, Pages 254-262
(student), Indiana University
A preliminary analysis of results from the consumer durables section of a broader study directed by Professor Ralph Day of Indiana University is presented. The results reflect wide variations over the 67 durable product classifications in: (1) the sample proportion reporting a purchase; (2) the relative importance of the items to users; and (3) the patterns of reported satisfaction and dissatisfaction. [The data for this study were provided by Professor Ralph Day, Indiana University. The author is grateful to Professor Day and to the reviewers for their comments on the original draft of this paper.]
The concern of consumer advocates, policy makers, and the consumer public with the apparent general rise in consumer dissatisfaction has led to a call for more and better research focusing on the products and services which seem to be responsible for the increase in the number of dissatisfied consumers. The design and implementation of "effective" consumer protection programs must overcome two basic constraints: insufficient funding and inadequate information. The first limitation suggests that since policy makers have finite resources with which to support their programs, they must attempt to discern the most useful way in which satisfaction/dissatisfaction (CS/D) research should be conducted. The central decision focuses on what kinds of data should be collected to facilitate the design of programs concerned with consumer satisfaction/dissatisfaction. Also, researchers may consider the related issue of aggregation. In assessing consumer dissatisfaction, should the unit of investigation be at the level of the individual, some particular segment of the consuming population or aggregated data on the entire population? Aggregate analysis is governed by data which may provide only very rough approximations of CS/D trends throughout society. Analysis of disaggregated data, while more costly and time-consuming, promises further understanding of the underlying determinants of consumer satisfaction/dissatisfaction. Although virtually all CS/D research to date has relied on individual measures, there have been some efforts to develop overall indexes of consumer satisfaction (Handy, 1976; Handy, 1977; Lingoes and Pfaff, 1972; A. Pfaff, 1972; M. Pfaff, 1976). It is not clear to what extent, if any, the index of consumer satisfaction (ICS) facilitates the actual decision-making process of policy makers. For example, at the industry level, there appears to be some reservation about the willingness of managers to rely on ICS as a decision-making tool (Pratt, 1972).
The adequacy of volunteered complaint information as a basis for assessing consumer dissatisfaction has received recent attention (Day and Bodur, 1977). They suggest that complaint letters may be misleading since they tend not to be representative either of the types of problems confronting consumers or of the types of people experiencing consumer problems. No further elaboration on specific shortcomings of volunteered complaint data (for example, "big ticket" bias) is provided here since these issues have been discussed elsewhere (Day and Bodur, 1977).
Policy makers require a workable basis for setting priorities and planning consumer protection programs. They depend on the availability of the kinds of information which will permit them to evaluate, systematically, existing levels of dissatisfaction with products and services. A recent study (Day and Bodur, 1977) has suggested that at least four kinds of information are required for diagnostic purposes:
1. Number of users in the population.
2. Proportion of users expressing dissatisfaction with the item.
3. The degree of importance associated with the consumption of a product or service.
4. How consumers resolve their dissatisfaction with an item.
They contend that these elements may provide policy makers with a basis for assessing the "need for protection" by consumers of particular goods and services.
PURPOSE OF THE STUDY
The primary contribution of this study consists of reporting preliminary results of analysis on durable goods data obtained by means of a survey research instrument which was developed as a part of a larger research project (Day and Bodur, 1977; Day and Landon, 1976). The current findings may be put into perspective if assessed in relation to the underlying objectives of the parent project. The fundamental goal of the project was the design of a research method which would furnish the kinds of information essential for evaluating consumer satisfaction/dissatisfaction over a comprehensive set of products and services. The operational objective was the development of a product-by-complaints matrix based on "better quality data (than volunteered complaints or 'single incident recall' can give us) of a descriptive nature" (Day, 1977). The matrix would be used by policy makers as a tool for diagnosing consumer dissatisfaction. [For further details on the parent project, see reports on 1974 study, Day and Landon, 1976; Day and Bodur, 1977.] At least the following issues were of interest to the designer of the research instrument:
* How and from whom should the data be collected?
* How do volunteered data provided by self-selected individuals differ from those obtained from a scientifically-designed probability sample? Which set provides more representative information about consumer problems over the entire population?
* If these data are collected from a probability sample, how free are they from "big ticket" bale and/or overpowering "demand characteristics'' of the data collection procedures?
In general, the relevant information required will consist of measures: (1) of the fraction of the population using the item and (2) of the distribution of users over the satisfaction/dissatisfaction scale for each separate product/service category. This paper examines the possibility that certain "patterns" reflecting different levels of satisfaction/dissatisfaction may exist for consumers over the various kinds of durable goods. These patterns, if confirmed, might increase the effectiveness of consumer protection programs by enabling policy makers to focus their attention on products with the highest rates of dissatisfaction among users.
The data for this research were obtained as part of a survey research project which utilized a set of data collection instruments designed for a larger national project currently in preparation. The actual data for the current study were acquired from self-administered questionnaires which were distributed through the "drop-off/pick-up" method to a probability sample of more than 600 dwelling units in one Midwestern city during the fall of 1976. Th instrument was extensively pre-tested. [For details on instrument design and pre-testing, see Day and Bodur, 1977.]
The questionnaires were quite lengthy, probing both consumer satisfaction/dissatisfaction and complaint behavior on an "aided recall" basis over approximately 200 categories of products and services. The preliminary findings which will be reported here relate exclusively to the usage rate/importance/satisfaction portion of the survey instrument which covered 67 categories of durable products. Items which were believed to be purchased and used by consumers in similar ways were grouped together into categories.
The initial task required respondents to indicate whether or not they had purchased any items from the category over a three-year recall period. Those who had made such a purchase were then asked to provide an indication of the relative importance of that category and of the relative extent of satisfaction or dissatisfaction with items contained in the category. The responses available to respondents for each item were as follows:
Purchase (?) and Importance of these products
N = I have not bought this product (any of these products) within the past three years.
L = This product is (these products are) of less importance to me or my family than some of the other durables we own.
H = This product is (these products are) highly important to me or my family.
How satisfied or dissatisfied I am with them
1 = I am quite satisfied with this product (these products).
2 = I am somewhat satisfied with this product (these products).
3 = I am somewhat dissatisfied with this product (these products).
4 = I am quite dissatisfied with this product (these products).
The instrument, itself, was designed to furnish more complete information than is available from volunteered complaint data. Also, it sought to avoid or reduce the bias arising from "single incident recall" surveys which force respondents to limit their responses to one particular experience. By providing data on all products the respondent has purchased over a specified period of recall, the instrument provides a basis for assessing the extent of consumer satisfaction/dissatisfaction with items in each separate product/service category. Another advantage is that it enables the researcher to tap the positive as well as the negative elements of consumers' reactions to a comprehensive set of durable goods.
One familiar approach to the study of CS/D has led to operationalizing the measure of dissatisfaction as the discrepancy between performance expectations for the product or service and actual outcomes (Anderson, 1973; Cardozo, 1965; Day, 1975; Day, 1977; Miller, 1976; Olshavsky and Miller, 1973; Summers and Granbois, 1976; Swan and Combs, 1976). Concern about expectations and ultimate performance has raised important methodological issues regarding their measurement. For example, expectations may be operationalized to include: (1) expectations of product performance, (2) expectations of costs and, (3) expectations of the impact on others (Day, 1977). The relevant research question is whether all sources of expectation need to be assessed simultaneously in order to obtain a representative overall measure of the construct. If so, how is each component of overall expectation to be weighted? Unfortunately, the measurement of perceived product performance is also troublesome. For example, one approach has underscored the need to examine the influence of both physical and psychological dimensions of product performance on consumer satisfaction (Swan and Combs, 1976). The rationale for the research instrument employed in this study is that it lets the respondent define what he means by "satisfied" or "dissatisfied." The subject determines for himself what the points on the scale mean. This enables the researcher to evaluate the pattern of self-defined satisfaction/dissatisfaction over respondents for each product category.
The questionnaire also contains the usual demographics and a set of 14 Likert-type attitude/opinion scales designed to elicit subjects' impressions of business, government and marketing. The balance of the paper will be devoted to the preliminary analysis conducted on the category-by-category responses obtained from 119 usable questionnaires.
Data on usage rate, importance, and extent of satisfaction/dissatisfaction are shown in percentages. To examine the associations between the satisfaction/dissatisfaction scores and demographic data, correlation analysis was used. Since the satisfaction/dissatisfaction scales probably should be considered ordinal in nature, use of a nonparametric correlation coefficient seems appropriate. The data consisted of a substantial number of cases classified into comparatively few categories. Since they contained a large number of tied ranks, the Kendall (tau-C) correlation coefficient was preferred to Spearman's rank order correlation coefficient (Nie, Hull, Jenkins, Steinbrenner and Bent, 1975).
During the preliminary analysis, attitude and selected demographic variables, which might be considered as interval-scaled data, were regressed, in step-wise fashion, onto the satisfaction scores. Since examination of predicted criterion scores versus residuals suggested that the data were not suitable for multiple regression analysis, the results are not presented here. Other multivariate techniques may be applied as the analysis of the results proceeds.
To demonstrate the kind of information contained by the data and its potential for use by policy makers, two sets of results are presented. First, the category-by-category responses denoting purchase, relative importance and satisfaction/dissatisfaction are summarized and briefly analyzed. Next the relationships between demographics and satisfaction are explored.
Product Category Analyses
Tables 1 through 4 contain percentage summaries on 67 categories of durable goods which are grouped in four sections of the durable products questionnaire: (1) housing and home furnishings; (2) appliances and personal care equipment; (3) durable items for entertainment, recreation and education; and (4) cars and other transportation durables. All four tables provide information obtained from 119 usable questionnaires. The first column in each table shows the percent of respondents who reported they had purchased the item within the three-year recall period. The next column shows the percentage of purchasers rating the item as relatively important. The rank order of percent rating the item as important is also provided. This is followed by the percentage distribution of satisfaction/dissatisfaction ratings for each category. As stated earlier, a primary objective of the research plan is to provide a basis for classifying items according to the severity of consumer problems they engender. The classification task is simplified if one relies on a combined measure of dissatisfaction for every product category under investigation. This information is covered by the final two columns of Tables 1 through 4 which show the sum of the two dissatisfaction ratings and the rank order of categories by percent dissatisfied.
In Table 1, the percentage of subjects who reported that they had purchased one or more items in any "Housing and Home Furnishings" category varies widely. Only 1.7% of the sample indicated a purchase of "swimming pool" while 88.0% said they had purchased "draperies, linens, etc." Only three other categories reveal more than 50% of respondents indicating purchase within the specified period of recall: "sofas, chairs"; "housewares, tableware, etc."; and "lamps, clocks, etc." The percentage of consumers who indicated that the various items were highly important to them or their families ranges from 35.4% for "electric blankets" to 100.0% for "single family house" and "heating installation." The percent of buyers who rated an item in one of the two dissatisfaction categories varies from 6.8% for "lamps, clocks, etc." to 37.5% for "mobile home."
A key assumption of the research plan is that the amount of dissatisfaction expressed for an item should be brought into perspective by considering it in relation to the proportion of the population using the product. For example, when a very small segment of the population indicates purchase of an item (e.g., 1.7% for "swimming pools"), reported rates of dissatisfaction probably do not have statistical reliability. It may seem intuitively logical to expect higher rates of reported dissatisfaction for widely-owned items than for products which are used by a relatively small segment of the population. The fallacy of this reasoning is exposed if one compares the rates of usage and dissatisfaction for "mobile homes" with those of "draperies, linens, etc." in Table 1. The category "draperies, linens, etc.", with a reported purchase rate of 88.0%, ranks fourteenth out of twenty according to the percentage of dissatisfied users. On the other hand, the greatest rates of dissatisfaction with "Housing and Home Furnishing" items appear to be associated with those purchased by a comparatively small percentage of the entire population. For example, "mobile homes" are ranked first by percent of dissatisfied buyers even though less than 10% of respondents reported purchase of the product. The number two item according to dissatisfaction is "swimming pools," another product characterized by limited ownership. In Table 4, 36.2% of respondents reported purchase of a used car within the recall period. It seems that approximately 41% of the used car owners were dissatisfied with the product. The apparent high rate of dissatisfaction with used cars reported in this study is corroborated by the amount of dissatisfaction with the item expressed in conventional complaint statistics. In Table 2, the rates of dissatisfaction reported for "Appliances and Personal Care Products" may be compared to those furnished by another study (Mason and Himes, 1973). It is not surprising that higher levels of dissatisfaction were expressed by respondents in the current study. In their research, Mason and Himes (1973) defined a complaint as "any dissatisfaction with an appliance purchased and for which a consumer seeks relief from an outlet in the channel of distribution." In this study, the survey instrument enabled subjects to register dissatisfaction with an item independent of specific complaint behavior. Comparison of the results of the two studies suggests that the differences in percentage reporting dissatisfaction may be ascribed to the number of dissatisfied consumers who do not undertake any formal search for redress. Table 3 reveals a similar pattern of results and will not be discussed here. The preceding examples are meant to suggest that greater understanding of the kinds of problems which consumers experience is possible if the number of dissatisfied users is considered in relation to the total number of users.
Relationship of Satisfaction and Demographics
One premise for this study is that writers of complaint letters are probably not representative of the full range of people experiencing dissatisfaction with products and services (Day and Landon, 1976; Day and Bodur, 1977; Warland, 1975). Here, the unit of interest is any consumer who feels dissatisfied with his purchase of a durable product, not a complainer per se. The central question is: do dissatisfied users of particular items have anything in common? This question implies that it may be useful to investigate relationships between demographic variables and satisfaction scores over a comprehensive set of durable products. Tables 5 through 8 contain correlation coefficients measuring the extent and direction of the relationships between demographics and satisfaction. Since the entire sample consisted of 119 respondents, cell sizes were checked to minimize the possibility of reporting spurious results. This examination revealed that the response distributions for the education and income variables were skewed slightly to the right of normal. All coefficients reported are significant at the .05 level.
Table 5 contains the correlation coefficients for the "Housing and Home Furnishings" section of the questionnaire. All of the demographic variables are significantly correlated with at least three of the durable product categories. Examination of the table shows varying patterns of relationships between demographics and satisfaction. Although none of the coefficients are large, most of the categories exhibit some relationship between one or more demographic variables and the satisfaction scores. The category with the largest number of significant relationships is the "single family house." On the other hand, no significant relationships appear to exist between demographics and satisfaction for "mobile home" owners. The variables with the largest number of significant relationships were marital status, employment status, home ownership and income. The number of significant relationships for each demographic variable varied widely over the groups of durable products shown on Tables 5 to 8. Another study (Day and Bodur, 1977) examined the relationships between demographic variables and satisfaction scores over a broad set of consumer service categories. They reported:
1. weak associations between demographics and responses on satisfaction;
2. widely varying patterns of significant relationships over these categories;
3. situational characteristics (e.g., marital status, home ownership) were more consistently related to satisfaction scores than most personal characteristics (e.g., education and sex).
The results of the present study parallel the earlier findings of Day and Bodur, 1977.
The sample statistics reported here are based on data which were obtained in one Midwestern city during the fall of 1976. Since a scientifically-determined probability sample was obtained, the results are probably representative of the sample population. However, any generalization of these results to any other population would be inappropriate.
This study is intended to stimulate more research into the underlying determinants of consumer satisfaction/dissatisfaction with durable products. Some fruitful areas for future research on durable goods may focus on:
(1) the role of product cost as a basis for predicting consumer satisfaction/dissatisfaction with an item;
(2) the extent to which consumer expectations can be manipulated by manufacturers or government policy makers;
(3) the identification and measurement of any "carry-over effect" associated with purchase satisfaction, and
(4) the impact of contemporary advertising on the gap between consumer expectations and perceived performance for widely-used durable products.
This paper has reported some of the preliminary results of analysis on data collected as part of a larger research project still in progress. These findings cover the section of the questionnaire which elicited satisfaction/ dissatisfaction scale responses. On the basis of these results, the discriminating power of the survey instrument as a tool for diagnosing consumer dissatisfaction seems to have been satisfactorily demonstrated. The questionnaire was designed to furnish information on rate of purchase, relative importance and satisfaction/dissatisfaction over 67 categories of durable products. The results showed widely varying relationships between the percent of a sample reporting purchase of a product and the percent of users experiencing dissatisfaction with the item. The relationship between demographic variables and satisfaction were generally weak and varied widely over product categories. On a category-by-category basis, however, interesting relationships between demographics and satisfaction scores were revealed.
The results reported here appear to have face validity. It is hoped that they will demonstrate the usefulness of the information conveyed by the data which was obtained by the survey instrument. There is room for much more research into consumer satisfaction/dissatisfaction with durable products.
SATISFACTION/DISSATISFACTION RATINGS: HOUSING AND HOME FURNISHINGS
SATISFACTION/DISSATISFACTION RATINGS: HOME APPLIANCE AND PERSONAL CARE PRODUCTS
SATISFACTION/DISSATISFACTION RATINGS: DURABLE ITEMS FOR ENTERTAINMENT, RECREATION, AND EDUCATION
SATISFACTION/DISSATISFACTION RATINGS: CARS AND OTHER TRANSPORTATION DURABLES
KENDALL'S TAU (C) CORRELATION COEFFICIENTS - DEMOGRAPHICS WITH SATISFACTION SCORES: HOUSING AND HOME FURNISHINGS
KENDALL'S TAU (C) CORRELATION COEFFICIENTS - DEMOGRAPHICS WITH SATISFACTION SCORES: HOME APPLIANCE AND PERSONAL CARE PRODUCTS
KENDALL'S TAU (C) CORRELATION COEFFICIENTS - DEMOGRAPHICS WITH SATISFACTION SCORES: DURABLE ITEMS FOR ENTERTAINMENT, RECREATION, AND EDUCATION
KENDALL'S TAU (C) CORRELATION COEFFICIENTS - DEMOGRAPHICS WITH SATISFACTION SCORES: CARS AND OTHER TRANSPORTATION DURABLES
Rolph E. Anderson, "Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance,'' Journal of Marketing Research, February 1973, pp. 38-44.
Alan R. Andreasen, "A Taxonomy of Consumer Satisfaction/ Dissatisfaction Measures," A paper presented at a workshop on the Conceptualization and Measurement of Consumer Satisfaction and Dissatisfaction, sponsored by the Marketing Science Institute (MSI) and National Science Foundation (NSF), April 1976, Chicago, Illinois.
Richard Cardozo, "An Experimental Study of Customer Effort, Expectations and Satisfaction," Journal of Marketing Research, Vol. 2, August 1965, pp. 244-249.
John A. Czepiel and Larry J. Rosenberg, "The Study of Consumer Satisfaction: Addressing the 'So What' Question," A paper presented at MSI and NSF Workshop, April 1976, Chicago, Illinois.
Ralph L. Day, "Extending the Concept of Consumer Satisfaction,'' Advances in Consumer Research, Vol. IV, 1977 (forthcoming).
Ralph L. Day and Muzaffer Bodur, "A Comprehensive Study of Satisfaction with Consumer Services," in Ralph L. Day, editor, Consumer Satisfaction/Dissatisfaction and Complaining Behavior, Division of Research, Indiana University, 1977, pp. 64-74.
Ralph L. Day and E. Laird Landon, "Collecting Consumer Complaint Data by Survey Research," in Beverlee B. Anderson (ed.), Advances in Consumer Research, Vol. III, 1976, pp. 263-268.
Charles R. Handy, "Monitoring Consumer Satisfaction with Food Products," A paper presented at MSI and NSF Workshop, April 1976, Chicago, Illinois.
Charles R. Handy, "Indexes of Consumer Satisfaction with Food Products: 1974 and 1976 Survey Results," A paper prepared for the Research Symposium on Consumer Satisfaction/Dissatisfaction and Complaining Behavior, Indiana University, April 21-22, 1977.
Donald J. Hempel, "Consumer Satisfaction with the Home Buying Process: Conceptualization and Measurement," A paper presented at MSI and NSF Workshop, April 1976, Chicago, Illinois.
James C. Lingoes and Martin Pfaff, "The Index of Consumer Satisfaction: Methodology," Proceedings, Association for Consumer Research, 1972, pp. 689-712.
William J. Lundstrom and Lawrence M. Lamont, "The Development of a Scale to Measure Consumer Discontent," Journal of Marketing Research, Vol. XII, November 1976, pp. 373-381.
Joseph B. Mason and Samuel H. Himes, Jr., "An Exploratory Behavioral and Socioeconomic Profile of Consumer Action About Dissatisfaction with Selected Household Appliances," Journal of Consumer Affairs, Vol. 7 (Winter 1973), pp. 121-127.
John A. Miller, "Studying Satisfaction: Modifying Models, Eliciting Expectations, Posing Problems, and Making Meaningful Measurements," A paper presented at MSI and NSF Workshop, April 1976, Chicago, Illinois.
Normal H. Nie, G. Hadlai Hull, Jean G. Jenkins, Karin Steinbrenner and Dale H. Bent, Statistical Package for the Social Sciences, 2nd ed., New York: McGraw-Hill Book Co., 1975.
Richard W. Olshavsky and John A. Miller, "Consumer Expectations, Product Performance and Perceived Product Quality," Journal of Marketing Research, Vol. 9, February 1973, pp. 19-21.
Anita B. Pfaff, "The Index of Consumer Satisfaction," Proceedings, Association for Consumer Research, 1972, pp. 713-737.
Martin Pfaff., "The Index of Consumer Satisfaction: Measurement Problems and Opportunities," in H. Keith Hunt, (ed.), Advances in Consumer Research, Vol. V, Association for Consumer Research, 1978, forthcoming.
H. B. Pickle and R. Bruce, "Consumerism, Product Satisfaction-Dissatisfaction: An Empirical Investigation," Southern Journal of Business, Vol. 7 (November 1972), pp. 87-100.
Robert W. Pratt, Jr., "The Index of Consumer Satisfaction and Corporate Marketing Policy," Proceedings, Association for Consumer Research, 1972, pp. 742-745.
John O. Summers and Donald H. Granbois, "Predictive and Normative Expectations in Consumer Dissatisfaction and Complaining Behavior," A paper presented at ACR Conference, October 1976, Atlanta, Georgia.
John E. Swan and Linda J. Combs, "Product Performance and Consumer Satisfaction: A New Concept," Journal of Marketing, Vol. 40, April 1976, pp. 25-33.
Rex H. Warland, Robert O. Herrmann and Jane Willits, "Dissatisfied Consumers: Who Gets Upset and Who Takes What Action," Journal of Consumer Affairs, Winter 1975, pp. 148-163.