The statistical tests used in the analysis of structural equation models with unobservable variables and measurement error are examined. A drawback of the commonly applied chi square test, in addition to the known problems related to sample size and power, is that it may indicate an increasing correspondence between the hypothesized model and the observed data as both the measurement properties and the relationship between constructs decline. Further, and contrary to common assertion, the risk of making a Type II error can be substantial even when the sample size is large. Moreover, the present testing methods are unable to assess a model's explanatory power. To overcome these problems, the authors develop and apply a testing system based on measures of shared variance within the structural model, measurement model, and overall model.
Several issues relating to goodness of fit in structural equations are examined. The convergence and differentiation criteria, as applied by Bagozzi, are shown not to stand up under mathematical or statistical analysis. The authors argue that the choice of interpretative statistic must be based on the research objective. They demonstrate that when this is done the Fornell-Larcker testing system is internally consistent and that it conforms to the rules of correspondence for relating data to abstract variables.
The American Customer Satisfaction Index (ACSI) is a new type of market-based performance measure for firms, industries, economic sectors, and national economies. The authors discuss the nature and purpose of ACSI and explain the theory underlying the ACSI model, the nation-wide survey methodology used to collect the data, and the econometric approach employed to estimate the indices. They also illustrate the use of ACSI in conducting benchmarking studies, both cross-sectionally and over time. The authors find customer satisfaction to be greater for goods than for services and, in turn, greater for services than for government agencies, as well as find cause for concern in the observation that customer satisfaction in the United States is declining, primarily because of decreasing satisfaction with services. The authors estimate the model for the seven major economic sectors for which data are collected. Highlights of the findings include that (1) customization is more important than reliability in determining customer satisfaction, (2) customer expectations play a greater role in sectors in which variance in production and consumption is relatively low, and (3) customer satisfaction is more quality-driven than value- or price-driven. The authors conclude with a discussion of the implications of ACSI for public policymakers, managers, consumers, and marketing in general.
Are there economic benefits to improving customer satisfaction? Many firms that are frustrated in their efforts to improve quality and customer satisfaction are beginning to question the link between customer satisfaction and economic returns. The authors investigate the nature and strength of this link. They discuss how expectations, quality, and price should affect customer satisfaction and why customer satisfaction, in turn, should affect profitability; this results in a set of hypotheses that are tested using a national customer satisfaction index and traditional accounting measures of economic returns, such as return on investment. The findings support a positive impact of quality on customer satisfaction, and, in turn, profitability. The authors demonstrate the economic benefits of increasing customer satisfaction using both an empirical forecast and a new analytical model. In addition, they discuss why increasing market share actually might lead to lower customer satisfaction and provide preliminary empirical support for this hypothesis. Finally, two new findings emerge: First, the market's expectations of the quality of a firm's output positively affects customers’ overall satisfaction with the firm; and second, these expectations are largely rational, albeit with a small adaptive component.
In marketing applications of structural equation models with unobservable variables, researchers have relied almost exclusively on LISREL for parameter estimation. Apparently they have been little concerned about the frequent inability of marketing data to meet the requirements for maximum likelihood estimation or the common occurrence of improper solutions in LISREL modeling. The authors demonstrate that partial least squares (PLS) can be used to overcome these two problems. PLS is somewhat less well-grounded than LISREL in traditional statistical and psychometric theory. The authors show, however, that under certain model specifications the two methods produce the same results. In more general cases, the methods provide results which diverge in certain systematic ways. These differences are analyzed and explained in terms of the underlying objectives of each method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.