Robert A. Peterson

A Meta-analysis of Cronbach's Coefficient Alpha

Despite some limitations, Cronbachs coefficient alpha remains the most widely used measure of scale reliability. The purpose of this article was to empirically document the magnitudes of alpha coefficients obtained in behavioral research, compare these obtained values with guidelines and recommendations set forth by individuals such as Nunnally (1967, 1978), and provide insights into research design characteristics that may influence the size of coefficient alpha. Average reported alpha coefficients ranged from .70 for values and beliefs to .82 for job satisfaction. With few exceptions, there were no substantive relationships between the magnitude of coefficient alpha and the research design characteristics investigated. Copyright 1994 by the University of Chicago.

Exploring the Implications of the Internet for Consumer Marketing

Past commentaries on the potential impact of the Internet on consumer marketing have typically failed to acknowledge that consumer markets are heterogeneous and complex and that the Internet is but one possible distribution, transaction, and communication channel in a world dominated by conventional retailing channels. This failure has led to excessively broad predictions regarding the effect of the Internet on the structure and performance of product and service markets. The objective of this article is to provide a framework for understanding possible impacts of the Internet on marketing to consumers. This is done by analyzing channel intermediary functions that can be performed on the Internet, suggesting classification schemes that clarify the potential impact of the Internet across different products and services, positioning the Internet against conventional retailing channels, and identifying similarities and differences that exist between them. The article concludes with a series of questions designed to stimulate the development of theory and strategy in the context of Internet-based marketing.

ANTECEDENTS AND CONSEQUENCES OF SALESPERSON JOB SATISFACTION: META ANALYSIS AND ASSESSMENT OF CAUSAL EFFECTS

A three-phase quantitative investigation of relationships involving salesperson job satisfaction was undertaken. First, the strength, valence, and consistency of pairwise relationships were assesse...

On the Use of College Students in Social Science Research: Insights from a Second-Order Meta-analysis

A second-order meta-analysis was conducted to assess the implications of using college student subjects in social science research. Four meta-analyses investigating response homogeneity (cumulative N > 650,000) and 30 meta-analyses reporting effect sizes for 65 behavioral or psychological relationships (cumulative N > 350,000) provided comparative data for college student subjects and nonstudent (adult) subjects for the present research. In general, responses of college student subjects were found to be slightly more homogeneous than those of nonstudent subjects. Moreover, effect sizes derived from college student subjects frequently differed from those derived from nonstudent subjects both directionally and in magnitude. Because there was no systematic pattern to the differences observed, caution must be exercised when attempting to extend any relationship found using college student subjects to a nonstudent (adult) population. The results augur in favor of, and emphasize the importance of, replicating research based on college student subjects with nonstudent subjects before attempting any generalizations. Copyright 2001 by the University of Chicago.

Measuring Customer Satisfaction: Fact and Artifact

Self-reports of customer satisfaction invariably possess distributions that are negatively skewed and exhibit a positivity bias. Examination of the customer satisfaction literature and empirical investigations reveal that measurements of customer satisfaction exhibit tendencies of confounding and methodological contamination and appear to reflect numerous artifacts. Implications and suggestions for research and practice are discussed.

Exploring the Implications of M-Commerce for Markets and Marketing

Business pundits have enthusiastically prognosticated about a seamless, mobile world where commerce occurs on an anywhere, anytime basis. This type of commerce has been referred to as mobile commerce or, more simply, m-commerce. However, there have been relatively few attempts to systematically explore the opportunities and challenges posed by m-commerce. This article investigates the implications of m-commerce for markets and marketing by means of a formal conceptualization of m-commerce, a space-time matrix that delineates the impact of mobile technologies, and a taxonomy of m-commerce applications.

The effect of effort on sales performance and job satisfaction

The authors address a fundamental gap in understanding how sales performance and job satisfaction are determined in an investigation of the sales force of a direct-selling organization. Results ind...

On the Use of Beta Coefficients in Meta-Analysis.

This research reports an investigation of the use of standardized regression (beta) coefficients in meta-analyses that use correlation coefficients as the effect-size metric. The investigation consisted of analyzing more than 1,700 corresponding beta coefficients and correlation coefficients harvested from published studies. Results indicate that, under certain conditions, using knowledge of corresponding beta coefficients to input missing correlations (effect sizes) generally produces relatively accurate and precise population effect-size estimates. Potential benefits from applying this knowledge include smaller sampling errors because of increased numbers of effect sizes and smaller non-sampling errors because of the inclusion of a broader array of research designs.

Relationship marketing and the consumer

ConclusionCollectively, the responses of the individuals interviewed reveal that there is a multitude of reasons for participating in a marketing relationship. Although many of the reasons are cognitive in nature, many others are affective. This affective “dimension” of relationship marketing remains unexplored but would seem to offer great promise for explaining the long-term nature of a firm-consumer relationship. Moreover, a liberal interpretation of the collective responses of the interviewed individuals suggests an alternative fundamental axiom of relationship marketing consistent with the definition of Angeles (1992):Consumers enter into a marketing relationship because they expect to receive positive value from their participation.Acceptance of this axiom would seem to suggest a conceptualization and research agenda much broader and potentially more productive than that associated with the axiom currently being offered.Sheth and Parvatiyar have raised a question that, despite its obviousness in retrospect, has yet to be addressed satisfactorily. Hopefully their article will prove to be the genesis of an answer. Although no one is likely to agree with all of their propositions, or even their speculation that relationship marketing will ultimately improve marketing productivity, Sheth and Parvatiyar have done the discipline a service by focusing attention on a neglected topic. Indeed, what they have effectively done is to challenge the discipline to debate what they believe to be the fundamental axiom of relationship marketing in consumer markets. Regardless of the outcome of this debate, marketing knowledge will have advanced.

A Meta-Analysis of Variance Accounted for and Factor Loadings in Exploratory Factor Analysis

A meta-analysis of two factor analysis outcome measures, the percentage of variance accounted for and the average (absolute) factor loading, in 803 substantive factor analyses was undertaken. The average percentage of variance accounted for was 56.6%, and the average (absolute) factor loading was 0.32. Number of variables factor analyzed, nature of the sample from which data were collected, sample size, number of factors extracted, and (minimal) number of scale categories employed influenced the percentage of variance accounted for in a factor analysis. Number of factors extracted, analytical approach, and number of variables analyzed influenced the average factor loading obtained in a factor analysis. Factor analysis of synthetic (random) data possessing the general structure as the observed data in the meta-analysis accounted for 50.2% of the variance in the data and produced an average factor loading of 0.21. The latter figures imply that many factor analyses have produced outcome measures of questionable meaningfulness.