Naresh K. Malhotra

Alternative Techniques for Assessing Common Method Variance An Analysis of the Theory of Planned Behavior Research

Each research domain carries the burden of examining the effects of common method variance (CMV) on published research within the domain. To focus on this concern in the context of the theory of planned behavior (TPB), this research empirically compares several methods of detecting the presence of and estimating the level of CMV in the TPB domain. These methods include various implementations of the marker variable technique and versions of the multitrait-multimethod (MTMM) technique. The results show that the marker variable technique provides estimates of CMV and CMV-corrected correlations and paths that are consistent with those produced using the other methods. Next, one implementation of the marker variable technique method is implemented post hoc on a large data set of published TPB studies. This analysis provides strong confirmatory evidence that the effects of CMV do not alter the substantive inferences of study results in prior research. Overall, these findings support putting to rest concerns about the adverse influence of CMV in the TPB domain.

The Influence of Common Method Variance in Marketing Research: Reanalysis of Past Studies Using a Marker-Variable Technique

Despite criticisms regarding the effects of common method variance (CMV) in marketing research, there have been only a limited number of studies that directly assess the pervasiveness of CMV biases. This paper is an attempt to quantify the ubiquity of such effects in the marketing literature by employing a relatively new method called the marker-variable technique (Lindell & Whitney 2001). A marker-variable is defined as a theoretically unrelated variable included in a study that can be used to estimate the effect of CMV. The technique is used as an analysis tool for assessing and controlling for CMV effects in published studies based on the reasoned-action framework (Fishbein & Ajzen 1975). The findings of this study show that after controlling for the effects of CMV, the majority of relationships between variables remain significant, even for relatively extreme values of CMV. Specifically, more than 80% of the 663 CMV-adjusted correlations that were computed remained significant. This leads to the conclusion that the influence of CMV may not be as problematic as once thought.

Conjoint Model with Artificial and Real Stimuli: A Comparative Assessment of Within and Cross-Domain Generalizability and Choice Prediction

In their review of the developments in conjoint methods, Green and Srinivasan (1990) report that a large number of studies have addressed validity issues. Most of these studies have involved tests of cross-validity using holdout set of profiles. These studies have shown very high internal validity for conjoint models when holdout samples are used (Green and Srinivasan 1990; Akaah and Korgaonkar 1983). However, mixed results have been reported for external validity thereby indicating, in some cases, a lack of convergence between models estimated from artificial and real stimuli (Holbrook and Havlena 1988; Holbrook et al 1985).

An Integrated Model Of Attitude And Choice: An Interaction Approach

Attitude, preference, and choice models in marketing still continue to be based on the multi-attribute paradigm despite the acknowledgement of affect (feelings and emotions) in brand attitude. Although voluminous work has been done in the area of affect, there has been a lack of research effort to unify the two streams within the attitudinal and choice framework. The objective of this paper is to unify the two streams of research and develop an integrated model of attitude and choice. We define “attitude” as a summarized evaluative judgment based on cognitive beliefs and its evaluative aspect (traditional multi-attribute model), and “affect” is reserved for valenced feeling states and emotions (Cohen and Areni 1991; Erevelles 1998). Our conceptualization of affect is drawn from category-based affective processing in the categorization literature. Fiske and Pavelchak (1986) distinguish between piecemeal versus category-based affective processing. The top-level affective tag may come from a conditioned response to the category label or may be the summation of lower level attribute-based affective tags (Cohen and Areni 1991). Drawing also from categorization theory, the essence of the proposed model is the parallel processing of brand-based attributes and holistic affect and their joint interaction (Fiske and Pavelchak 1986; Dabholkar 1994). For both overall attitude and choice, interaction effect is modeled at the multidimensional expectancy value component (EVC) level. Incorporating interaction effect in the proposed integrated model is a central theme of this study.