural similarity (Hansen, 1997). This proposition was tested in this study.

Sample and data collection

The study was carried out in Denmark with respondents who had completed the "Kompas" questionnaire and other standard questions in the Gallup Marketing Index for the third quarter of 1997. A quota sample of 350 respondents was drawn from that sample and asked to fill out the SVI. This procedure led to a database comprising:

$AKompas" values

$SVI values

$Marketing variables

Analyses

The starting point for the analyses was the SVI. First, the dimensionality of the SVI was assessed through factor analysis and MDS, followed by correlation and factor analyses aiming at reducing the 57 items to 20 that would cover the ten motivational domains. The "Kompas" items were also analysed using factor analysis and MDS. Second, "Kompas" and SVI items were pooled and analysed using MDS and the items of the two systems are compared individually.

Analyses I: The Schwartz Value Inventory SVI

The multidimensional scaling of the 57 value items (Figure 3) showed a solution that was quite in line with the theoretical model (see Fig. 1), especially when taking into account the four higher-order dimensions. Because of the large number of items it is however difficult to overlook the total diagramme.

For further analysis, therefore, it seems preferable to reduce the number of items to a more parsimonious size. The reduction to 20 items, with two representing each of the ten motivational domains, was achieved by a combination of correlation and factor analysis. First, all 57 items were correlated with each other and only items that correlated with r>.400 and within their respective higher-order domain were selected for factor and MDS analyses. The final solution is shown in Figure 4.

FIGURE 3

MDS SOLUTION OF THE 57 ITEMS IN THE SVI (STRESS=.1638, RSQ=.9199)

FIGURE 4

MDS SOLUTION OF THE SELECTED 20 ITEMS IN THE SVI (STREE=.1626, RSQ=.9568)

TABLE 2

FACTOR SOLUTIONS OF "KOMPAS" 1996-1998

Analyses II: "Kompas"

The test database consists of 350 respondents all having completed the full Marketing Index Questionnaire with the 113 attitude questions from which 22 are used for the computation of the "Kompas" scores. There is not much reason to expect that the "Kompas" solution based on this sample should deviate from the original "Kompas" solution based on larger samples. However, the fact that non-response occurs partly because some respondents refuse to complete the more complex Schwartz questionnaire, and partly because of people returning incomplete questionnaires may still affect the scores.

To test whether this is the case, the traditional "Kompas" two-dimensional varimax rotated factor analytical solution was computed. The factor loadings for the 22 questions are shown in the last column of Table 2. In the firstcolumns the loadings are shown from the 1996, the 1997 and the 1998 databases. It appears that the test database reproduces the original factor loadings remarkably well. Hence, there is no doubt that the computed scores for the respondents on the two dimensions: modern-traditional and social-individual orientation fully represent the same scores as they appear in the larger, national databases for the years 1996B1998.

The "Kompas" items were then submitted to a MDS, using 20 out of the 22 statements (i.e., ignoring the two items representing a mixture of traditional-individual). The results are shown in Figure 5. I can be seen that the two dimensions traditional-modern and individually-socially orientation reappears if this unrotated solution is turned 45H.

Analyses III: SVI and "Kompas"

Based on the theoretically derived four higher-order motivational dimensions of the SVI values and the empirically derived four "Kompas" dimensions, and looking at fig. 4 and 5 the following assumptions about their relationships were made:

$SVI / Self-transcendence (UNI, BEN) & "Kompas" Traditional, "Kompas" Collectivism

$SVI / Self-enhancement (ACH, POW) & "Kompas" Individualism, "Kompas" Modernity

$SVI / Conservationism (CON, TRA,SEC) & "Kompas" Traditional, "Kompas" Individualism

$SVI / Openness to change (SDI, STI, HED) & "Kompas" Modernity

To test whether these assumptions hold, correlation analysis was conducted. The results are shown in Tables 3 to 6 and confirm, although to a varying degree, the assumptions (note that only coefficients significant at the 0.05 level or below are reported).

FIGURE 5

MDS OF 20 "KOMPAS" ITEMS (STRESS=.1687, RSQ=.8616)

Based on these findings, the 20 SVI items and the 20 "Kompas" items were pooled and entered in a MDS analysis. The result is shown in Figure 6.

The comparison can also be carried out in a reversed way.

In Figure 7 a standard diagram is shown for the average value of a typical SVI item: #true friendship’. It can be seen that this basic value appears to be present in very different degrees in the Kompas segments. It is very expressed among the traditional and traditional individually oriented respondents and much less with the modern, modern socially oriented respondents. That is, the traditional (individually) oriented consider true friendship as being much more important than the modern (socially oriented) respondents.

To analyse the data, 57 such diagrams can be constructed. To do so and to look at these would be rather tedious and patterns would not easily be detected. For this reason a slightly different approach has been chosen. A data matrix of 20 by 9 average scores has been established. This is reproduced in Exhibit 1. In each cell of this matrix the average value score among respondents belonging to a particular Kompas segment is computed. In the first cell of the table it can be seen that there are respondents who have answered the question about #loyal’ with an average of 7.11. In the second last column of the exhibit it can be seen that this score is very similar to that for the center group. Also the value of #loyalty’ is most expressed among the traditional socially oriented ones.

To summarise the findings of Exhibit 1, table 7 has been set up. Here the 20 SVI statements are ordered in clockwise order (see fig. 1) and the average Kompas Score for the segment most closely corresponding with the SVI domains is shown together with the average score for all segments on the SVI questions analysed.

Thus the two systems relate at a micro level as well as at a macro level. At the micro level the individual Schwartz items meaningfully describe or complement the existing descriptions of the Kompas Value segments. At the macro level the overall imensions of the to systems relate meaningfully and domains of the Schwartz system are related to corresponding segments in the Kompas system.

It is possible to study these relations in more detail and to identify those more structural differences the comparison in Table 7 also suggests. Here the relatively limited number of observations calls for more advanced statistical analyses of the data.

TABLE 3

SVI / SELF-TRANCENDENCE & "KOMPAS" TRADITIONAL, "KOMPAS" COLLECTIVISM

TABLE 4

SVI / SELF-ENHANCEMENT & "KOMPAS" INDIVIDUALISM, "KOMPAS" MODERNITY

TABLE 5

SVI / OPENNESS TO CHANGE & "KOMPAS" MODERNITY

TABLE 6

SVI / CONSERVATIONISM & "KOMPAS" TRADITIONAL, "KOMPAS" INDIVIDUALISM

FIGURE 6

MDS OF POOLED SVI AND "KOMPAS" ITEMS (STRESS=.1887, RSQ=.8971)

FIGURE 7

KOMPAS DIAGRAM FOR THE SCHWARTZ VALUE QUESTIONS 'TRUE FRIENDSHIP'

TABLE 7

20 SVI ITEMS AND THEIR AVERAGE SCORE AND THEIR SCORE ON THE 'TARGET' KOMPAS SEGMENTS TO WHICH THE STATEMENTS ARE HYPOTISED TO BELONG

EXHIBIT 1

EXHIBIT 1 (CONT.)

CONCLUSIONS

1. The Kompas System composed of 8 equally large segments is demonstrated to be valid, stable, reliable, and sensitive in studies of consumption and media behaviour.

2. The Schwartz Value Domain System, an internationally accepted and tested value system, is introduced based upon which meaningful comparisons can be made across borders and within groups.

3. The empirically based Kompas System, working with derived values, reflects the same kind of value structure as the theoretically and on more basic values base Schwartz Value Domain System.

4. In addition to the international description of the Kompas Value Segments based upon behaviour, attitudes, opinions and socio/demographic variables the segments can meaningfully be described in terms of the single items in the Schwartz Value battery and in terms of the more aggregated domains of the same theory.

5. The stability of the Schwartz values domains being well documented in research across borders suggests that one can meaningfully look for the two Kompas dimensions also across borders.

6. Experiences with value domains and local variations in these may provide useful guidance in adapting the Kompas measurement system to different cultures.

7. In contexts where intellectual abilities of respondents or others reasons makes it difficult to work with the somewhat demanding value domain questions, the more straightforward Kompas measurement approach may substitute.

The present presentation is the first introduction to an approach relating the study of consumer values to theoretical social character approaches from social psychology and sociology. Much more insight is to be gained. The data used here are reported in more detail elsewhere and future studies with the use of the Schwartz Value Domain System, the Kompas System or the two together may give us still further insight in the more basic values underlying the pragmatic value dimensions used in most consumer segmentation and media research.

Such a theoretical foundation is badly needed In spite of what is sometimes otherwise claimed all values systems in contemporary commercial use rely upon more or less the same questions. Any contemporary value questionnaire contains a considerable number of questions going back to the original Mitchell 81983) and the original Leo Burnet and other style questionaires (Wells & Tigert, 1971).

REFERENCES

Beckmann, S. C., Christensen, A. G., & Kilbourne, W. E. (in press). Contemporary world-views and environmental concern: Comparing different population segments in Denmark (CEC Working paper no. 10). Copenhagen: Copenhagen Business School, Department of Marketing, Research group "Consumption, environment & culture".

GSrling, T., Lindberg, E., Montgomery, H., & Waara, R. (1985). Beliefs about the attainment of life values (Umes Psychological Reorts 181): University of Umes, Dept. of Psychology.

Grunert, K. G., Grunert, S. C., & Beatty, S. E. (1989). Cross-cultural research on human values. Marketing and Research Today, 17(1), 30-39.

Grunert, S. C., & Juhl, H. J. (1995). Values, environmental attitudes, and buying of organic foods. Journal of Economic Psychology, 16(1), 39-62.

Grunert, S. C., & Muller, T. E. (1996). Measuring values in international settings: Are respondents thinking "real" life or "ideal" life? In L. A. Manrai & A. K. Manrai (Eds.), Global perspectives in cross-cultural and cross-national consumer research (pp. 169-186). New York: The Haworth Press.

Grunert, S. C., & Scherhorn, G. (1990). Consumer values in West Germany: Underlying dimensions and cross-cultural comparison with North America. Journal of Business Research, 20(2), 97-108.

Grunert-Beckmann, S. C., & Askegaard, S. (Eds.). (1997). "Seeing with the mind’s eye": An exploratory study of the use of pictorial stimuli in value research. Hillsdale, NJ: Lawrence Erlbaum.

Hansen, F. (1997). From Life Style to Value Systems to Simplicity. In J. W. Alba & J. W. Hutchinson (Eds.), Advances in Consumer Research (Vol. 25, pp. 181-195). Provo, UT: Association for Consumer Research.

Hofstede, G., & Bond, M. H. (1984). Hofstede’s culture dimensions: An independent validation using Rokeach’s Value survey. Journal of Cross-Cultural Psychology, 15, 417-433.

Homer, P. M., & Kahle, L. R. (1988). A structural equation test of the value-attitude-behavior hierarchy. Journal of Personality and Social Psychology, 54(4), 638-646.

Johnson, S. M., & Bolstad, O. D. (1973). Methodologic issues in naturalistic observation: Some problems and solutions for field research. In L. A. Hamerlynch, L. C. Handy, & E. J. Mash (Eds.), Behavior change: Methodology, concepts, and practice . Champaign, IL: Research Press.

Kahle, L. R., Beatty, S. E., & Homer, P. (1989). Consumer values in Norway and the United States: A comparison. Journal of International Consumer Marketing, 1(4), 81-92.

Kahle, L. R. E. (1983). Social values and social change: Adaptation to life in America. New York: Praeger.

Kluckhohn, C. (1951). Values and value-orientations in the theory of action: An exploration in definition and classification. In T. Parsons & E. A. Shils (Eds.), Toward a general theory of action (pp. 388-433). New York: Harper & Row.

Maslow, A. H. (Ed.). (1959). New knowledge in human values. New York, NY: Harper.

Mitchell, A. (1983). Nine American lifestyles: Who we are and where we are going. New York, NY: MacMillan.

Montgomery, H., Drottz, B.-M., GSrling, T., Persson, A.-L., & Waara, R. (1985). Conceptions about material and immaterial values in a sample of Swedish subjects. In H. BrandstStter & E. Kirchler (Eds.), Economic Psychology, Proceedings of the 10th IAREP Annual Colloquium (pp. 427-437). Linz: Trauner.

Munson, J. M., & McQuarrie, E. F. (1988). Shortening the Rokeach Value Survey for use in consumer research. In M. J. Houston (Ed.), Advances in consumer research (Vol. 15, pp. 381-386). Provo, UT: Association for Consumer Research.

Pittel, S. M., & Mendelsohn, G. A. (1966). Measurement of moral values. A review and critique. Psychological Bulletin, 66, 22-35.

Puohiniemi, M. (1995). Values, consumer attitudes and behaviour. An application of Schwartz’s value theory to the analysis of consumer behaviour and attitudes in two national samples. Helsinki: University of Helsinki, Department of Social Psychology.

Raffee, H., & Wiedmann, K.-P. (1987). Dialoge 2: Konsequenzen fnr das Marketing. Hamburg: Gruner & Jahr.

Rokeach, M. J. (1968). Beliefs, attitudes, and values. San Francisco, CA: Jossey Bass.

Rokeach, M. J. (1973). The nature of human value. New York, NY: Free Press.

Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advance and empirical tests in 20 countries. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 1-65). Orlando, FL: Academic Press.

Schwartz, S. H., & Bilsky, W. (1987). Toward a universal psychological structure of human values. Journal of Personality and Social Psychology, 53, 550-562.

Tolman, E. C. (1951). A psychological model. In T. Parsons & E. A. Shils (Eds.), Toward a general theory of action (pp. 279-364). New York: Harper & Row.

Triandis, H. C. (1985). Collectivism vs. individualism: A reconceptualization of a basic concept in cross-cultural psychology. In C. Bagley & G. C. Verma (Eds.), Personality, cognition, and value: Cross-cultural perspectives of childhood and adolescence . London: MacMillan.

Vernon, P., & Allport, G. W. (1931). A test for personal values. Journal of Abnormal and Social Psychology, 26, 231-248.

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