Subscale distance and item clustering effects in self-administered surveys: A new metric

The authors explore the effect of a form of question context on responses to a computer-mediated marketing research survey. As an increasing proportion of marketing research is conducted through computer interfaces, the pool of potential context effects is rapidly expanding. The authors conduct an experiment using a multi-item scale that consists of five dimensions and manipulate three such context effects: explicit item labeling, item presentation (alone/grouped), and subscale items presented contiguously or not. In a refined analysis of variance, the authors use a special one-dimensional case of the spatial and attribute-based distance metric proposed by Hoch, Bradlow, and Wansink (1999) to explain subscale variance, replacing the indicator variable for clustering used in a standard analysis of variance. This metric provides a scalar measure of how much variation exists in the order of presentation of items within a subscale (the subscale distance). This analysis indicates a significant decrease in subscale variance (increased reliability) with decreasing subscale distance but no-longer-significant effects due to labeling and grouping. The authors discuss implications of their findings for researchers conducting surveys in computer-mediated environments.

Duke Scholars

Author Gavan J. Fitzsimons Fuqua School of Business