Advances in Consumer Research Volume 11, 1984 Pages 285-290
CONSUMER ETHNOCENTRISM: THE CONCEPT AND A PRELIMINARY EMPIRICAL TEST
Terence A. Shimp, University of South Carolina
Casual empiricism reveals considerable variability in the attitudes and behaviors of consumers toward products imported into the U.S. from other countries. Increasingly greater number of American consumers have adopted foreign-made produces, but an unknown number of consumers eschew the purchase of these imported products. The specific motivations underlying either form of behavior remain a virtually unexplored phenomenon in consumer behavior. This study addresses this issue by introducing the concept of "consumer ethnocentrism" and presenting empirical evidence from a preliminary test of the concept.
The general concept of "ethnocentrism" was introduced and used descriptively by Sumner in 1906 (Adorno et al. 1950) and remains a venerable concept in the fields of anthropology, sociology, and social psychology. Ethnocentrism was originally conceptualized as a purely sociological concept that distinguished between ingroups (those groups with which an individual identifies) and outgroups (those regarded as antithetical to the ingroups). It now, however, is recognized that ethnocentrism is a psycho-social phenomenon with relevance to individual-level personality systems as well as to the more general cultural and social analytic frameworks (Levine and Campbell 1972). The present undertaking is most sympathetic to the individual-level aspect of the multi-faceted notion of ethnocentrism. The concept of "consumer ethnocentrism" is designed to capture individual consumer cognitions and emotions as they relate to product offerings from other countries (i.e., "outgroups"). In the fullest sense of ethnocentrism, these product symbol from other countries may represent objects of contempt to the ethnocentric consumer. whereas the products of one's own national group are objects of pride and attachment (cf. Levine and Campbell 1972).
The general applicability of ethnocentrism to the study of consumer behavior has been recognized by others (e.g., Berkman and Gilson 1978; Markin 1974). There, however, is no known work that has reconstructed the concept to be more finely suited to the specific study of consumer behavior. The term "consumer ethnocentrism" labels a concept that captures the spirit of Jacoby's (1978) plea for domain-specific constructs. The concept is used here to represent consumers' beliefs in the superiority of their own country's products. This perception is postulated to transcend mere economic and functional considerations, and, instead, to have a more noble foundation rooted in morality. That is, consumer ethnocentrism is intended to capture the notion that some consumers believe it is somehow wrong to purchase foreign-made products, because it will hurt the domestic economy, cause the loss of jobs, and, in short, because, from their point of view, it is plainly unpatriotic.
The consumer ethnocentrism concept is postulated to be one component of a complex, multifaceted construct involving consumers, cognitive, affective, and normative orientations toward foreign-made products. The domain of this general construct spans object-based beliefs and attitudes (i.e., perceptions of product quality, value, etc.), normative-based beliefs and attitudes (i.e., perceptions of whether one should or should not purchase foreign-made products), and personalistic-based considerations of what mode of behavior (product choice) is in the consumer's best personal interest. Consumer ethnocentrism, as a dimension of this general consumer-orientation-toward-foreign-products construct, is designed to capture normative-based beliefs that buying domestic products is somehow good for the country, whereas purchasing non-domestic products is deleterious to the economy, the country, and to fellow citizens.
At the individual consumer level, ethnocentric tendencies should be determined in large part from one's socialization experiences. The family unit would be expected to be the primary socialization agent, but adult opinion leaders, peers, and mass media would also influence a child's ethnocentric orientation during the prime period of early childhood socialization. Consumer ethnocentric orientation would perform a function very much like those performed by racial attitudes and religious dogma--namely, in provide the child with a sense of identity and belongingness. This orientation, once formed, should carry into adulthood with few changes except in those instances where one's socialization influences are markedly altered
Though no one has previously formulated a specific concept such as this, consumer behavior researchers have studied ethnocentrism if only indirectly. Research has examined such issues as attitudes toward foreign products, biased consumer perceptions, and related matters (e.g., Anderson and Cunningham 1972; Bannister and Saunders 1978; Etzel and Walker 1974; Gaedeke 1973; Rierson 1966; Schooler 1971; Schooler and Wildt 1968).
Who are consumer ethnocentrics, what are their identifying characteristics, and how do they differ from non-ethnocentrics in their perceptions, evaluations, intentions, and behavior with respect to foreign-made products? This research was undertaken to address these questions. The automobile was selected as the focal product category, because it may be imbued more than most any other in U.S. society with cultural and symbolic significance.
VARIABLES AND GENERAL HYPOTHESIS
The general hypothesis tested was that consumer ethnocentrics have less favorable perceptions and evaluations of foreign autos, that they perceive important referents not to favor their purchasing foreign cars, and that they are less likely than non-ethnocentrics to own or to intend to purchase foreign-made automobiles. A number of criterion variables were measured to test for specific differences between consumer ethnocentrics and non-ethnocentric;, Included were all the standard Fishbein measures (Ajzen and Fishbein 1980; Fishbein and Ajzen 1975), the consumption-specific measures designed by Warshaw (1980), and other measures idiosyncratic to the study of foreign cars. Various independent variables were also measured Research has revealed that attitudes toward foreign products are related to age, sex, and socio-economic characteristics (Adorno et al. 1950; Bannister and Saunders 1978; Schooler 1971), thus offering rationale in the present research for examining these related variables
Data were collected in waves from a consumer panel located in the two Carolinas Responses were obtained at one wave or another from nearly 1200 households, though data are presented only for the subsample of 863 who provided responses to the consumer ethnocentrism measure This sample is older (mean age 47+) and somewhat upscale in socio-economic characteristics Though unrepresentative of the population in the true sense of the word, this alone does not diminish the generalizability of the study results. Indeed, upscale consumers actually are more representative of the automobile buying public, as many lower SES consumers have literally been Priced out of the new automobile market.
The initial wave of data collection obtained measures on a variety of demographic and socio-economic variables, including measurement of social class using Hollingshead's (1949) education and occupation factors, which were coded and combined into five social class categories. The most extensive data collection occurred at a later wave, which obtained the Fishbein and Warshaw measures alluded to above and also the measure of consumer ethnocentrism. Operationalizations of value expectancy and subjective norm items followed the technique devised by Oliver and Berger (1979) Several points are pertinent regarding these measurements First, the questionnaire adhered to the recommendation that measurement be specific with regard to action, target, context, and time (Ajzen and Fishbein 1980). This was accomplished with the instruction: "In this section of the questionnaire a variety of questions are asked about your purchasing a new automobile within the next year or so. Some of the questions refer to foreign-made cars. Interpret this as meaning standard Japanese or German cars (for example, Honda Accord, Toyota Corona, VW Rabbit) rather than luxury foreign cars such as Mercedes or Rolls Royce."
The choice of behavioral consequences and referents to serve as value expectancy and subjective norm constituents was made on the basis of personal observation augmented with precedents from prior automobile research (e.g., Holbrook 1978; Mazis and Ahtola 1975; Vinson, Scott, and Lamont 1977). Nine specific consequences associated with purchasing a foreign car were selected: fuel economy, car payments, dependability, status, safety, workmanship quality, ease of serviceability, appearance, and innovativeness of engineering. Four referent groups were chosen: friends and neighbors, spouse, family other than spouse, and work associates. Motivation to comply measures associated with these referents were situation-specific rather than general. Respondents were asked whether others' opinions mattered as they concerned the choice of a car, not whether their opinions mattered in some general, unspecified behavioral domain (see Miniard and Cohen 1979, 1981 for elaboration).
Other Fishbein constructs (i.e., Aact, SN, and BI) were measured using accepted procedures. Warshaw's more consumption-specific constructs also were operationalized. These included measures of the relative affordability of foreign-made cars and relative desirability (excluding affordability considerations). All measures adhered to the procedures suggested by Warshaw (1980).
MEASUREMENT OF CONSUMER ETHNOCENTRISM
Operationalizing consumer ethnocentrism was complicated by the absence of a suitable measurement scale. The famous F and E scales constructed by Adorno et al. (1950) to measure the general construct of ethnocentrism and related constructs were considered inappropriate for the present research because of excessive generality and datedness. More recent ethnocentrism scales have been developed, but these are domain specific and too specialized for the present undertaking (cf. Chang and Ritter 1976; Warr, Faust, and Harrison 1967). Failure to locate an acceptable ethnocentricity scale necessitated a specialized measure. An open-ended question was designed: "Please describe your views of whether it is right and appropriate for American consumers to purchase products that are manufactured in foreign countries." The question was posed in a projective fashion ("American consumers") and was directed at respondents' perceptions of the morality of purchasing foreign-made products ("is it right and appropriate?"). The choice of an open-ended question is justified, as the issue of consumer ethnocentrism is complex and the relevant dimensions are unknown (cf. Selltiz et al. 1959, p. 262). Two coders were trained to independently classify responses as indicating either an ethnocentric or non-ethnocentric orientation. The intercoder agreement exceeded 98 percent, and the 11 (of 863) disputed coding decisions were reconciled and assigned to a single category. Of the 863 respondents, 304 (35%) were classified as "consumer ethnocentrics" and 559 (652) as "nonCethnocentrics."
Responses classified as "ethnocentric" reflected various themes, the most frequent being that consumers should support American companies, oftentimes with the addendum that the economy is in trouble and needs help. The following response is illustrative: "I do not believe we should buy foreign products. If we bought more American-made products, we would have less unemployment. I prefer, if it is at all possible, to stick to our own products. " Several themes also appeared in the non-ethnocentric responses. The most frequent comments were that American consumers have the right to choose whatever products they want in order to get the best value; that foreign competition promotes increased quality of domestic products; that trade unions are responsible for the non-competitive position of American firms; and so f orth.
Background Characteristics of Consumer Ethnocentrics and Non-Ethnocentrics
An initial consideration is whether ethnocentrics and non-ethnocentric consumers are fundamentally different in background characteristics. Results presented in Table 1 show that the two groups are roughly equivalent in life cycle stage and racial character, but clear differences are manifest in educational achievements, income levels, and social class attainment. In particular, consumer ethnocentrics reflect significantly lower education, income, and social class levels. The social class finding is based on analysis of only the three major classes in Hollingshead's (1949) classification, due to the paucity of respondents classified as upper or lower class. Analysis of the three mass classes indicates an identical proportion, 43 percent, of ethnocentrics and non-ethnocentrics in the middle class. However, 31 percent of the ethnocentrics are upper middle class and 26 percent lower middle class, whereas the proportions for non-ethnocentrics are 44 and 13 percent, respectively.
COMPARATIVE PROFILE OF CONSUMER ETHNOCENTRICS AND NON-ETHNOCENTRICS
Differences Between Ethnocentrics and Non-Ethnocentrics Toward Foreign Cars
A series of tests examined the general hypothesis that consumer ethnocentrics are fundamentally different than non-ethnocentrics in how they perceive and evaluate foreign cars. The first set of results compares the two groups on a number of specific disaggregated variables--beliefs, evaluations, normative beliefs, etc. Next, the two groups are compared on a social class-by-social class basis to rule out a possible alternative explanation for the previous findings. Differences between ethnocentrics and non-ethnocentrics in terms of aggregated cognitive structure and subjective norm variables provide a final set of findings.
Disaggregated Findings. Multivariate analysis (MANOVA) was performed on each of several sets of dependent variablesC beliefs, evaluations, normative beliefs, motivations to comply, intentions, affordability considerations, and desirability considerations. With exception of the evaluation and motivation to comply measures, all analyses were multivariate significant (all p's < .001), thereby supporting the general research hypothesis that consumer ethnocentrics and non-ethnocentrics differ in their beliefs, purchase intentions, etc. but not in their evaluations of automobile attributes or in their motivations to comply with referents. The following discussion of univariate findings is appropriately restricted to the five sets of dependent variables that achieved multivariate significance.
The first findings in Table 2 reveal that consumer ethnocentrics' perceptions toward foreign automobiles were uniformly less favorable--they were significantly more likely to agree that owning a foreign car would require high car payments (the only negative consequence) but significantly less likely to acknowledge all but one (status) of the eight possible positive consequences attendant with foreign car ownership.
Respondents' normative beliefs provide a second set of findings. Consumer ethnocentrics' normative beliefs were significantly less positive for all four referent groups. Thus, consumer ethnocentrics personally were not as favorably inclined toward foreign cars, not did they perceive that others favored their purchasing foreign-made automobiles.
Generic product purchase intentions, "brand" intentions, and other Warshaw (1980) variables provide the last set of findings in Table 2. The results show that: consumer ethnocentrics were significantly less likely to purchase a new automobile within the next year or so (i.e., BIy); they were significantly less likely to purchase a foreign-made automobile (BI ), and they were significantly less likely to purchase a foreign car given that they do purchase a new automobile (BI ); there were not statistically significant differences/between ethnocentrics and non-ethnocentrics in their perceptions of how difficult or easy it would be to afford a new automobile, but a new foreign car was perceived as less affordable by ethnocentrics and non-ethnocentrics; the desire to purchase a new automobile was stronger for non-ethnocentrics and so was their desire to own a foreign car.
Social Class Findings. The foregoing findings demonstrate clear and persistent differences between consumer ethnocentrics' and non-ethnocentrics' perceptions and intentions regarding foreign automobiles. The possibility remains, however, that these differences are attributable to some other factor which confounds or moderates the ethnocentrism variable. The best candidate for this role is social class, which previously was shown to be lower for consumer ethnocentrics. All twenty results presented in Table 2 were reanalyzed on a class-by-class basis. These analyses were limited to Hollingshead's (1949) three middle classes, because, as noted previously, the number of respondents in the upper and lower classes was insufficient.
DIFFERENCES BETWEEN BELIEFS, PERCEPTIONS, AND INTENTIONS OF CONSUMER ETHNOCENTRICS AND NON-ETHNOCENTRICS
The findings for each social class generally reproduced the overall findings. This was determined by comparing the statistical conclusion (i.e., to reject or not to reject the null at a = .05 or lower) for each of the twenty tests on each social class (i.e., nine personal beliefs, four normative beliefs, and seven Warshaw zeasures) against the overall sample conclusion from Table 2.
A "mismatch" was registered when a specific within-social class result led to a different statistical conclusion than that from the overall sample result. Of the 20 comparisons for each social class, the number of "mismatches" was five, two, and five for the upper middle, middle, and lower middle classes, respectively. The absence of any systematic pattern in these mismatches suggests that the fundamental differences between consumer ethnocentrics and non-ethnocentrics are real rather than artifactually conditioned by social class confounds.
Further demonstration is afforded by results from a series of seven MANOVAs, one for each of the seven sets of dependent variables (beliefs, evaluations, normative beliefs, motivations to comply, intentions, affordability considerations, and desirability considerations). Each set represented a vector of dependent variables, and ethnocentrism and social class as factors in 2 x 3 MANOVAs. The primary purpose of these analyses was to examine for significant multivariate interaction effects between the two factors, the presence of which would have suggested that ethnocentrism is confounded with social class. Not one of the seven multivariate interaction effects was statistically significant, thereby removing social class as an alternative explanation for the previously observed differences between ethnocentrics and non-ethnocentrics.
Aggregated Findings. Another set of tests for differences between ethnocentrics and non-ethnocentrics was accomplished by aggregating various attitudinal, cognitive structure, and subjective norm variables. Three standard Fishbein variables were formed: Aact, Xb.e., and i:NB MC.. Differences between consumer ethnocentrics and non-ethnocentrics were tested by analysis of covariance for each of the three aggregated criterion variables. Ethnocentrism, social class, and sex were factors and age was a covariate. Past research on foreign products (Bannister md Saunders 1978; Schooler 1971) along with common-sense justify the addition of sex and age to these analyses. One might expect that younger consumers and perhaps males would be more favorably disposed toward foreign automobiles.
The findings in Table 3 reveal that social class was insignificant in every analysis. Sex had significant effects zn the cognitive structure (Sbiei) variable, with males possessing more favorable structures. Age was statistically significant for only the Aact variable; younger consumers lad more favorable attitudes. Finally, the ethnocentrism variable attained statistical significance in all three analyses; ethnocentric consumers were uniformly less favorably disposed toward foreign cars.
SUMMARY RESULTS FOR ATTITUDINAL AND NORMATIVE INFLUENCE VARIABLES
Automobile Ownership. A final interest is whether consumer ethnocentrics and non-ethnocentrics differ in foreign automobile ownership. Respondents listed the makes and models of all automobiles they owned, and households were subsequently classified as "owning none" or "owning one or more" non-luxury Japanese or other imported automobiles. Cross-classification of the ownership and ethnocentrism variables uncovered a strong relationship: 35 percent of the non-ethnocentrics owned at least one foreign car, whereas only 10 percent of the consumer ethnocentrics were foreign car owners (X2 corrected = 60.18, 1 df, p < .001). Next, each of the three mass social classes was analyzed separately to remove the potentially confounding effects of social class. Twenty-one percent of the upper middle class ethnocentrics owned a foreign-made auto compared with 41 percent of the non-ethnocentrics (X2 corrected = 5.99, 1 df, p < .01). Only seven percent of the middle class ethnocentrics owned a foreign car in contrast to the 40 percent of non-ethnocentric owners (X2 corrected = 26.03, 1 df, p < .001). Likewise, only six percent of the lower middle class ethnocentrics owned a foreign car, whereas 20 percent of the non-ethnocentrics were owners (X2 corrected = 2.47, 1 df, p = .12).
Ethnocentrism characteristically involves strong pro-ingroup sentiments combined with strong anti-outgroup sentiments (Adorno et al. 1950). As a sociopsychological concept, ethnocentrism is designed to account for why people identify with groups, believe their groups are best, and regard other groups (outgroups) as antithetical to ingroups. Applied to the domain of consumption, the concept of ethnocentrism (Consumer Ethnocentrism) is a potentially valuable explanan of why some consumers have negative orientations toward products manufactured outside their native country. In the present study, respondents designated as "consumer ethnocentrics" ascribed to the position that purchasing foreign automobiles is wrong because it hurts the country, the domestic economy, and fellow citizens. "Non-ethnocentrics," who did not share these sentiments, held significantly more favorable beliefs, attitudes, and intentions toward foreign automobiles.
An obvious and legitimate concern regarding this study is the issue of measurement quality: Was the measure used (a single open-ended question) too crude, and was consumer ethnocentrism in fact measured? It is undeniable in response to the first issue that a single open-ended question lacks precision, is less desirable than a multi-item scale, and that a psychometrically-sound scale to measure consumer ethnocentrism would unquestionably be preferable. However, the fact remains that such a scale is unavailable. It is furthermore notable that the consistently significant differences detected would have been even stronger had a more Precise measure of consumer ethnocentrism been available.
A second issue is whether the measure of consumer ethnocentrism, imprecision aside, did indeed tap this construct rather than some other. This of course is a validity issue. Criterion validity has been demonstrated, as ethnocentrics were predictably different along a variety of measures. The high degree of intercoder reliability in classifying respondents into the two ethnocentric groups fulfilled a necessary (but not sufficient) condition toward establishing construct validity, but the unavailability of a second consumer ethnocentrism measure precluded a determination of convergent validity. Discriminant validity likewise was not established, as there conceivably are other untested variables with which consumer ethnocentrism is fundamentally inseparable. However, it can be said with confidence that consumer ethnocentrism is not a mere manifestation of more basic social class factors.
Additional work obviously is needed before it can be concluded with assurance that consumer ethnocentrism is a useful concept. A valid scale adhering to accepted psychometric procedures is a needed next step. Such a scale would transform consumer ethnocentrism from a dichotomous or "specific non-variable," as operationalized in this research, to a continuous or "general variable." The latter is always superior due to greater discriminatory ability and because it permits more powerful statistical analysis (Hage 1972).
Although incidental to the present paper, it may be of interest to note that efforts have recently been undertaken to construct a psychometrically-sound scale to measure consumer ethnocentrism and related phenomena. The proposed Consumer Orientation toward Foreign Products Scale is conceptualized as a multifaceted construct that includes cognitive, affective, and normative elements. The construct's domain spans consumers ' object-based beliefs and attitudes (i.e., perceptions of foreign product quality, value, etc.), normative-based beliefs and attitudes (i.e., perceptions of whether one should or should not purchase foreign-made products), and personalistic-based considerations centering on consumers' personal interests and welfare.
The proposed scale and the study presented herein are based on the belief that this global form of consumer choice behavior (domestic versus foreign products) has strategic implications. In terms of private policy, knowledge of the number and identifying characteristics of consumer ethnocentrics would be useful in formulating market segmentation and promotional strategies. For example, knowing the magnitude of ethnocentric attitudes within a targeted market segment would indicate whether appeals to patriotism are appropriate. Moreover, the application of a psychometrically-sound scale in companies' periodic consumer research studies would enable a determination of the role played by ethnocentric attitudes in influencing product choice behavior
Adorno, T.W., Else Frenkel-Brunswik, Daniel J. Levinson and R. Nevitt Sanford (1950), The Authoritarian Personality, New York: Harper & Row.
Ajzen, Icek and Martin Fishbein (1980), Understanding Attitudes and Predicting Social Behavior, Englewood Cliffs, NJ: Prentice-Hail. Inc.
Anderson, W.T. and William H. Cunningham (1972), "Gauging Foreign Product Promotion," Journal of Advertising Research, 12 (February), 29-34.
Bannister, J.P. and J.A. Saunders (1978), "UK Consumers' Attitudes Towards Imports: The Measurement of National Stereotype Image," European Journal of Marketing, 12, 562-70.
Berkman, Harold W. and Christopher C. Gilson (1978), Consumer Behavior: Concepts and Strategies, Encino, CA: Dickenson Publishing Co., Inc.
Chang, Edward C. and Edward H. Ritter (1976), "Ethnocentrism in Black College Students," The Journal of Social Psychology, 100, 89-98.
Etzel, Michael J. and Bruce J. Walker (1974), "Advertising Strategy for Foreign Products," Journal of Advertising Research, 14 (June), 41-44.
Fishbein, Martin and Icek Ajzen (1975), Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research, Reading, MA: Addison-Wesley.
Gaedeke, Ralph (1973), "Consumer Attitudes Toward 'Made In' Developing Countries," Journal of Retailing, 49 (Summer), 13-24.
Hage, Jerald (1972), Techniques and Problems of Theory Construction in Sociology, New York: John Wiley & Sons, Inc.
Holbrook, Morris B. (1978), "Beyond Attitude Structure: Toward the Informational Determinants of Attitude," Journal of Marketing Research, 15 (November), 545-56.
Hollingshead, August B. (1949), Elmtown's Youth: The Impact of Social Class on Adolescents, New York: John Wiley & Sons Inc.
Jacoby, Jacob (1978), "Consumer Research: A State of the Art Review," Journal of Marketing, 42 (April), 87-96.
Levine, Robert A. and Donald T. Campbell (1972), Ethnocentrism: Theories of Conflict, Ethnic Attitudes, and Group Behavior, New York: John Wiley & Sons, Inc.
Markin, Rom J. (1974), Consumer Behavior: A Cognitive Orientation, New York: MacMillan Publishing Co., Inc.
Mazis, Michael B. and Olli T. Ahtola (1975), "A Comparison of Four Multi-Attribute Models in the Prediction of Consumer Attitudes," Journal of Consumer Research, 2 (June), 38-52.
Miniard, Paul W. and Joel B. Cohen (1979), "Isolating Attitudinal and Normative Influences in Behavioral Intentions Models," Journal of Marketing Research, 16 (February), 102-110.
Miniard, Paul W. (1981), "An Examination of the Fishbein-Ajzen Behavioral Intentions Model's Concepts and Measures," Journal of Experimental Social Psychology, 17, 309-339.
Oliver, Richard L. and Philip K. Berger (1979), "A Path Analysis of Preventive Health Care Decision Models," Journal of Consumer Research, 6 (September), 113-22.
Rierson, Curtis C. (1966), "Are Foreign Products Seen As National Stereotypes?," Journal of Retailing, 42 (Fall), 33-40.
Rierson, Curtis C. (1967), "Attitude Changes Toward Foreign Products," Journal of Marketing Research, 4 (November), 385-87
Schooler, Robert D. (19071), "Bias Phenomena Attendant to the Marketing of Foreign Goods in the U.S.," Journal of Inter- national Business Studies, 2 (Spring), 71-80.
Schooler, Robert D. and Albert R. Wildt (1968), "Elasticity of Product Bias," Journal of Marketing Research, 5 (June), 199-38.
Selltiz, Claire, Marie Jahoda, Morton Deutsch and Stuart W. Cook (1971), Research Methods in Social Relations, New York: Holt. Rinehart and Winston.
Vinson, Donald E., James E. Scott and Lawrence M. Lamont (1977), "The Role of Personal Values in Marketing and Consumer Behavior," Journal of Marketing, 4 (April), 44-50.
Warr, Peter B., Judith Faust and Godfrey J. Harrison (1967), "A British Ethnocentrism Scale," British Journal of Social Clinical Psychology, 13, 145-55.
Warshaw, Paul R. (1980), "A New Model for Predicting Behavioral Intentions: An Alternative to Fishbein," Journal of Marketing Research, 17 (May), 153-72.