Abstract
US consumer sentiment toward the practice of marketing has been measured via annual survey for over two decades in this continuing study. The following report presents the measuring instrument, a test of the measurement model, and some updated findings. Examination of selected demographic variables indicates that they are only marginally related to the consumer sentiment construct. Yet the enduring import of the project is description of the secular trend of aggregate consumer attitude toward marketing in the context of the world’s largest national economy.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In the future, the retailing and product items will be updated to capture changes in the consumer landscape such as more service categories and online purchasing. To maintain continuity, parallel measures and re-indexing will be needed. For now, nothing prevents consumer respondents from interpreting “stores” as service providers, which actually is consistent with modern parlance.
The weighting procedure is retained for conceptual authenticity and psychometric conservatism based upon the Wilkie-Pessemier advice (1973, p. 436), though use of the measure unweighted produces negligible difference in results. Weight standardization, as suggested by Bass and Wilkie (1973), also does not materially alter the results. Index numbers are calculated as
$$100 \times \left( {{\text{raw score difference from measure's lower limit}}\,\, \div \,\,1984{\text{ raw score difference from lower limit}}} \right).$$Some other subtleties of operationalization are discussed in Appendix B.
Tests of reliability and convergent and discriminant validity were undertaken using the generally accepted techniques at the onset of the project 20 years ago. This manuscript relies on contemporary methods and more extensive data to modernize the original validation.
This internal consistency test, in the confirmatory factor analytic context, is sometimes erroneously termed convergent validity, though it is based upon similar ingredients or items within a given measure rather than maximally different multiple measures (Campbell and Fiske 1959) as required for true convergent validation.
For brevity, and perhaps the reader’s relief, only Y2004 and 1984–2004 pooled results are reported for the composite reliability and variance extracted segments.
Geopolitical events may initially seem remote to the subject matter but would have direct economic consequences, even for the economic actors known as consumers. The primary transmission mechanism could be through uncertainty which affects risk-adjusted returns in consumer, as well as financial markets and in turn consumer sentiment. For specific example, the US final victory in the Cold War produced an immediate “peace dividend” in the early 1990s, involving a reallocation of national income away from military investment and toward consumption.The main discontinuity in the ICSM era is 1991. Dummy variable analysis decomposing the ICSM time series into pre- and post-1991 segments finds a significant improvement in trend for the 1991–2004 period (b = 1.95, p = .0006), possibly attributable to the signal events contemporaneous with the 1991 turning point, e.g., demise of communism.
Even visual inspection of Figure 3 reveals which components contribute most to the overall index movements reviewed here. For instance, the 1983, 1990 and 1994 upturns are driven primarily by product and price (along with retail service in 1990). The 1987 and 1993 declines and the 1999 recovery are common to all elements. Year-over-year significant differences are summarized elsewhere in “Longitudinal Findings.”
Although the pairwise analysis does not conform to the assumptions of the Duncan test (Klockars and Sax 1986, pp. 47–50), the concept of “experimentwise” a-error rate is still meaningful. Here it is .33, i.e., a 33% probability that at least one of the significant differences is due to chance. Most of the coefficients listed would not be significant per the stricter, but low-power, Scheffé correction, yet the series analysis is offered as the best available summary portrait of the trend highlights. The 1984-2001 difference would be significant under any of the widely used adjustments, including Scheffé.
For consistency with the originally published method and results, and primarily in the interest of offering a meaningful metric (the −200/+200 scale), reported scores are the simple weighted-sum composites of scale items as described in section “ICSM Measurement,” rather than the latent underlying factors (Σξ, η) assimilated from LISREL factor loadings (λ). Technical import is nil because of the high correlation between measures computed from the two approaches (r ≈ .99). A convenient benefit is that the accompanying index numbers apply equally to both the raw scores and the underlying construct. In later analysis, estimates of the pure constructs (ξ, η) are used.
The experimentwise or cumulative a-error rate for the enumerated set of comparisons for the product and price scales is .37.
For this initial phase of demographic analysis, tests reported involve only the contemporary 2004 data, rather than those from the prior 1988–2003 period. Though the analysis is descriptive, it is exploratory in spirit, to be followed by tests involving the long-term data. Also, for this and subsequent analysis, the synthesized latent factors (ξ, η,) are used for ICSM and its four elements rather than the indicator-based measures consisting of composites of x-scores. Rationale was given previously (see footnote 8).
LISREL structural equation modeling (S.E.M.) could have been used to relate the objective demographic measures to the four components of ICSM, but not the ICSM construct itself which has no independent indicators apart from the component factors (had ξ constructs been modeled as reflective facets such a model would have been S.E.M.-compatible, but not conceptually genuine). Therefore, to avert inconsistent treatment within the same analysis (Table 10) – which would have yielded negligible gain in precision – the latent variables are synthesized from the indicators (x) and coefficients (λ) derived from the analysis described in Figure 1 and Tables 3, 4, 5, 6 and 7.
Five-point scale response options labeled from “Agree Strongly” to “Disagree Strongly,” R: reverse scored. Asterisk indicates item retained after purification.
References
Atuahene-Gima, K., & Murray, J. Y. (2004). Antecedents and outcomes of marketing strategy comprehensiveness. Journal of Marketing, 68, 33–46.
Barksdale, H. C., & Darden, W. R. (1972). Consumer attitudes toward marketing and consumerism. Journal of Marketing, 36, 28–35.
Barksdale, H. C., Darden, W. R., & Perreault Jr., W. D. (1976). Changes in consumer attitudes toward marketing, consumerism, and government regulation: 1971–1975. The Journal of Consumer Affairs, 10, 117–139.
Barksdale, H. C., & Perreault Jr., W. D. (1980). Can consumers be satisfied? MSU Business Topics, 28, 19–30.
Barta, P., & Chaker, A. M. (2001). Consumers voice rising dissatisfaction with companies. The Wall Street Journal, A2, A13 (May 21).
Bass, F. M., & Wilkie, W. L. (1973). A comparative analysis of attitudinal predictions of brand preference. Journal of Marketing Research, 10, 262–269.
Boshoff, C., & DuPlessis, A. P. (1991). South African consumer sentiment toward marketing. Port Elizabeth, South Africa: Vista University. Working paper.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait–multimethod matrix. Psychological Bulletin, 56, 81–105.
Carsky, M., Dickinson, R. A., & Canedy III, C. R. (1998). The evolution of quality in consumer goods. Journal of Macromarketing, 18, 132–143.
Clymer, A. (1983). How Americans rate big business. The New York Times Magazine, 69, (June 8).
Davis, H. L. (1970). Dimensions of marital roles in consumer decision-making. Journal of Marketing Research, 7, 168–177.
Davis, H. L., & Rigaux, B. P. (1974). Perception of marital roles in decision processes. Journal of Consumer Research, 1, 51–62.
Fama, E. F., & MacBeth, J. D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636.
Farmer, R. N. (1967). Would you want your daughter to marry a marketing man? Journal of Marketing, 31, 1–3.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press.
Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E. (1996). The American Customer Satisfaction Index: Nature, purpose, and findings. Journal of Marketing, 60, 7–18.
Francese, P. (2003). Consumers today. American Demographics, 25, 28–29.
Gaski, J. F., & Etzel, M. J. (1985). A proposal for a global, longitudinal measure of national consumer sentiment toward marketing practice. In E. Hirschman & M. Holbrook (Eds.) Advances in consumer research (pp. 65–70). Provo, UT: Association for Consumer Research.
Gaski, J. F., & Etzel, M. J. (1986). The index of consumer sentiment toward marketing. Journal of Marketing, 50, 71–81.
Gaski, J. F., & Etzel, M. J. (2005). National aggregate consumer sentiment toward marketing: A thirty-year retrospective and analysis. Journal of Consumer Research, 31, 859–867.
Hatcher, L. (1994). A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, NC: SAS Institute.
Hilsenrath, J. E. (2001). Shopper satisfaction takes another hit. The Wall Street Journal, A2, A4 (November 19).
Hoyer, W. D., & MacInnis, D. J. (2004). Consumer behavior (3rd ed.). Boston: Houghton Mifflin.
Hustad, T. P., & Pessemier, E. A. (1973). Will the real consumer activist please stand up: An examination of consumers’ opinions about marketing practices. Journal of Marketing Research, 10, 319–324.
Klockars, A. J., & Sax, G. (1986). Multiple comparisons. Beverly Hills, CA: Sage.
Lundstrom, W. J., & Lamont, L. M. (1976). The development of a scale to measure consumer discontent. Journal of Marketing Research, 13, 373–381.
Magnuson, W. G., & Carper, J. (1982). Caveat emptor. In D. Aaker, & G. Day (Eds.) Consumerism (pp. 267–278). New York: Free.
Morrow, J. (2003). A place for one. American Demographics, 25, 25–29.
Packard, V. (1957). The hidden persuaders. New York: D. McKay.
Popkin, J., et al. (2001). Small business share of economic growth. In Small Business Research Summary, no. 211 (January), U.S. Small Business Administration, Office of Advocacy. Washington: US Small Business Administration. http://www.SBA.gov/advo/research/rs211.pdf.
Robertson, T. S. (1976). Low-commitment consumer behavior. Journal of Advertising Research, 16, 19–27.
Robertson, T. S., Zielinski, J., & Ward, S. (1984). Consumer behavior. Glenview, IL: Scott, Foresman.
Rothschild, M. L. (1987). Advertising. Lexington, MA: D. C. Heath.
Spencer, J., & Albergotti, R. (2004). Customer-satisfaction index climbs. The Wall Street Journal, D2, (Feb. 18).
The Gallup Report (1986). Public mood is brightest on record. (No. 246, March), 2–4.
Todd, S., & Lawson, R. (2001). Consumer sentiment: New Zealanders’ views on marketing. In S. Chetty & B. Collins (Eds.), Bridging marketing theory and practice. Proceedings, ANZMAC Conference. New Zealand: Massey University ISBN 0-473-08206-3 (CD-ROM).
Wee, C. H., & Chan, M. (1989). Consumer sentiment toward marketing in Hong Kong. European Journal of Marketing, 23 (4), 25–39.
Wilkie, W. L. (1986). Consumer behavior. New York: Wiley.
Wilkie, W. L., & Moore, E. S. (1999). Marketing’s contributions to society. Journal of Marketing, 63 (Special Issue), 198–218.
Wilkie, W. L., & Pessemier, E. A. (1973). Issues in marketing’s use of multi-attribute attitude models. Journal of Marketing Research, 10, 428–441.
Acknowledgements
The author thanks colleagues at the University of Notre Dame, especially Mike Etzel and Hang Li for their long-term contributions to this project. Current and former Synovate (previously Market Facts) executives Verne Churchill, Robert Toll, Norman Kane, John Hageman, Mike Nugent, Kate Permut, and John Vidmar are acknowledged as well.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A
Questionnaire Items Footnote 14
ICSM
Product Quality
The quality of most products I buy today is as good as can be expected.
*I am satisfied with most of the products I buy.
*Most products I buy wear out too quickly. (R)
Products are not made as well as they used to be. (R)
*Too many of the products I buy are defective in some way. (R)
*The companies that make products I buy do not care enough about how well they perform. (R)
*The quality of products I buy has consistently improved over the years.
Price
*Most products I buy are overpriced. (R)
*Businesses could charge lower prices and still be profitable. (R)
*Most prices are reasonable considering the high cost of doing business.
Competistion between companies keeps prices reasonable.
Companies are unjustified in charging the prices they charge. (R)
*Most prices are fair.
*In general, I am satisfied with the prices I pay.
Advertising
Most advertising provides consumers with essential information.
*Most advertising is very annoying. (R)
*Most advertising makes false claims. (R)
*If most advertising were eliminated, consumers would be better off. (R)
*I enjoy most ads.
Advertising should be more closely regulated. (R)
*Most advertising is intended to deceive rather than to inform consumers. (R)
Retailing or Selling
*Most retail stores serve their customers well.
*Because of the way retailers treat me, most of my shopping is unpleasant. (R)
*I find most retail salespeople to be very helpful.
Most retail stores provide an adequate selection of merchandise.
In general, most middlemen make excessive profits. (R)
*When I need assistance in a store, I am usually not able to get it. (R)
*Most retailers provide adequate service.
Appendix B
On Measurement
The Reflective–Formative Issue
Readers may initially detect a trace of formative, versus reflective, character in some of the advertising and distribution/retail items in Appendix A. For instance, statements incorporating “essential information,” “annoying,” and “false claims” would seem to capture different dimensions of the applicable construct, would they not? Despite the appearance, however, recognize that the inherent construct can be and is here defined constitutively as (sentiment toward) total functional benefit relative to cost or downside, or satisfaction with service output relative to input – in this case helpful (information) versus unhelpful (deceptive, annoying) stimuli. Therefore, the item “... advertising provides... essential information,” embodies the benefit side or numerator of the relation with the cost or input implicit. “Most advertising is very annoying” and “... makes false claims” would then represent the downside or denominator, leaving the output or benefit implicit (use of items representing different parts of a construct enabled variety of stimuli presentation, essential in reflective multi-item scales). All items, thus understood, proxy the total construct.
Equivalent rationale justifies some of the distribution/retail statements. Given an effective conceptual definition of (sentiment toward) service output/input or benefit/cost (analogous to the above), “the way retailers treat me” and “selection of merchandise” are partial surrogates for the service output side of the relation, and “middlemen make excessive profits” proxies the input.
Readers who still regard some scale items as problematic in this way should note that most of the potential offenders were removed in purification. Furthermore, on the empirical side, validation evidence including factor analysis and coefficient α, first appearing in Gaski and Etzel (1985, 1986), supports unidimensionality of the four individual scales.
The Affective–Cognitive Issue
The Index of Consumer Sentiment toward Marketing is a measure of an affective construct, sentiment – explicitly favorable or unfavorable sentiment toward an object of attitude, in this case marketing. Inspection of the scale items of Appendix A reveals many to be literally statements of cognitions, however. Reconciliation occurs by noting that cognitions may serve as indirect measures of affect. This practice is common and defensible when the target characteristic is a motherhood attribute, i.e., one of universal approbation, for which greater perceived presence is tantamount to more favorable affect. Examples would be personal freedom, absence of pain, economic gain, adequate oxygen to breathe, and, in this context, traits such as good product quality, fair prices, truthful advertising, and helpful retail service. Questionnaire items measuring perceptions of these qualities are clearly representing affective reactions as well (technically because of the valence imparted by the “good,” “fair,” “truthful,” and “helpful” anchors). Hence, a measure of affective sentiment emerges legitimately from the items shown in the preceding appendix.
Rights and permissions
About this article
Cite this article
Gaski, J.F. The Index of Consumer Sentiment toward Marketing: Validation, Updated Results, and Demographic Analysis. J Consum Policy 31, 195–216 (2008). https://doi.org/10.1007/s10603-008-9066-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10603-008-9066-5