An Investigation of Consumers' Attitudes Toward Complaining

Marsha L. Richins, Portland State University
ABSTRACT - Several researchers have documented the value of the consumer complaint mechanism. No study, however, has systematically investigated consumer attitudes toward complaining about unsatisfactory products. This paper reports exploratory and descriptive research to determine the domain of consumers' attitudes toward complaining and the relationship between these attitudes and actual complaint behavior. In empirical analysis, three attitude domains emerged. Different groups of consumers were shown to differ in their attitudes, and attitudes were significantly related both to propensity to complain and self-reported complaint behavior.
[ to cite ]:
Marsha L. Richins (1982) ,"An Investigation of Consumers' Attitudes Toward Complaining", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 502-506.

Advances in Consumer Research Volume 9, 1982      Pages 502-506

AN INVESTIGATION OF CONSUMERS' ATTITUDES TOWARD COMPLAINING

Marsha L. Richins, Portland State University

ABSTRACT -

Several researchers have documented the value of the consumer complaint mechanism. No study, however, has systematically investigated consumer attitudes toward complaining about unsatisfactory products. This paper reports exploratory and descriptive research to determine the domain of consumers' attitudes toward complaining and the relationship between these attitudes and actual complaint behavior. In empirical analysis, three attitude domains emerged. Different groups of consumers were shown to differ in their attitudes, and attitudes were significantly related both to propensity to complain and self-reported complaint behavior.

INTRODUCTION

Recent research has indicated that dissatisfactions with products and services are relatively common among consumers, and these dissatisfactions frequently cause considerable expense and inconvenience. A relatively small percentage of these dissatisfactions, however, result in actual complaints to the business involved in the transaction (Thomas and Shuptrine 1975). While customer complaints can be a headache for retailers and producers, they do provide companies with a chance to remedy the dissatisfaction and ultimately please the customer. Unvoiced complaints may have more serious consequences, especially in competitive markets where consumers may switch their loyalty to rival products or outlets (Hirschman 1970). Dissatisfied consumers also seem quick to tell others of their unpleasant experiences, magnifying the chances of lost sales. In addition, frequent experiences of consumer dissatisfaction may culminate in a negative attitude toward business, which is correlated with demands for more government regulation and intervention in the marketplace. Thus, it is in the interest of business to improve customer feedback.

One area which seems particularly relevant in the understanding of complaint behavior consists of consumers' attitudes toward complaining. It is now well accepted in the consumer behavior and psychology literatures that attitudes correlate with behavior. While some studies have examined indirect attitudinal correlates of complaining such as attitudes toward business and consumerism (Robinson 1977 Warland, Herrmann, and Willits 1975), at this time no study has systematically examined the relationships between actual attitudes toward complaining and complaint behavior.

The preponderance of work in attitudes toward complaining has been carried out by Bearden and his colleagues. In these studies, attitudes toward complaining were measured rather narrowly on three bipolar scales in response to the statement, "Complaining directly to the people responsible for unsatisfactory automobile repairs or service is ...." The three adjective pairs were wise-foolish, good-bad, and beneficial-harmful. Respondents were also asked about attitudes toward interactions with service personnel; psychosocial risk in complaining due to conflicts, worry, or embarrassment; two items concerning moral and social norms about complaining; and other variables.

In the work reported by Bearden and Teel (1980), the three-item attitudes toward complaining scale was used as the dependent variable. In other reports, the three-item attitudes toward complaining scale was used as an independent variable. In an exploratory study, Bearden, Crockett, and Teel (1979) found this scale, the moral norm item, and psychosocial risks significantly related to self-reported intention to complain. In larger studies (Bearden, Crockett, and Graham 1980; Bearden, Teel, and Crockett 1980), path analyses were conducted with the dependent variable a propensity to complain scale. Of the variables in the proposed causal chain, the three-item attitude scale bore the strongest relationship to propensity to complain. These studies by Bearden and his colleagues suggest that attitudes toward complaining do indeed bear a relationship to consumer complaints.

This study is a further investigation of attitudes toward complaining and has two goals:

1. To determine the nature of attitudes toward complaining. The literature cited shows some confusion concerning what should be included in the domain of attitudes toward complaining: Robinson used a measure of attitudes toward business, while Bearden measured direct perceptions of the complaining experience. In addition, Bearden's measure was limited in scope and its derivation was not reported.

2. To examine the relationship between attitudes toward complaining and behaviors. Previous studies correlated attitudes with intention to complain and propensity to complain rather than actual complaint behavior. This study assesses behavioral relationships by examining the relationship of attitudes with (1) propensity to complain, (2) actual complaint behavior, and (3) known group membership.

In addition to these major objectives, demographic correlates of attitudes toward complaining were investigated.

METHODOLOGY

Research to meet the goals described above was carried out in two stages. The first involved establishing a domain for attitudes toward complaining. The only published measure of attitudes toward complaining concerned only the wisdom, goodness, and beneficiality of complaining (Bearden and Teel 1980). The various definitions of attitude in the social psychology literature, however, suggest that such an approach is unnecessarily limited. Thurstone (1928), for instance, suggested that attitude denotes:

the sum total of a man's inclinations and feelings, prejudice or bias, preconceived notions, ideas, fears, threats, and convictions about any specified topic . . . (p. 530).

More recently, Fishbein (see Fishbein and Ajzen 1975) has characterized attitude or affect toward an attitude object as resulting from cognitions or beliefs about an attitude object. In fact, many researchers have measured attitudes by measuring beliefs (e.g., Lundstrom and Lamont 1976; Rossiter 1977) and that perspective is followed here. Thus, the important task in establishing a domain for attitudes toward complaining involves determining consumer cognitions about the act of making complaints to businesses. This was done through a review of literature on complaining behavior and depth interviews with adult consumers.

During depth interviews conducted by trained interviewers in the respondents' homes, 16 consumers (9 women and 7 men) were asked about their own complaint behavior and their perceptions of the complaint experience. Six areas were consistently mentioned by these individuals as affecting their feelings toward the complaint act. Two of these appear to bear a negative relationship with complaint behavior with the remaining areas positively related.

1. The affect, usually negative, experienced when one complains. Most respondents indicated that making a complaint may involve unpleasant conflicts, embarrassment, and annoyance. A few mentioned, however, that under some circumstances complaining can be a pleasurable, or at least cathartic, experience. It should be noted that such individuals were a distinct minority. The literature provides some treatment of this dimension. Bearden et al. (1979) measured psychosocial risk and Richins (1980) showed a negative correlation between perceived psychological costs of complaining and actual complaint behavior.

2. The extent of objective costs or trouble involved in complaining. Many consumers mentioned that making complaints is, quite simply, a lot of trouble and involves time, and occasionally, monetary costs. The greater the perceived costs, the lower the likelihood of complaining. This aspect of complaint perceptions was also addressed by Richins (1980).

3. Business responsiveness A major concern on the part of consumers interviewed was the efficacy of complaining and retailers' willingness to adjust complaints. Some believed complaining rarely resulted in remedying the dissatisfaction, while others generally considered complaints effective. One may hypothesize a positive relationship between perceptions of business responsiveness and complaint behavior.

4. Expectation of societal benefits. This variable also appears to bear a positive relationship with complaining. Several respondents felt that registering complaints has societal as well as personal benefits. These individuals were quick to point out that if enough people complained about a particular product, it would eventually be improved or removed from the marketplace. When queried, however, other individuals did not agree with this perspective.

5. Personal norms concerning complaining. Many consumers were concerned whether complaining is appropriate behavior, some stating they did not like to be seen or see themselves as complainers or trouble-makers and that this sometimes inhibited them from requesting remedy for their dissatisfactions. Another group had a distinctly opposite point of view, that people should complain when dissatisfied and in fact had a moral obligation to do so. This aspect of attitude was addressed in Bearden's studies by his normative variables.

6. Situational variables. During the interviews, a number of other issues were brought up by consumers, including blame for a particular dissatisfaction's occurrence, the role of the item's cost and importance to the consumer in determining whether a complaint would be registered, and the frequency with which the consumer patronized the business where the unsatisfactory product or service was purchased. Since these variables are situational in nature and refer to specific dissatisfaction instances rather than complaining in general, they were omitted from further study.

This first phase of the study provided insight into the nature of attitudes toward complaining and revealed a broader scope than encompassed by earlier exploratory work. Five general dimensions of attitudes toward complaining emerged: the extent and nature of affect associated with complaining, objective costs, perceptions of business responsiveness, expectations of societal benefits from complaining, and norms concerning complaining.

Based on the work described above, 31 attitude statements tapping the five domains were developed and administered in a pretest to a convenience sample of 43 student and 14 adult consumers. Redundant items and those with low variance were removed, and a few ambiguous statements were reworded. The remaining 15 items, scaled on a five-point Likert scale from strongly agree to strongly disagree, are shown in Table 1.

TABLE 1

ATTITUDES TOWARD COMPLAINING

In the second phase of the study, this questionnaire was mailed in fall, 1980, under the auspices of a state university to three groups which received identical cover letters and questionnaires. Four hundred were mailed to a random sample of residents of a western urbanized area (including the suburban and rural fringe) with a population greater than 1 million. Questionnaires were also mailed to 212 consumer members of a consumer protection group who resided in the same area and to 198 individuals in the area who had, within the last year, registered a complaint with a government agency or a private consumer protection group. This mailing and a reminder postcard resulted in the response rates listed in Table 2. The overall response rate of 44.0 percent was higher than typical of consumer surveys and was due to greater than average responses from the consumer group members and third party complainers. It is plausible that these two groups have greater interest in consumer issues and would thus be more likely than the population at large to respond to such a survey (Suchman 1962).

TABLE 2

RESPONSE RATES FOR THE SAMPLE

RESULTS AND DISCUSSION

Structure of Attitudes

To determine whether attitudes toward complaining are structured into the five domains hypothesized, a principal components analysis was conducted. Since five domains were pre-specified, five factors were rotated for the initial analysis. Using a varimax solution, the last two factors each had only one item with high loadings, and the eigenvalue for the fifth factor was less than 1.00. Two additional rotations were carried out for three and four factors. The three-factor solution appeared to be most interpretable and is Presented in Table 3. The three factors bear resemblance, but not exact similarity, to the expected domains. No single dimension for affect emerged. Rather, affective items were spread across two of the three factors. Factor one reflects a combination of domains 2 and 3 and part of domain 1 concerning objective costs, business responsiveness, and negative affect, respectively. Essentially, this factor seems to represent consumers' answers to the question, "Is complaining worth the trouble?" The fact that these domains loaded together suggests that instead of examining the costs of complaining separately from the likelihood of remedy, or business responsiveness, the consumer weighs them together in forming an attitude. The second factor reflects norms concerning complaining and follows very closely the hypothesized domain 5. The third factor corresponds to domain 4, societal benefits of complaining.

TABLE 3

PRINCIPAL COMPONENTS ANALYSIS OF ATTITUDE ITEMS

While the results of this analysis provide some support for the structure of attitudes proposed, a note of caution is in order. From this type of study it is impossible to determine whether the listed domains are exhaustive of attitudes toward complaining and adequately represent attitude structure. Although accepted methods for determining domains were followed (see Bohrnstedt 1970; Churchill 1979), it is still possible that important cognitions affecting complaint attitudes and behavior have been omitted.

Attitude/Behavior Relations

The first analysis to assess the usefulness of attitudes toward complaining was a known-groups comparison similar to that performed by Lundstrom and Lamont (1976) in assessing consumer discontent. Factor scores on the attitude scale for the three groups in the sample -- third-party complainers, consumer activists, and the general population were compared using one-way analysis of variance. Results shown in Table 4 indicate that significant differences exist among the three groups for the first and second factors but not the third. For both of the significant analyses, Scheffe tests were performed to determine the nature of these differences. At p < .05, for the first factor the third-party complainer group was significantly different from the other two groups, which were not significantly different from each other. Thus, third-party complainers tend to see the costs of complaining as higher and business responsiveness lower than do other consumers. Given their recent experiences in complaining to businesses, this is not surprising. On the second factor, norms concerning complaining, the general population subsample was significantly more likely to believe that people should not complain when dissatisfied than were the third-party complainers. Members of the consumer protection group were midway between these two groups on the norm factor but did not differ significantly from either. These analyses indicate that different population subgroups do indeed differ in attitudes toward complaining.

TABLE 4

MEANS OF THREE GROUPS ON ATTITUDE FACTOR SCORES

Fishbein (see Fishbein and Ajzen 1975) and others have proposed that beliefs influence attitudes which in turn affect intentions concerning the attitude object. Intentions then affect behaviors. While this formulation is appropriate for examining specific behaviors (e.g., making a complaint in response to a specific dissatisfaction), it requires a slight modification when dealing at the general level (e.g., examining the number or type of complaints a person might make over a period of time). As illustrated in Figure 1, the causal flow for the general case may be defined as follows: beliefs influence attitudes which in turn affect a general tendency (or propensity) to behave in a certain way toward the attitude object. This general tendency further affects actual behaviors, but in any specific instance is moderated by situational variables such as the nature of the specific dissatisfaction, dependence on the product, and difficulty of making a complaint. This framework is used in this study to examine the relationship between complaint attitudes and behaviors.

FIGURE 1

RELATIONSHIPS AMONG COMPLAINT BEHAVIOR CONSTRUCTS

Before presenting the analysis of the relationships described in Figure 1, an elaboration of measures of propensity to complain and complaint behavior used in the study is necessary. Propensity to complain was measured using a six-item scale similar to those used by Bearden et al. (1980) and Robinson (1977), and included such items as, "I am probably more likely to return an unsatisfactory product than most people I know," and "I would attempt to notify store management if I thought service in a store was particularly bad." An earlier administration of the scale to 177 college students yielded acceptable levels of internal consistency (alpha = .79). Alpha for this administration was .74.

To measure complaint behavior, respondents were asked to indicate whether they had performed any of eight complaint behaviors ranging from mild (not leaving a tip at a restaurant when the service was poor) to more extreme (writing a letter of complaint to a business). An interval level Guttman scale was created out of four of these dichotomous items relating to complaints to retailers. The coefficients of reproducibility and scalability were .87 and .58, respectively, meeting the requirements for a scalogram analysis by Guttman (1944) and others (see Edwards 1957).

The relationship between attitudes toward complaining and actual complaint behavior was analyzed using the framework described in Figure 1. First, the association between attitudes and propensity to complain was examined using a multiple regression analysis with propensity as the dependent variable and the three attitude factor scores as independent variables. A significant R of .50 was obtained with all betas significant at the p < .01 level (see Table 5). Thus, individuals who see complaining costs as relatively low and businesses willing to remedy the complaint (worth the effort), who possess a norm positive toward complaining, and who believe that complaining is beneficial for society are more likely to complain than those who do not share these views. The strongest relationship between attitudes and propensity was found for the norms component of attitudes.

TABLE 5

MULTIPLE REGRESSION RESULTS

The relationship between attitudes and actual behavior, as measured by the Guttman scale, was also examined. Because of situational variables intervening between propensity and actual behavior (as noted in Figure 1), this relationship was expected to be weaker than the association between attitudes and propensity. A multiple regression analysis yielded a significant R of .38. All betas were significant at the p < .01 level (see Table 5), except one significant at p < .05. Consistent with the findings pertaining to propensity to complain, factor 2 relating to norms clearly bore the strongest relationship with actual behavior.

Unlike the findings of the preceding analysis, however, factor 1 bore a negative relationship to actual complaint behavior. Those scoring higher on the Guttman complaining scale were more likely to perceive complaint costs as high and business responsiveness lower. This seems to be due to the nature of the study sample; namely, the presence of third-party complainers in the sample influenced this result. When a second regression analysis using only general population and consumer protection group member subgroups was performed, results shown in the two right-hand columns of Table 5 were obtained. A test for equality between the two sets of coefficients (Chow 1960) confirms that third-party complainers should be treated as a separate population for this analysis (F3,350 = 4.76, p < .01). The analysis excluding this group shows no significant relationship between factor 1 and actual complaint behavior. Previous research examining situation-specific variables (Richins 1980), however, has shown that perceptions of costs and benefits of making a complaint about a particular dissatisfaction (situational influences as illustrated at the bottom right of Figure 1) do correlate with complaint behavior. Thus, it appears that for consumer complaints, situation-specific perceptions of costs associated with making the complaint are more potent predictors than are generally held attitudes about complaint efficacy as measured by factor 1 of this data set.

In a final analysis to-determine the relationship between propensity to complain and actual behavior as indicated by the arrow at the right of Figure 1, the propensity scale was correlated with the Guttman scale of complaint behaviors. The resulting correlation of .41 was significant (p < .01) and accounted for about 17 percent of the variance. Unexplained variance is probably due to a number of factors. Measurement error, for one, reduces the size of correlation coefficients. A correction for attenuation (Nunnally 1967) due to imperfect reliability of the propensity scale raises r and r2 to .48 and .23, respectively, though measurement error in the behavioral variable is still not accounted for. The remaining unexplained variance may be due to situation-specific variables described above.

Demographic Correlates of Attitudes Toward Complaining

The survey form included five standard demographic items: sex, marital status, education, age, and income. Few of these showed significant relationships with attitudes toward complaining factors, and those relationships which did emerge were rather weak. Men scored higher on factor 3, believing that complaining has greater benefits to society. This relationship accounted for a very small proportion of variance in attitudes, however (w2 = .014). Age showed a positive relationship with norms concerning complaining (w2 = .08), and education was related to factor 1 such that those with higher education levels were more likely to believe complaining is worth the effort (w2 = .035).

CONCLUSIONS

This study has examined the nature of attitudes toward complaining. Three dimensions emerged: whether complaining tends to be worth the trouble involved (a balancing of objective and psychological costs with perceptions of business responsiveness to complaints), the individual's norms concerning complaining, and perceptions of societal benefits likely to result from complaining.

The relationship between these attitudes and actual complaint behavior was demonstrated. In general, individuals with more positive attitudes toward complaining were shown to possessed greater propensity to complain and reported undertaking more complaint actions. Results of this study combined with previous research (Richins 1980) also suggest that perceptions of some aspects of specific complaint situations may overpower nonspecific attitudes in determining actual complaint behavior.

In reviewing correlational studies such as these, the causal direction of the variables involved may be questioned. It is not clear whether people are more likely to complain because of their pro-complaining attitudes or that people develop their complaining attitudes to maintain consistency with their behaviors, as self-perception theorists would propose. With respect to complaining behavior, it is likely that both cases are operative. Certainly people with unfortunate complaint experiences (such as the third-party complainers in this sample) are likely to develop negative attitudes toward complaining while those who have greater success in registering complaints will tend to have more positive attitudes. Longitudinal research monitoring consumers' attitudes and behaviors over time would help clarify these causal relations as well as reduce the effects of potential self-report bias, which is difficult to control in cross-sectional studies.

Business has much to benefit from customer feedback, both positive and negative. Complaints provide a dual benefit: they may alert the distribution channel that an ongoing problem exists which needs correction, and they provide businesses a second chance to satisfy a customer who will continue patronizing their stores and buying their products. Of the attitudes examined in this study, personal norms about complaining bore the strongest relationship to complaint behavior. This suggests that by improving customers' perceptions of the social acceptability of providing negative feedback to businesses, feedback can be increased. While most businesses are neither equipped nor motivated to undertake large scale efforts to influence such attitudes, efforts on a more limited scale would be relatively inexpensive and easily undertaken.

REFERENCES

Bearden, William, Crockett, Melissa, and Graham Stephen (1980), "Consumers' Propensity to Complain and Dissatisfaction with Automobile Repairs," in Refining Concepts and Measures of Consumer Satisfaction and Complaining Behavior, (eds.) H. Keith Hunt and Ralph L. Day, Bloomington, IN: Indiana University, pp. 35-43.

Bearden, William, Crockett, Melissa, and Teel, Jesse E. (1979), "Alternative Framework for Predicting Consumer Complaining," in 1979 Educators' Conference Proceedings, (eds.) Neil Beckwith et al., Chicago: American Marketing Association, pp. 239-243.

Bearden, William and Teel, Jesse E. (1980), "An Investigation of Personal Influences on Consumer Complaining," Journal of Retailing, 56, pp. 3-200

Bearden, William and Crockett, Melissa (1980), "A Path Model of Consumer Complaint Behavior," in Marketing in the 80's (eds.) Richard P. Bagozzi et al., Chicago: American Marketing Association, pp. 101-104.

Bohrnstedt, George W. (1970), "Reliability and Validity Assessment in Attitude Measurement," in Attitude Measurement, (ed.) G. F. Summers, Chicago: Rand-McNally, pp. 80-99.

Chow, Gregory C. (1960), "Tests for Equality Between Sets of Coefficients in Two Linear Regression," Econometrics. 28, pp. 591-605.

Churchill, Gilbert A. (1979), "A Paradigm for Developing Better Measures of Marketing Constructs," Journal of Marketing Research, 16, pp. 64-73.

Edwards, Allen L. (1957), Techniques of Attitude Scale Construction, new York: Appleton-Century-Crofts.

Fishbein, Martin, and Ajzen, Icek (1975), Belief. Attitude. Intention and Behavior, Reading, MA: Addison-Wesley.

Guttman, Louis (1944), "A Basis for Scaling Quantitative Data," American Sociological Review, 9, pp. 139-150.

Hirschman, Albert 0. (1970), Exit. Voice and Loyalty, Cambridge: Harvard University Press.

Lundstrom, William J., and Lamont, Lawrence M. (1976), "The Development of a Scale to Measure Consumer Discontent," JournAl of Marketing Research, 13, pp. 373-391.

Nunnally, Jum C. (1967), Psychometric Theory, New York: McGraw-Hill.

Richins, Marsha L. (1980), "Consumer Perceptions of Costs and Benefits Associated with Complaining," in Refining Concepts and Measures of Consumer Satisfaction and Complaining Behavior (eds.) H. Keith Hunt and Ralph L. Day, Bloomington IN: Indiana University, pp. 50-53

Robinson, Larry M. (1977), "Consumer Complaint Behavior of New Car Owners: Development and Test of a Theoretical Model," unpublished Ph.D. Dissertation, The Ohio State University.

Rossiter, J.R. (1977), "Reliability of a Short Test Measuring Children's Attitudes Toward TV Commercials," Journal of Consumer Research, 3, pp. 179-184.

Suchman, Edward A. (1962), "An Analysis of Bias in Survey Research," Public Opinion Quarterly, 26, pp. 102-111.

Thomas, William R., and Shuptrine, F. Kelly (1975), "The Consumer Complaint Process: Communication and Resolution," Business and Economic Review, 21, pp. 13-22.

Warland, Rex R., Herrmann, Robert 0., and Willits, Jane (1975), "Dissatisfied Consumers: Who Gets Upset and Who Takes Action," Journal of Consumer Affairs, 9, pp. 148-163.

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