The comparison of one-item-at-a time to multiple-items-at-a-time measurement of a brand equity framework for predicting brand choices in real markets

Publication Type:
Thesis
Issue Date:
2010
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail01Front.pdf10.61 MB
Adobe PDF
Thumbnail02Whole.pdf143.01 MB
Adobe PDF
NO FULL TEXT AVAILABLE. Access is restricted indefinitely. ----- This thesis compares two methods of preference elicitation, namely one-item-at-a-time measurement known as monadic rating scales and multiple-items-at-a-time measurement known as discrete choice experiments requiring respondents to trade off among items. Although the rating scale approach has been widely used, there are serious concerns with its use e.g., response style biases (Bachman and O'Malley 1984). Discrete choice measurement has well established behavioral and measurement theory, namely random utility theory (Thurstone 1927; Luce 1959; Churchill 1979; McFadden 1986; McFadden 2000) and has the advantage of ameliorating low discrimination across items. To our knowledge, no previous research has compared these two preference elicitation approaches regarding the prediction of real market choices. To enable us to compare the two methods, we use the Erdem and Swait (1998) Brand Equity Framework, because it is built upon information economics and signaling theory (Spence 1974) and is fully compatible with random utility theory (McFadden 1974). To further examine the validity of the framework, we test it using banking and mobile phones service categories given its previous testing in predominantly goods categories. Prior to conducting our empirical comparisons, we enhance the relevance of the Erdem and Swait (1998) Brand Equity Framework in services by refining its measurement items using banking and mobile phone brands. We begin our comparisons by examining the ability of measurement models to predict market choice. First, we compare models based on a relatively new multiple-items-at-a- time approach known as Best-Worst Scaling (Finn and Louviere 1992; Marley and Louviere 2005) with one-at-a-time ratings based models based on confirmatory factor analysis (CFA). We find the Best-Worst approach outperforms the CFA approach in predicting real market choices. Second, we base our comparisons on the structural model underlying the Erdem and Swait (1998) Brand Equity Framework using main effects only models. The multiple- items-at-a-time preference elicitation method is based on discrete choice experiments, whilst the traditional one-item-at-a-time approach is based on monadic rating scales. We find that discrete choice models outperform rating scale models in predicting market choices. Third, we extend our comparisons of the structural model by adding two-way, threeway, and four-way interactions and again find that discrete choice models consistently outperform rating scale models in predicting market choices. Our empirical results show that multiple-items-at-a-time measurement approaches consistently outperform the one-item-at-a-time measurement approach in predicting real market choices. We believe our study lays a firm foundation for future research in this exciting area.
Please use this identifier to cite or link to this item: