About AMA

Executive Summary
Practitioners favor single-item measures for practical reasons, but most academic marketing researchers believe that multiple-item measures are necessary to represent validly the major constructs (variables) studied in marketing. Indeed, academic journals in marketing will not accept articles with single-item measures of major constructs. Academic researchers Aimee Drolet and Donald Morrison, in the Journal of Service Research (2001) argue in favor of single-item measures on the practical ground that, in at least some cases, a single item provides most of the "information" in a multiple-item scale. The only theoretical argument for single-item measures has been made by John Rossiter in his article in International Journal of Research in Marketing (2002), which introduces his alternative procedure for scale development, titled C-OAR-SE.

According to C-OAR-SE, a single-item measure is appropriate if the object being rated is simple and unambiguous (e.g., an advertisement, a brand) and the attribute on which the object is being rated is also simple and unambiguous to the raters (e.g., liking of the advertisement, overall evaluation of the brand). Rossiter calls these "doubly concrete" constructs to distinguish them from more complex "abstract" constructs in which the object, the attribute, or both have multiple meanings to raters and thus require multiple items to represent them. However, this is a purely theoretical argument, and to date, there has been no test of whether a single-item measure of a doubly concrete construct is as valid as a multiple-item measure of the same construct. There has been no test of whether the researcher "loses anything" by using a single-item measure.

One such test—and the most practical one—is whether the single-item version can predict equally as well as the multiple-item version. The current study offers a test of the predictive validity of single-item versus multiple-item measures of two widely studied, doubly concrete constructs: attitude toward the ad (A_Ad) and brand attitude (A_Brand). Bergkvist and Rossiter devise single-item measures of A_Ad and A_Brand with the straightforward procedure of taking one item from a typical three-item measure of each of these two constructs. They then use the single-item and multiple-item versions of A_Ad to predict the single-item and multiple-item versions of A_Brand for four print advertisements for new brands, providing four replications of the test. They use two criteria to assess predictive validity: the coefficient of correlation, r, between the two variables and the multivariate squared correlation, R², which allows for the possible effect of a second predictor variable, Beliefs_Brand, added to A_Ad to predict A_Brand.

The findings clearly demonstrate by both criteria (r and R²) that the single-item measures predict equally as well as the multiple-item measures.

The conclusion from the study is that it is unnecessary and wasteful to use multiple-item scales to measure basic constructs in marketing (in which "doubly concrete" is the operational boundary definition of "basic"). The tests are based on A_Ad and A_Brand, but the single-item recommendation is also intended for other constructs in marketing that involve basic objects (e.g., companies, retailers, salespersons, sales promotions) and basic attributes (e.g., perceptions, intentions, satisfaction). Marketing research practitioners have always used single-item measures for constructs such as these, and academic researchers and academic journal editors can now follow without concern about losing validity (r) or explanatory power (R²).

Biography
Lars Bergkvist is Senior Lecturer in Marketing and a member of the Marketing Research Innovation Centre at the University of Wollongong in Australia. He received his doctoral degree from the Stockholm School of Economics and has several years of practitioner experience as a media analyst and market researcher. His research interests are in the areas of brand management, marketing communications, and marketing measurement.

John R. Rossiter is Research Professor of Marketing and a member of the Marketing Research Innovation Centre at the University of Wollongong in Australia. He is also Visiting Professor of Marketing in the Rotterdam School of Management at Erasmus University in the Netherlands. He received his doctoral degree from the University of Pennsylvania and has taught at the Wharton School, Columbia University, University of Technology Sydney, and the Australian Graduate School of Management. He is best known for his books with Larry Percy on advertising and promotion management (McGraw-Hill 1987, 1997), with Peter Danaher on media planning (Kluwer 1998), and with Steven Bellman on marketing communications (Pearson Prentice Hall 2005). He is on the editorial board of several leading journals in consumer behavior and advertising. In recent years, his theoretical and research interests have extended to the broad fundamental topics of marketing knowledge (authoring articles in Marketing Theory [2001, 2002]) and marketing measurement (authoring an article in International Journal of Research in Marketing [2002] that introduced a new procedure for scale development). He is a Fellow of the American Psychological Association, the Australian and New Zealand Marketing Academy, and the Market Research Society of Australia.

Journal of Marketing Research, Vol. XLIV, No. 2, May 2007
View Table of Contents.