Perceptions of Consumer Concern By Buiness, Legislators and Services: a Multivariate Linear Model

William R. Darden, University of Arkansas
Thomas J. Stanley, Georgia State University
Roy Howell, University of Illinois
ABSTRACT - This article presents the results of a study of consumer perceptions of "consumer concern" by selected industries. The major thrust of the paper focuses on the idea that different kinds of consumers have differing perceptions as to the "consumer concern" of different businesses, services and agencies.
[ to cite ]:
William R. Darden, Thomas J. Stanley, and Roy Howell (1982) ,"Perceptions of Consumer Concern By Buiness, Legislators and Services: a Multivariate Linear Model", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 507-514.

Advances in Consumer Research Volume 9, 1982      Pages 507-514


William R. Darden, University of Arkansas

Thomas J. Stanley, Georgia State University

Roy Howell, University of Illinois


This article presents the results of a study of consumer perceptions of "consumer concern" by selected industries. The major thrust of the paper focuses on the idea that different kinds of consumers have differing perceptions as to the "consumer concern" of different businesses, services and agencies.


With the rapid growth of the consumerist movement, much research has been undertaken to develop a better understanding of major consumer issues and how various groups perceive them (Barksdale and Darden, 1972; Hustad and Pessemier, 1973; Landon, 1977). One aspect of these undertakings is explaining the relationship between the discontent of consumers and their personal characteristics (Aaker and Day, 1974; Barksdale and Darden, 1972; Hustad and Pessemier, 1973; Landon, 1977; Hill and Garner, 1974). This study will introduce two new concepts which would seem to be important in the study of consumer activism and present the results of research designed to test their role in the process.

Previous Research

A recent study for the Sentry Insurance Company, Consumerism at the Crossroads (1977) has made a valuable contribution. As a part of that study 1,510 consumers evaluated how well different industries served the consumer. The results show that consumers as a group perceive major differences between industries on how well they serve the consumer.

Throughout the Sentry study, evidence documents consumer concern about lack of sensitivity by members of the business community. Senior-level business managers were found to be "... out of touch with what consumers find most worrisome in their dealings with business" (p.iv).

Determinants of Consumerism

One weakness of Sentry's study is that it did not study the characteristics of consumers who are most dissatisfied with their treatments by particular industries. Previous studies to provide some interesting findings of a general nature. Barksdale and Darden (1972) found age and political predisposition relevant. A study of Canadian consumers who complained about business practices found "... the average consumer complainer... a middle aged, well-educated, affluent, managerial-professional man or woman" (Liefield, et. al., 1972).

Hustad and Pessemier indicate that anti-business (pro consumerism) consumers were more often liberal in their political views and had higher educational and occupational status levels (1973). Those becoming upset with business and marketing practices and taking action, "... were better educated, earned high incomes and were frequently in the top social class...and were more liberal" (Hustad and Pessemier, 1973, p. 323).

Lundstrom and Kerin (1976) found that consumers discontented with business tend to be older males with high incomes and education, and more professional in occupation. Others (Liefield, et al., 1973; Miller, 1973; Pfaff and Blivice, 1977) have noted the positive relationship between income, education, and occupational status and consumer discontent with business.

Miller (1973) indicates that, in contrast to previously cited studies, discontented consumers are likely to be younger, more mobile, more educated, and in an earlier stage of the family life cycle. This view of the discontent consumer as younger is supported by Warland, et al., (1972), who find that the contented consumer "appears to be the oldest and the most politically conservative group." Pfaff and Blivice (1977) also indicate that "satisfaction increases with age, activism decreases with age."

Previous research would thus tend to support the conclusion that the discontented consumer is likely to be more educated and to have a higher income and higher occupational status (in essence, of a higher social class). The evidence with regard to age and stage in the family life cycle is less clear, with some studies indicating that the typical discontented consumer is of middle age (Liefield, et al., 1955, p. 79), while others characterize the discontented consumer as younger (Warland, et al., pp. 153155; Pfaff and Blivice, 1977, pp. 119-112).

Perceived Consumer Concern and Price

Reasonableness Constructs

This paper proposes that two constructs - (1) perceptions of an industry's "concern toward consumers" and (2) "price reasonableness" of an industry, as internalized by consumers - are variables that are largely neglected in the study of consumerism. The concern, which consumers feel is held for them by the different business and public institutions, and the perceptions of price reasonableness, held also by consumers, would seem to be two sets of variables which may moderate between satisfaction and the choice of actions taken in the case of initial satisfaction. Thus, in the framework of the model of Post-purchase Consumer Processes (Andreasen 1977), the concerns of industries toward consumers and the reasonableness of their pricing policies (as perceived by a consumer) could be expected to influence his/her purchase expectations. When the process of comparing product or service performance to expectation yields dissatisfaction, levels of perceived consumer concern and price reasonableness exert an influence on the choice of action taken by the consumer.

These two constructs (perceived consumer concern and price reasonableness) are consistent with the orientation of the consumer discontent scale developed by Lundstrom and Lamont (1976). A substantial number of items in the scale that measures their construct deal with the degree to which consumers feel that business (in general) is concerned about their welfare and is practicing reasonable pricing policies.


1. It is hypothesized that the demands of particular stages of the consumer life cycle (Wells and Gubar 1966) color how people perceive the concern for consumers on the part of those offering products and services to households. This should be reflected in their attitudes about the consumer concern and price reasonableness for those businesses theY seek to have serve them.

2. Income is suggested to be positively related to how consumers perceive the consumer concern and price reasonableness held by particular business and public sectors. Those with higher incomes are more likely to be experienced in purchasing products and thus to know better their options within a particular business or public service.

3. Political Predisposition is hypothesized to relate both to the perceived consumer concern and to the price reasonableness constructs. Warland, et al., (1975) suggest that conservative consumers are less likely to be "upset" in the marketplace. Conservatives generally place responsibility within the individual, rather than with outside forces (Williamson, et al.. 1977).

4. Occupation is hypothesized in this study to relate to both the perceived consumer concern and price reasonableness constructs. It seems logical that executives have empathy with those executives who also provide them with products and services. Skilled and non-skilled workers may not understand the motivations of managers who provide their products and services.


Based on the findings of the Sentry study (1977) and Andreasen's work (1977), it seems appropriate to examine simultaneously the individual perceptions that a consumer holds toward several businesses and related institutions. In this case, we are dealing with a multivariate dependent vector variable, rather than a generalized consumerist attitude toward "business" (with the concomitant unrealistic univariate analysis). As suggested by Wind and Denny (1974), multivariate analysis of variance (MANOVA) may be used to examine such dependent response surfaces for possible association with predictor factors.

The objective of this paper is to construct and to test such a model. We propose that such a model is more isomorphic to reality and is less likely to lead to contradictions in research findings (Bock and Haggard, 1968). If MANOVA gives overall significant results, then univariate analysis is appropriate in determining the nature of the relations. The symbolic model tested is a General Linear Model (referred to as Model I in this study), and is represented by (Bock and Haggard. 1968):



(a) Yijk1 is a three element vector containing measures of the ith consumer's perceptions of the concern held first by business (1), second by legislators (2), and third by professional services (3). The ith consumer holds the jth political predisposition, is in the kth stage of the family life cycle, and works in the 1th occuPation.

(b) u... is the grand mean vector of Yijk1, tj, Bk and o1 are each three elements vectors that represent the independent effects on Yijk1 due to consumer holding the jth political predisposition (PP), being in the kth stage of the family life cycle (FLC) and employed in the 1th occupation.

(c) rjBko1, tjBk, tjo1, Bko1 are three element victors that represent all possible interactions among the main effects.

(d) Zijk1 is income (treated as a covariate and centered to mean-zero).

(e) sijk1 is a three element vector variable that is assumed to be random and to have a null vector as its expected value.

This somewhat complicated model allows the exploration of the complex relationships between patterns of consumer characteristics and patterns of consumer perceptions of the concern for consumers on the part of several businesses (in this case private and public services). Thus many of the problems inherent in successive univariate tests of relationship are not encountered (see Cooley and Lohnes, 1971; Bock, 1963; and Darden and Rao, 1980).

The approach involves modeling the response surface (in this case consumer perceptions of consumer concern held by managements of selected businesses and public services) in a sequential fashion with "effect coding" for the effects of each factor (see Cramer 1967). The significance of the last factor is found by testing the reduction in the sum of squares and cross-products matrix found with the introduction of the last factor into the full linear model. [For a discussion of the use of multiple regression in MANOVA, when non-orthogonal designs are encountered, see Woodward and Overall, 1977.] The multiple regression approach to multivariate analysis of variance also allows the researcher to address problems inherent in designs with unequal cell sizes (Perreault and Darden, 1975).

Price Reasonableness Model

The same modeling approach can be used to test yet another response surface of consumers, one that reflects their appraisal of the performance of private and public services. In this case, the price reasonableness of the selected services, as perceived by respondents, is the surface to be modeled. Andreasen notes that, "... it was found that when the type of problem was investigated, the sole problem in fourteen percent of the purchases was that it 'cost too much."' (1977, p. 18). Andreasen (1977) also indicates that the perception of prices as unreasonable may be a surrogate for poor performance.

As suggested by the literature (Anderson, et al., 1976; Barksdale and Darden, 1972; and Lundstrom and Kerin, 1976), demographic characteristics would appear to explain some of the variation in price reasonableness attitudes. Earlier work in social psychology (Hollingshead and Redlich, 1958) suggests that social class is also a significant causative factor. In fact, stage in family life cycle, income, occupation and political predisposition are again viewed as the explanatory factors. Thus "price reasonableness" can be modeled in the same fashion as was the "consumer concern" model. Equation 2 reflects the same predictors, but with a different response surface


respondent attitudes about the "price reasonableness" of (1) retailers, (2) professional services, and (3) utilities. We refer to this as Model II


Analysis Approach

Exploration of these ideas proceeded through two stages:

1. First, a sample of consumers from the State of New York was obtained and analyzed to determine the applicability of the two models.

2. Responses from a sample of top executives from For- tune's 500 was gathered.

Since the Fortune's 500 top executives all tended to be high on social class, and were at later stages of the family life cycle, their attitudes toward consumer discontent in both domains were compared with those of the consumer sample. The purpose of this analysis is to compare how those instrumental in shaping business policy differ from a cross-section of consumers with respect to consumer discontent.

Consumer Data

The consumer data were obtained by questionnaires mailed during the summer of 1976. Respondents were randomly selected from New York State telephone directories. The sample represented all consumers throughout the state who had listed phone numbers. Selection of consumer respondents was made in proportion to the population within every directory area of the state.

Twelve hundred consumers were mailed questionnaires. A postcard was also mailed to each potential respondent approximately three weeks after the initial mailing of the questionnaire. Three hundred and eighty-five, approximately 32 percent, of those mailed were returned and were usable.

Prior studies suggest that late responders to mail surveys have characteristics that are similar to non-responders. [See for example A. N. Oppenheim (1966, p. 34).] Thus a comparison of the responses of those replying early with those replying late should give one basis for evaluating the extent of non-response bias. This was accomplished in two ways. First, the responses of those in the first wave (262 questionnaires) were compared with those received after mailing the reminder post card (123 questionnaires). Second, the time required for the receipt of each questionnaire was checked for relation with the questionnaire response profile. Both tests suggest little differences between early and late responders. Thus some support is provided for the representativeness of the data.

Consumer respondents were representative of the New York State head-of-household population; the age and income characteristics of this consumer sample were highly congruent with the latest census data. (Table 1)

Fortune's 500 Executives

Four hundred top executives were randomly selected from the Fortune 500 Companies. Each was sent a self-administered questionnaire and a cover letter requesting that they fill out the instrument; 185 returned fully completed and usable questionnaires, in a response rate of approximately 46%. Three kinds of top executives returned the questionnaire:

1. Corporate Presidents. Fifty-seven respondents were top executive officer within the corporation.

2. Marketing Vice-President. These executives are in direct control of corporate and product advertising, product planning, dealer relations, pricing of products, and managing product services. 111 of the executives were in this category.

3. Vice-President for Advertising and/or Public Relations. Seventeen respondents held positions with primary responsibility for advertising and/or public relations for the firm.

Fortune's 500 Executives filled out the same questionnaire as did the random sample of consumers from the state of New York.


Respondents were asked their opinions about the general concern for consumer needs and desires shown by supermarkets, oil companies, automobile manufacturers, utility companies, telephone companies, state legislators, U.S. legislators, medical doctors, hospitals, and attorneys (Table 3-A). Respondents were also asked to indicate their feelings about the prices generally charged consumers by supermarkets, department stores, automotive dealers, medical doctors, attorneys, hospitals, telephone companies, and utility companies (Table 3-B). Those responding also provided information about their political predisposition (conservative. moderate. or liberal). stage in family life cycle, and social class.







Responses to the price items and to the concern items were factor analyzed by Hotelling's method of Principal Axes with varimax rotation (Cooley and Lohnes, 1971). Factor loadings were used to categorize specific industries into broader industry classifications. The concern for consumer items followed the hypothesized factor patterns; (l) business, (2) government representatives, and (3) professional services. The loadings on the sectors in the "price reasonableness" domain suggested, also as hypothesized, three patterns of factors falling under the general heading of (1) retailers, (2) professional services, and (3) utilities (Table 3-A). All businesses had factor loading that exceeded .57 on some one factor, illustrating convergent validity, while those same items Loaded low on all other factors, indicating reasonable discriminant validity.

The loadings from each of the two factor analyses appeared to provide evidence of sufficient construct validity, and the Cronbach Alpha coefficients for the sectors grouped under each factor demonstrated strong item reliability (Tables 3-A and 3-B). These scaling results are considered very good (Cronbach, 1953) and lead us to summate sectors grouped under a factor to obtain "summated" perceptions for each domain (first, creating a "consumer concern" response surface, and second creating a "price reasonableness" response surface). A score for each dimension was obtained by summing the corresponding ratings of items that were suggested as being basically the same dimension by factor analysis (Tables 3-A and 3-B). For example, the price reasonableness of utilities was constructed by adding each respondent's judgments of price reasonableness toward telephone companies to that of utility companies. The summated scores for each respondent are then divided by the number of items for that construct to place the score back into the range of the original verbal rating scale.




Model I: Consumer Concern by Businesses and Services

Table 4 contains the statistics that result from testing the "consumer concern" model (symbolically depicted in equation 1). The analysis proceeds by first considering the covariate and interactions among the main effects. Then the significance of each main effect is considered in turn.

Income. We see that income (the covariate in this model) does relate significantly to the "consumer concern" of select services as perceived by New York consumer (P z .05). Section I of Table 6 contains discriminant analysis loadings that suggest the nature of this relationship. In general, consumers with higher incomes tend to perceive greater concern for consumers on the part of businesses (supermarkets, oil companies, automobile manufacturers, and utilities) and professional services (medical doctors, hospitals, attorneYs).

These results appear to conflict with the findings of prior research. For example, Liefield, Edgecombe and Wolfe (1975), Warland, Herrmann, and Willits (1975), Pfaff and Blivice (1977) and Miller (1973) all find higher incomes are associated with higher levels of consumer activism. Yet it is not necessarily true that those most active in the consumerist sphere cannot perceive higher levels of concern on the part of private and public services. In fact, it seems quite likely that those making higher incomes are more involved in business affairs, and therefore are more knowledgeable in this area. These same persons may simultaneously be likely to have empathy for the viewpoint of those offering services (with whom they mix) and yet likely to know their rights in business and stand up for them. It may be a matterS in this case, of knowing one's options and not being intimidated by nonpersonal business institutions.

As related to Andreasen's Post-purchase Consumer Process (1977), this finding may indicate that the individuals' choice of action upon being dissatisfied may be affected by the degree of concern for the consumer perceived as being held by the business or institution. If the consumer perceived as being held by the business or institution. If the consumer perceives little concern for his or her well-being on the part of the business or institution, the consumer may not take action, as this would be useless if the business doesn't "care." On the other hand, if business is perceived as concerned, action on the consumer's part may be perceived as being a more effective response to dissatisfaction.



Interactions. One of the most interesting findings in the study is that political predisposition (PP), family life cycle (FLC), and occupation (OC) does not significantly interact with each other to create independent effects on the "consumer concern" response function (Section II of Table 4.)

Political Predisposition. By far the most interesting finding of the study is the high overall association between political predisposition and perceptions of "consumer concern" on the part of select private and public institutions (P < .001:. Examination of the univariate F-ratios in section III of Table 4 suggests that most of this relationship is with the "business" factor; that is, political predisposition appears to relate strongly with respondent perceptions of "consumer concern" by private companies.

It comes as no great shock that "conservative" respondents report higher levels of concern than do "liberals" (Section II, Table 6.) Nevertheless, the finding provides some construct validity for the study and suggests that political orientation has a greater role in determining consumerist attitudes than had been previously reported by Barksdale and Darden (1972). In this case, the study provides a raison d'etre for the findings of Warland, Herrmann and Willits (1975). If conservative consumers perceive greater concern on the part of businesses and public service bodies (as suggested in the test of our model), it seems likely that they will not perceive "mistreatment" in the marketplace - or when they do, they will be less likely to take "action."

This finding, when compared to the finding of greater perceived concern among those with higher income in this study and previous findings of higher levels of activism among those with higher income (Liefield, Edgecombe and Wolfe, 1975; Warland, Herrmann and Willits, 1975; Pfaff and Blivice, 1977; Miller, 1973) would seem to indicate that the interaction between income and political predisposition may be an important factor in the consumer's choice of action, while (as noted previously) not affecting perceived concern.

Family Life Cycle. Table 4 shows that the three stages of the family life cycle [Wells and Gubar's original classification of the family life cycle is based on (1) marital status, (2) age, and (3) age of youngest child. They include: (1) Bachelor, (2) newly married, (3) Full Nest Stage I, (4) Full Nest Stage II, (5) Empty Nest I, (6) Empty Nest II, and (7) Solitary Survivors. In this study, we use three stages in the FLC: (1) under 40, single or married without children; (2) under or over 40, married with children at home; and (3) over 40, married or single, and no children at home.] are indeed strongly related to "consumer concern" by businesses. Examination of the univariate F-ratios (see Section rv of Table 4) shows, however, that only the "Legislators" consumer concern factor contributes significantly to this association (p < .03). Thus those in the middle stages of the family life cycle perceive higher levels of consumer concern or the part of state and national legislators than do those in earlier or later stages (Table 6). This finding helps document the origins of the "Grey Power" movement across the United States.

Occupation. Studies of consumer activism (behavior) have consistently reported that occupation is an important determinant of complaint behavior (Liefield, et al., 1975; Pfaff and Blivice, 1977; Diamond, et al., 1976). However, when the response surface being studied is respondent perceptions of "consumer concern on the part of businesses," less evidence is available. The model in this study shows little contribution by occupation to the explanation of "consumer concern" of businesses as perceived by New York consumers, when considered with the effects of income adjusted for as a covariate.

Model II: "Price Reasonableness" of Business and Services.

Since the "price reasonableness" response surface (Model II) is within the same interest area as is "perceived consumer concern" (Model I), we expect that the same set of household characteristics to be related to this second response surface in a similar fashion. The statistics testing the estimates of factor effects for Model II are contained in Table 5. We note immediately that the hypothesized interaction effects (Section II) are not significant.

Income. As in Model I, income is strongly related to how respondents rated the "reasonableness" of the pricing policies of the three dimensional business response surface (F-ratio = 3.2; p < .02). Again, those consumers with higher incomes perceive businesses and services as having more reasonable pricing policies. Thus the pricing policies of retailers, professionals, and utilities are viewed as more reasonable by high income consumers and, conversely, viewed as less reasonable by low income respondents.

Political Predisposition. Political predisposition also shows strong overall relationship with respondent perceptions of how reasonable are the pricing policies of businesses. Rao's multivariate F-ratio is significant at the .01 level, indicating that at least one of the three response factors is also significant in a meaningful fashion (Section III of Table 5).

Table 6 reinforces our image of those respondents that reflect favorable perceived attitudes toward the motives of businesses toward consumers. Table 5 suggests that political predisposition is related to perceptions of "price reasonableness" of retailers (p < .05) and of utilities (p < .001), and Table 6 shows that politically conservative respondents perceive the pricing policies of both retailers and utilities as being more reasonable than do their liberal counterparts.





FLC and OC. Both family life cycle and occupation do not show overall relationship with consumer perceptions of the reasonableness of retailers, professional services and utilities. Even without the overall test of Model II, univariate analysis would have strongly indicated no relation on all three dimensions of the response surface.


One of the more interesting issues explored in this study is the hypothesis that top executives - with more experience in businesses and with a more global view of the intentions and objectives of administrative policies have more favorable perceptions of the "consumer concern" and the "reasonableness of pricing policies" of administrators of businesses, elected representatives and public services. In an attempt to provide support for this thesis, the perceptions of New York consumers and Fortune's 500 executives were compared in both these response areas.

The data in Table 7 support this hypothesis to a considerable degree. Multivariate analysis of variance - for both response surfaces - is significant at the .001 level. The Fortune's 500 executives view the "consumer concern" of businesses, elected representatives and professional persons (doctors, lawyers, etc.) as much higher than do the New York consumers. The canonical correlation produced by the general linear model (Cramer, 1967; Perreault and Darden, 1975) suggests the linear relationship to be strong (Rc = .52). In the same fashion, executives perceive the pricing policies of retailers, professionals, and public services to be more reasonable than do New York state consumers (Rc = .52).

Some specific conclusions include the following:

1. Stage in family life cycle, political predisposition and occupation do not appear to interact to create independent effects on perceptions of "consumer concern" and reasonableness of pricing policies" by the businesses and services in this study.



2. Fortune's 500 executives have uniformly more favorable perceptions of the "consumer concern" and the "reasonableness of pricing policies" of the businesses than do their New York State consumer counterParts.

3. Income and political predisposition of consumers significantly relate to the response surfaces in both model I and Model II.

4. Stage in family life cycle is significantly related to perceptions of "consumer concern" as analyzed in model I. It is interesting that these are most favorable for those f ami lies who have children at home

5. Respondents uniformly have a more unfavorable perception of the "consumer concern" held by legislators (state and national); the exceptions include "liberals" and "professionals" (medical doctors. lawyers. etc.)

6. "Professionals" (doctors, lawyers, etc.) are perceived in general as having less concern for consumers. This finding is somewhat unexpected; however, it probably is tied to the soaring costs of medical services and professional fees in general.

7. Respondents in general perceive "retailers" as having more reasonable pricing policies than do "professionals" and "utilities" (Table 6). Consumers have personal contact with retailing establishments through sales clerks and other personnel. The producing firm is probably viewed as setting prices, ameliorating consumer attitudes toward retailers.

8. The mean scores in Table 6 in general support the findings of the Sentry study. Means for all categories either approach "unconcerned" for consumer concern or "unreasonable" for pricing policies - and for all businesses, services and agencies. Thus consumer perceptions of the "consumer concern" of businesses - and its elected representatives - is not high. The same kind of judgment can be tentatively made with respect to how "reasonable" consumers view the pricing policies of similar agencies.


Aaker, David A. and Day, George S. (1974), Consumerism- Search for the Consumer Interest, New York, The Free Press.

Anderson, Rolph E., (1976), "Consumer Profiles of Complainants Versus Noncomplainants: An Exploratory Study," in Proceedings, (Houston, Southern Marketing Association).

Andreasen, Alan R. (1977), "A Taxonomy of Consumer Satisfaction/Dissatisfaction Measures," The Journal of Consumer Affairs, 11 (Winter), 11-24.

Barksdale, Hiram C., and Darden, William R. (1972), "Consumer Attitudes Toward Marketing and Consumerism," Journal of Marketing, 36 (October), 28-35.

Barksdales, Hiram C., Darden, William R., and Perreault, William O., Jr. (1976), "Changes in Consumer Attitudes Toward Marketing, Consumerism and Government Regulation: 1971-1975," The Journal of Consumer Affairs, 10 (Winter), 117-139.

Bock, Darrell R. (1963), "Programming Univariate and Multivariate Analysis of Variance," Technometrics, 5(February), 95-117.

Bock, Darrell R. and Haggard, Ernest A. (1968), "The Use of Multivariate Analysis of Variance in Behavioral Research," in Handbook of Behavioral Research, ed. D. K. Whitla, Reading. Mass.: Addison-Wesley.

Bourgeois, Jacques C. and Barnes, James G. (1976), "Consumer Activists: What Makes them Different?", in Advances in Consumer Research, ed. Beverlee B. Anderson, Proceedings of the Sixth Annual Conference of the Association for Consumer Research, 3, 73-80.

Bourgeois, Jacques C. (1979), "Viability and Profile of the Consumerist Segment," Journal of Consumer Research, 5(March), 217-228.

Cooley, William W., and Lohnes, Paul R. (1971), Multivariate Data Analysis, New York: John Wiley & Sons, 96-167.

Cramer, Elliot M. (1967), "The Revised MANOVA Program," Lab Report University of North Carolina Psychometric Lab, Chapel Hill, N.C.

Cronbach, Lee J., (1953), "Coefficient Alpha and the Internal Structure of Tests," Psychometrika, 16, 297-334.

Czepiel, John A., and Rosenberg, Larry J. (1976), "Consumer Satisfaction: Toward an Integrative Framework," Proceedings, Southern Marketing Association, 169-71.

Darden, William R. and Rao, C. P. (1980), "A Linear Covariate model of Warranty Behavior and Warranty Attitudes," Journal of Marketing Research, 16(February), Forthcoming.

Diamond, Steven L. Ward, Scott, and Faber, Ronald (1976), "Consumer Problems and Consumerism: Analysis of Calls to a Consumer Hot Line," Journal of Marketing, 40(January), 58-62.

Feldman, L. P. (1976), Consumer Protection: Problems and Perspectives, St. Paul: West Publishing Company.

Hill, C. J., and Garner, Sarah (1974), "Consumerism and the Corporation: The Marketing Concept or Public Relations?," Proceedings. Southern Marketing Association, 203-205.

Hollingshead, August B., and Redlich, Fredrick C. (1958), Social Class and Mental Illness: A Community Study, new York: John Wiley and Sons, 398-407.

Hustad, Thomas P. and Pessemier, Edgar (1973), "Will the Real Consumer Activist Please Stand Up: An Examination of Consumer's Opinions About Marketing Practices," Journal of Marketing Research, 8(August), 319-24.

Landon, E. Laird, Jr. (1977), "Consumer Satisfaction Differences Between Industry and Government," in Conceptualization and Measurement of Consumer Satisfaction and Dissatisfaction, ed. H. Keith Hunt, Marketing Science Institute, 355-364.

Liefield, J. P., Edgecombe, F. H. C., and Wolfe, Linda (1975), "Demographic Characteristics of Canadian Consumer Complainers," The Journal of Consumer Affairs, 9(Summer) 9 73-80.

Lundstrum, W. and Kerin, R. (1976), "Psychological and Demographic Correlates of Consumer Discontent," in Proceedings, ed. Howard C. Schnider, Eighth Annual Meeting of AIDS, 312.

Lundstrum, W. and Lamont, L. W. (1976), "The Development of a Scale to Measure Consumer Discontent," Journal of Marketing Research. 13(November) 373-91

Miller, John A. (1973), "Who Is the Discontented Consumer?", American Marketing Association Combined Proceedings, 486-490.

Miller, K. E. and Sturdivant, F. D. (1977), "Consumer Responses to Socially Questionable Corporate Behavior," Journal of Consumer Research, 4 (June), 1-70

Oppenheim, A. N. (1966), Questionnaire Design and Attitude Measurement, New York: Basic Books. Inc., 34-40.

Perreault, William D., Jr. and Darden, William R. (1975), "Unequal Cell Sizes in Market Experiments: Use of the General Linear Motel," Journal of Marketing Research, 12 (August). 333-42.

Pfaff, Martin and Blivice, Sheldon (1977), "Socioeconomic Correlates of Consumer and Citizen Dissatisfaction and Activism," in Consumer Satisfaction. Dissatisfaction and Complaining Behavior, ed. Ralph L. Day, Indiana University, 115-123.

Plummer, Joseph T. (1977), "Life Style, Social, and Economic Trends Influencing Consumer Satisfaction," in Conceptualization and Measurement of Consumer Satisfaction and Dissatisfaction, ed. H. Keith Hunt, Marketing Science Institute, 382-408.

Sentry Insurance Company (1977), Consumerism at the Crossroads.

Warland, Rex H., Herrmann, Robert O., and Willits, Jane (1972), "Dissatisfied Consumers: Who Gets Upset and Who Takes Action," The Journal of Consumer Affairs, 9(Fall).

Wells, William D. and Gubar, George (1966), "The Life Cycle Concept in Marketing Research," Journal of Marketing Re- search, 3 (November), 355-63.

Williamson, John B., Jerry F. Boren and Linda Evans (1977), "The Sociology of Social Problems, in Social Problems: The Contemporary Debates, 2nd Edition, Boston: Little, Brown, pp. 1-29.

Wind, Yoram and Denny, Joseph (1974), "Multivariate Analysis of Variance in Research on the Effectiveness of TV Commercials," Journal of Marketing Research, 11 (May), 3-9.

Woodward, J. Arthur and Overall, J. E. (1975), "Multivariate Analysis of Variance by Multiple Regression Methods," Psychological Bulletin, 82(January), 21-31.

Zaichkowsky, Judy and Liefield, John (1977), "Personality Profiles of Consumer Complaint Letter Writers," in Consumer Satisfaction, Dissatisfaction. and Complaining Behavior, ed. Ralph L. Day, Indiana University, 126-129.