Consumer Problems and Complaints: a National View

Marc A. Grainer, Technical Assistance Research Programs (TARP)
Kathleen A. McEvoy, King Research
Donald W. King, King Research
ABSTRACT - This study reports preliminary findings of a survey of a randomly selected cross section (N=2513) of the nation's households. Research issues addressed include: prevalence and generic types of consumer problems experienced, industries subject to consumer problems, degree of deprivation resulting from consumer problems, and complaining behavior. Results are compared with findings from other relevant studies that address these issues.
[ to cite ]:
Marc A. Grainer, Kathleen A. McEvoy, and Donald W. King (1979) ,"Consumer Problems and Complaints: a National View", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 494-500.

Advances in Consumer Research Volume 6, 1979      Pages 494-500

CONSUMER PROBLEMS AND COMPLAINTS: A NATIONAL VIEW

Marc A. Grainer, Technical Assistance Research Programs (TARP)

Kathleen A. McEvoy, King Research

Donald W. King, King Research

ABSTRACT -

This study reports preliminary findings of a survey of a randomly selected cross section (N=2513) of the nation's households. Research issues addressed include: prevalence and generic types of consumer problems experienced, industries subject to consumer problems, degree of deprivation resulting from consumer problems, and complaining behavior. Results are compared with findings from other relevant studies that address these issues.

Since the early 1970's, the consumer satisfaction/dissatisfaction (CS/D) literature has expanded rapidly. Researchers have reported the results of numerous studies describing causes of (Anderson, 1973; Cohen and Goldberg, 1970; and Suprenant, 1977) and indices of (Day and Landon, 1976; Miller, 1976; and Pfaff, 1977) CS/D. Other investigators have studied levels of CS/D with selected groups of products and services (A. C. Nielsen Co., 1977; Best and Andreasen, 1976; Day and Bodur, 1977; and Handy and Pfaff, 1975).

During the same period, governmental decisionmakers and consumer advocates have become interested in this field of research. Their interest has centered principally on those studies where the findings deal with levels of CS/D by product/service category. Given the finite resources available for consumer affairs programs, effort must be channeled carefully into those problem areas of the greatest need. The level of CS/D with the various product/service groupings has been used in the setting of consumer protection priorities as a principal factor in determining the consumer population's "need for protection'' (Day and Bodur, 1977).

In the past, the complaints received by consumer protection agencies have served as a surrogate measure for the general population's CS/D with products and services. There are, however, two significant problems with this practice. First, only a small percentage of such complaints ever reach governmental agencies or consumer advocacy organizations (Andreasen and Best, 1977; Best and Andreasen, 1976; and Day and Bodur, 1978). Complaints received by private manufacturers and retailers are generally not available for public review. Therefore, only a small, potentially unrepresentative sample of complaints is available for review. Second, the literature suggests that complaints do not constitute a representative sample of those problems experienced by the consumer population at large (Best and Andreasen, 1976; and Warland, et al., 1975).

Survey research techniques offer a much more direct and valid method for determining the level of CS/D with product/service groupings. The survey, with respondents selected on the basis of probability, allows the investigator to determine representative levels of satisfaction and complaint submission. While much survey work has been done in this area, most of the reported studies have concerned rather narrow lists of product/service groupings (A. C. Nielsen Co., 1977; and Handy and Pfaff, 1975). To date, CS/D with comprehensive listings of product/service categories bas been measured in only two sets of studies, the Indiana University (IU) study (Ash, 1978; Day and Bodur, 1977; and Day and Bodur, 1978) and the Center for Study of Responsive Law (CSRL) study (Andreasen and Best, 1977; and Best and Andreasen, 1976).

While each of these studies has made a significant contribution to the field, research design constraints limit, to some extent, the generalizability of their findings. The survey in the IU study utilized a probability sample of dwelling units in a single midwestern city. Therefore, as the researchers point out, generalization of their findings "to the national population would be inappropriate." In the CSRL study, the respondents were asked to evaluate their experiences with 26 products and eight services; a much less comprehensive listing of products and services than the nearly 200 items listed in the IU questionnaire. The CS/D findings are, therefore, limited to these 34 product/ service groupings. Further, the 34 cities in which the CSRL telephone household interviews were conducted were not randomly selected. There was a substantial oversampling of households from six cities, and no rural households were included in the sample. [CSRL presents a persuasive case that this over-sampling of households does not bias the results. A high degree of stability is shown when responses by households from the six oversampled cities are compared with responses by households from the other 28 cities.]

This paper reports the results of a third survey that evaluates consumer experiences with goods and services generally. The sample for this survey consisted of a randomly selected cross-section of the nation's households. No category of product/service was excluded from the interview schedule.

The survey was designed and fielded as one component of a larger study of consumer complaint-handling practices conducted by Technical Assistance Research Programs, Inc. (TARP) under the sponsorship of the White House Office of Consumer Affairs (TAP, P, 1975; TARP, 1976a; TARP, 1976b; and TARP, 1978). The study's questionnaire was designed to provide the consumer affairs community with a data base for priority setting and policymaking.

While the space available for this paper does not allow for a complete listing of the study's findings, the following questions are addressed:

What percentage of consumer households have problems with the goods and services they purchase?

Which generic types of consumer problems (false advertising, incorrect billing, etc.) are the most prevalent?

Which goods and services cause the most consumer problems?

What degree of deprivation is suffered as a result of consumer problems?

What percentage of the consumer households that experienced problems with the goods and services they purchased took some action to obtain redress?

What percentage of the complaints submitted by problem households was "satisfactorily" resolved?

Parallel findings from the CSRL and IU studies are addressed where appropriate. [When comparing TARP findings with CSRL and IU results, the differing time qualifiers used by these studies must be considered. In the TARP study, data were collected for all problems which occurred during the year preceding the interview, regardless of when the "defective" product or service had been purchased. In the CSRL and IU studies, dissatisfaction/problem data were collected for all products or services purchased within a defined timeline. If a consumer purchased an auto four years before the interview, dissatisfaction with a transmission failure that occurred six months before the interview would not be recorded in the CSRL and IU studies. This problem would have been reported by the TARP study. On the other hand, if the auto had been purchased two years before the interview but the transmission had failed 18 months before the interview, the problem would have been reported by IU but not CSRL or TARP. The CSRL study has a smaller dissatisfaction/problem universe than either the IU or TARP studies. The IU and TARP studies have overlapping but different dissatisfaction/problem universes.]

METHODOLOGY

The data base for this study was collected via a nationwide survey consisting of personal interviews conducted with respondents found in a random sample of households. The survey sample was based on a national probability sample of households, consisting of 240 randomly chosen block clusters of households found throughout the continental United States. A sample of household addresses to be interviewed was chosen within each block cluster. A total of 4,327 households was selected by this design. Interviewers were given specific household locations as a starting point for each cluster, and they were instructed to follow a preselected route throughout the cluster. The survey was fielded in February 1976.

Of the 4,327 households chosen, 2,513 persons were interviewed, yielding a completion rate of about 60 percent. (There were 139 vacant homes and these are not included in the sample base.) Households not included consisted of 1,675 nonrespondents; 930 respondents not available for interviews after the initial visit and two callbacks; 714 potential respondents who refused to be interviewed; and 31 respondents who were not interviewed for other reasons. The interviewers questioned male or female heads of households, if available (or other adult household members, if not), about consumer problems they had experienced.

LEVEL OF CONSUMER PROBLEMS

A standard approach to measuring consumer dissatisfaction by survey is the "aided recall" method where respondents are asked how satisfied they were with listed categories of products and/or services purchased within a given time frame. The TARP study followed this approach with one significant modification. Instead of providing a comprehensive listing of product/service categories, respondents were handed a card containing a comprehensive list of consumer problems: ". . " here is a list of some of the problems (list provided in Table 1) that occasionally happen when people have bought or ordered things, received services, or when they try to get something repaired. Have you or anyone in your household had any problems like these in the past year? Which of these problems happened?" To make sure that any problems not on the prepared list were counted, respondents were also asked: "what other problems not on this list happened?" Product/service data were then collected for each respondent-identified problem by asking: "what product or service was involved in this problem?" In sum, the order in which the product/service and type of consumer problem data elements were collected is reversed from the approach applied by the CSRL and IU studies. [This approach was followed for four reasons. First, the identification of generic sets of consumer problems is as important for the setting of consumer protection priorities as is identifying the industries in which the problems occur. This order of questioning allows for direct measurement of the problem dimension without the potentially biasing effect of a preceding industry screening item.

Second, there is a continuing debate in the CS/D literature concerning the inherent subjectivity and, therefore, ambiguity of responses to standard satisfaction survey questions (Hunt, 1977). (In the CSRL study, nearly as many non-price related consumer problems were reported by respondents satisfied with their past year's purchases as by those who were dissatisfied with the preceding year's purchases.) While the problem categories listed in Table 1 are certainly open to subjective interpretation, the ambiguity associated with this formulation is within more limited boundaries than the standard survey questions which are used to operationalize CS/D.

Third, the aim of the study was to identify what the respondents considered to be significant problems. While a "single incident recall" type question might unduly bias responses toward problems with the big ticket items, it was thought that an "aided recall" type question, which provided a comprehensive listing of product/service categories, would not provide enough discrimination between major and minor problems. The method used was a compromise thought to mitigate each of these potential problems.

Fourth, reasons of economy and interviewing technique foreclosed using the IU type of product/service listing. The questionnaire, which included many items not reported in this paper, would have taken too long to administer if the IU approach had been followed.]

Before reporting the findings, it must be noted that this research only touches on the dissatisfaction end of the CS/D continuum. It does not follow that, because X percent of the households reported problems, the remainder of the sample was satisfied with the products and services they had purchased during the preceding year. The feelings of non-problem households probably ranged from mild dissatisfaction to very satisfied. From a consumer protection standpoint, it is the dissatisfaction/problem end of the continuum that is most relevant.

Of the national sample households, 32.4 percent (814/2513) reported consumer problems that had occurred during the previous year. These households reported a total of i,582 problems, for an average of 1.9 occurrences for households reporting problems. The per-sample-household rate of non-price problems reported by the CSRL study of 2.4 (divided nearly equally between "strong" and "weak" problems) is more than three times greater than the 0.7 problems per-sample-household of the TARP study. [Non-price related problems are excluded from the problem base as being more a function of the general rapid escalation in inflation at the time of the CSRL study than of perceived industry malfeasance. "Weak" problems were identified by a probe: "How could it have been better for your household?"; asked of all satisfied purchasers. The "strong" problems, those identified without a follow-up question, are more comparable to the type of significant problems reported in the TARP study.]

TYPES OF CONSUMER PROBLEMS

Table 1 reports the percentages of problem households which experienced the various generic problems during the year preceding the survey. Three out of four of the most prevalent problems concerned basic deficiencies in the quality of products and services purchased. This is consistent with the findings reported by the IU (for services and intangible goods) and CSRL studies. Studies of consumer problems covering more limited product groupings (A. C. Nielsen Co., 1977; and Handy and Pfaff, 1975) also support this trend. [The listing of generic problems as "aided recall" items elicited more detailed problem data in the marketing/business practice area (store did not have product advertised for sale, failure to receive delivery, etc.) than the open-ended follow-up probes utilized in the CSRL study. In the IU study, the "aided recall" probe for generic problems, which was administered as a follow-up question to respondents who were "highly dissatisfied" with a product or service, also elicited detailed marketing/business practice problem data. Such differentiation is useful for the setting of consumer protection policy. This data is of special relevance to agencies with enforcement and regulatory responsibilities.]

PROBLEM INDUSTRIES

The eight industries in which products or services were subject to the most consumer problems were, in descending order of problem prevalence: automobile, appliance (radio and television), all other major and minor appliances, mail service, clothing, telephone, food and household items. In the automobile and appliance industries, less than half of the problems concerned marketing/business practices or basic product quality. Most of the problems reported were with the services provided by repair facilities. Both the CSRL and IU studies report a relatively high prevalence of dissatisfaction/problems in these two industries. [In the automobile and appliance industries, both the CSRL and IU studies report higher dissatisfaction/problem rates with repair services than with the quality of recently purchased items.]

DEGREE OF DEPRIVATION

[Households reporting more than one consumer problem during the year preceding the survey were asked: "I'd like to ask you a few questions about only one of the problems that happened; that is, the most serious problem. Which problem was that?" When only one problem was reported, that problem was considered to be the household's most serious. The data presented in the following sections concern these most serious problems.]

TABLE 1

MOST PREVALENT TYPES OF CONSUMER PROBLEMS

When setting consumer protection priorities, it is not enough to simply identify the most prevalent problem categories or the "offending" industries. In addition, some method must be utilized to weight consumer problems according to their relative "importance." (Given two problems of equal prevalence, a regulatory agency's efforts should be concentrated upon correcting the causes of the more "important" problem.) The CSRL study does this weighting by selecting the products and services to be included in its questionnaire. The IU study asks the respondents in its survey to rate the importance of each item they reported as purchased during the qualifying period. TARP attempted to formulate importance in terms of the actual or potential deprivation suffered as a result of the consumer problem. Three measures of deprivation were used: (1) actual or potential injury occurrence, (2) time lost from work, school, etc., and (3) potential financial loss. [An index of relative problem deprivation, consisting of these and other measures of problem severity, could be developed for use in the setting of consumer protection priorities.]

Of the problem households, 11.5 percent (94/814) reported that their most serious consumer problem had caused or had the potential for causing injury. Nearly 60 percent (55/94) of these households reported the potential for injury.

Of the problem households, 14.0 percent (114/814) reported lost time due to their most serious problem. Having to stay home from work, travel time connected with problem resolution, and waiting for repair/deliverymen were cited as the principal causes for lost time.

Respondents from each problem household were asked: "Would you or anyone in your household have been out of any money if the (most serious) problem was not corrected? How much?" Of the respondents, 61.9 percent (504/814) felt that a financial loss was possible. Of the 504 households that stated there was a potential financial loss, 451 specified the amount that could have been lost. The average potential loss for the 504 households that believed economic deprivation would occur in the absence of problem correction was estimated to be $142. (Using the 814 problem households as a base, and, thereby, including the 310 households that did not foresee their most serious problem resulting in a financial loss, the estimated average loss would have been $88.) This average cost estimate is the upward limit on loss that could be suffered as a result of the respondents' most serious problem. The range of responses was from $1 to $3,000. [Estimating the upward limit of potential economic deprivation on the basis of responses to a single questionnaire item is imprecise business. An extensive series of follow-up questions would be necessary to clarify the ambiguities associated with this finding. (Did respondents include consequential damages in their calculation? Was the measure of loss based on original purchase price, replacement cost, cost of repair? etc.) The potential loss estimate must, therefore, be interpreted cautiously.]

Action to Obtain Redress - Rate of Complaint Submission

With reference to the most serious problem reported, every problem household was asked: "Did you or someone in your household take some action to get this problem corrected or did you let it go and not do anything about it?" This question was worded to elicit reports of complaints (actions seeking problem correction). Sixty-nine percent (562/814) of the problem households submitted (a) complaint(s) in an effort to resolve their most serious problem.

In the CSRL study, a less focused complaint question ("Did anyone in your household do anything about it [the 'strong' or 'weak' problem]?") was used. Unlike the TARP complaint question that was asked in reference to only the household's most serious consumer problem, the CSRL item was used as a follow up for all identified problems, serious or otherwise. CSRL respondents voiced complaints regarding approximately 40 percent of their problem purchases. [Of the responses elicited by the CSRL question, 18.5 percent described such non-complaint actions as "exit" actions (e.g., modifications in buying behavior) and "self-help" (e.g., paying for a repair by independent expert).]

However, when complaint rates were computed for "strong" non-price problems, the subgrouping of CSRL problems most comparable to the TARP study's most serious problem category, the complaint rate increased to 52.0 percent. The CSRL "strong" non-price problem complaint rates for appliance repair, automobile repair, televisions, and automobiles, the products and services reported by the TARP respondent households as being most often subject to problems, ranged from 75.0 percent to 67.6 percent. Therefore. the CSRL data are supportive of the TARP findings. [In the IU study, 78.2 percent of the respondent households reporting dissatisfaction with service or intangible good purchases took some form of action. Of the actions, 53.6 percent of the actions would be classified as complaints according to the TARP definition. (According to the IU nomenclature, these were public actions; split evenly between requests for redress and complaints.) The IU study does not report a ratio between complaining households and problem purchases that is comparable to the complaint rates presented in the CSRL or TARP studies.]

In answer to the question: "Who was contacted first to try to correct the (most serious) problem?", more than 90 percent of problem households which took action reported submission of their complaint to the manufacturer or retailer. Less than 10 percent of the complaints were submitted to government agencies or private consumer advocacy organizations. Both the CSRL and IU studies report a similar reliance on the manufacturer and retailer to resolve consumer complaints.

"SATISFACTORY" RESOLUTION OF COMPLAINTS

Satisfactory resolution was defined from the consumer's point of view. The respondent households which initiated complaint action about their most serious consumer problem were asked: "Which of these statements (card containing statements listed in Table 2 shown to respondent) best describes your feeling about what happened as a result of your efforts to get the problem corrected?"

Table 2 indicates that more than 40 percent of the complaining households obtained a satisfactory resolution (more-than-asked-for, complete-satisfaction, and accept-able-solution assessment categories) to their most serious problem. More than 10 percent of the respondent households obtained "mixed" results (did-get-something assessment category) while more than 40 percent of the complaining households reported completely unsatisfactory results. Nearly 70 percent of the most serious consumer problems reported by the respondent households were, therefore, not satisfactorily resolved. [This rate consists of the sum of the not satisfied complainants (both "mixed" and completely unsatisfactory results) and the households which took no action to resolve their most serious problem divided by the total number of problem households.]

TABLE 2

ASSESSMENT OF EFFORT TO GET PROBLEM CORRECTED

In the CSRL study, 52.7 percent of the complaints about "strong" non-price problems were satisfactorily re- solved. [Satisfaction was determined on the basis of responses to, "What was the result?", an open-ended question. On the basis of the reported method by which responses to this question were coded, the definitions of satisfactory resolution used in the CSRL and TARP studies are, in general terms, comparable.] The rate of satisfactorily resolved complaints for appliance repair, automobile repair, televisions, and automobiles ranged from 48.4 percent to 25.4 percent. The percentage of unremedied "strong" non-price problems reported in these four purchase categories ranged from 55.7 percent to 73.4 percent. These rates of satisfactorily resolved complaints and unremedied problems are consistent with the trends suggested by the TARP study.

SUMMARY OF FINDINGS

The study's principle findings include:

32.0 percent of the national sample experienced one or more consumer problem during the year preceding the study. [Because the study was designed to investigate significant consumer problems, 32.0 percent problem household rate should be viewed as a very conservative estimate of the level of consumer problems generally. The rate most probably would have been higher if the questionnaire had been designed to also identify the more minor types of consumer problems.]

An average of 1.9 problems was experienced by these problem households.

Three of the four most prevalent consumer problems reported concerned basic deficiencies in the quality of products and services purchased.

The majority of consumer problems reported concerned marketing/business practices.

The eight industries subject to the most consumer problems were (in descending order of problem prevalence) the automobile, appliance (radio and TV), all other major and minor appliances, mail service, clothing, telephone, food, and household item industries.

In the automobile and appliance industries, service-related problems predominated.

11.5 percent of the problem households reported that their most serious consumer problem had caused or had the potential for causing injury.

14.0 percent of the problem households reported lost time from work, school, etc. due to their most serious consumer problem.

61.9 percent of the problem households felt that a financial loss, averaging an estimated $142, was possible if their most serious consumer problem was not corrected.

69.0 percent of the problem households submitted (a) complaint(s) in an effort to resolve their most serious consumer problem.

Most complaints regarding a household's most serious consumer problem were submitted to the manufacturer or retailer. Few were submitted to government agencies or consumer advocacy organizations.

More than 40 percent of problem households which initiated complaint action to resolve their most serious consumer problem reported totally unsatisfactory results.

Nearly 70 percent of the most serious consumer problems reported by the national sample were not satisfactorily remedied.

Because of the probability-based sampling design utilized for the study's survey, these findings have generalized applicability to the nation's households.

DISCUSSION OF FINDINGS

When discussing consumer protection issues, the rhetoric of industry often consists of three defensive positions: (1) The number and seriousness of consumer problems suffered by the general population is not significant;

(2) Only a small, vocal minority of consumers complain about the problems they experience with products and services; and (3) The great majority of those complaints about products and services, which are registered, are resolved to the satisfaction of the consumer. [It is often further argued that, even where industry is not totally at fault, a meaningful effort to resolve the problem will be made in order to retain the consumer's good will and future patronage.] The data do not support these contentions. What is suggested, instead, is the need for the private sector to upgrade its complaint-handling capabilities. [Enlightened capitalism argues for business to develop techniques to identify those consumers whose problems have not generated complaints. While such consumers may not be unhappy enough to complain, they may be dissatisfied enough to change brands. Effective consumer problem resolution could prove a useful marketing technique.] How such upgrading may be most effectively accomplished is a worthwhile issue for future research.

The consumer advocacy movement has long argued that there is a need for consumer protection. One policy adopted to protect such consumer interests has been the establishment of governmental agencies and private, voluntary organizations which, among other responsibilities, mediate disputes that consumers and business cannot resolve by themselves. The data document quite well the need for consumer protection. (Consumer problems exist in large numbers, and a sizable percentage of complaints is not satisfactorily resolved.) However, only a small percentage of households which complain about serious consumer problems seeks aid from the newly established third party mediation agents. Pinpointing the explanation for this latter finding is an important topic for future research. Explanations for this finding may include: consumer perception that these mediation agents are ineffective complaint handlers, lack of consumer awareness that these agencies exist, and/or the inability of these organizations to increase their caseloads without additional funding. [TARP's ongoing White House Office of Consumer Affairs survey of consumer complaint-handling practices in governmental agencies and private, voluntary organizations suggests that caseloads have become so heavy that additional funding would be required to significantly increase the volume of complaints presently being handled.]

The CSRL, IU, and TARP studies indicate the feasibility of utilizing survey data as a consideration in setting consumer protection priorities. The survey methodology provides the consumer affairs analyst with a data base generalizable to specified target populations and capable of pinpointing consumer problems by: (1) prevalence, (2) generic type, (3) "offending" industry, (4) importance (degree of deprivation), and (5) population segments at risk. To be put to its best advantage, the survey methodology should be utilized on a longitudinal basis. This methodology is appropriate for monitoring local as well as national consumer problem trends.

Critical review of the differing survey techniques reported in the literature is needed. Operational issues (order of items in questionnaire, telephone vs. personal administration of interview schedule, etc.) must be explored in depth before this social science methodology can be utilized to its fullest potential as an applied public policy technique.

REFERENCES

A. C. Nielsen Company (1977), A New Look at Consumer Attitudes Toward Product Quality.

Anderson, R. (1973), "Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance," Journal of Marketing Research, 10, 38-44.

Andreasen, A. R., and Best, A. (1977), "Consumers Complain - Does Business Respond?" Harvard Business Review, 55, 93-101.

Ash, S. B. (1978) "A Comprehensive Study of Consumer Satisfaction With Durable Products," Advances in Consumer Research, Volume V, ed. H. K. Hunt, Chicago: Association for Consumer Research, 254-262.

Best, A., and Andreasen, A. R. (1976), Talking Back to Business: Voiced and Unvoiced Consumer Complaints, Washington, D.C.: Center for Study of Responsive Law.

Cohen, J. B., and Goldberg, M. E. (1970), "The Dissonance Model in Post - Decision Product Evaluation," Journal of Marketing Research, 7, 315-321.

Day, R. L., and Bodur, M. (1977), "A Comprehensive Study of Satisfaction with Consumer Services and Intangibles," Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, ed. R. L. Day, Bloomington, Indiana: School of Business, Indiana University, 64-74.

Day , R. L., and Bodur, M. (1978), "Consumer Response to Dissatisfaction With Services and Intangibles," Advances in Consumers Research, Volume V, ed. H. K. Hunt, Chicago: Association of Consumer Research, 263-272.

Day, R. L., and Landon, E. L. (1976), "Collecting Comprehensive Consumer Complaint Data by Survey Research," Advances in Consumer Research, Volume III, ed. B. B. Anderson, Chicago: Association for Consumer Research, 263-268.

Handy, C. R., and Pfaff, M. (1975), Consumer Satisfaction with Food Products and Marketing Services, Agricultural Economic Report No., 281, Washington, D.C.: Economic Research Service, United State Department of Agriculture.

Hunt, H. K., ed. (1977), Conceptualization and Measurement of Consumer Satisfaction and Dissatisfaction, Cambridge, Massachusetts: Marketing Science Institute.

Miller, J. (1976), "Exploring Some Alternative Measure of Consumer Satisfaction," Marketing: 1776-1976 and Beyond, ed. K. L. Bernhardt, Chicago: American Marketing Association, 661-664.

Pfaff, M. (1977), "The Index of Consumer Satisfaction: Measurement Problems and Opportunities," The Conceptualization of Consumer Satisfaction and Dissatisfaction, ed. H. K. Hunt, Cambridge, Massachusetts: Marketing Science Institute, 36-71.

Suprenant, C. (1977), "Product Satisfaction as a Function of Expectations and Performance," Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, ed. R. L. Day, Bloomington, Indiana: School of Business, Indiana University, 36-37.

TARP (1975), First Portion of Federal Central Office Phase of Contract No. HEW-OS-74-292, Feasibility Study to Improve Handling of Consumer Complaints: Evaluation and Inventory Reports, NTISPB-244182, Washington, D.C.: White House Office of Consumer Affairs.

TARP (1976a), Second Portion of Federal Central Office Phase of Contract No. HEW-OS-74-292, Feasibility Study to Improve Handling of Consumer Complaints: Evaluation and Inventory Reports, NTIS-PB-263081, Washington, D.C.: White House Office of Consumer Affairs.

TARP (1976b), A National Survey of the Complaint-Handling Procedures Used By Consumers, NTIS-PB-263082 Washington, D.C.: White House Office of Consumer Affairs.

TARP (1978), Consumer Problems, Consumer Complaints, and the Saliency of Consumers Advocacy Organizations: A Review of the Literature and the Results of National Survey, Washington, D.C.: National Science Foundation (forthcoming).

Warland, R. H., Herrmann, R. O., and Willitts, J. (1975), "Dissatisfied Consumers: Who Gets Upset and Who Takes Action," The Journal of Consumer Affairs, 9, 148-163.

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