Choice Task Complexity and Decision Strategy Selection

Abstract

The psychology, the marketing consumer behavior and, to a much smaller extent, the economics literature have long reported evidence that decision makers utilize different decision strategies depending upon many factors (person-specific, task-specific, etc.). Such observations have generally failed to affect the specification of choice models in commercial practice and academic research, both of which still tend to assume an utility maximizing, full information, indefatigable decision maker. This is true whether the models deal with Stated Preference (SP - from hypothetical elicitations) or Revealed Preference (RP - from actual market decisions) choice data. This paper, which deals only with SP data, addresses the following issues: (1) does task complexity affect decision strategy selection in experimental choice tasks? (2) does the cumulative cognitive burden created by multiple choice scenarios done in sequence affect the selection of decision strategy by respondents? Our contribution is two-fold: (1) we introduce decision strategy selection as an explicit factor in aggregate choice models via the mechanism of latent classes, which are assumed to be a function of task complexity; (2) we demonstrate, for a particular set of data, that within the scope of an SP choice task, respondents did indeed make use of multiple decision strategies as choice set complexity changed and as the SP task progressed. We examine the import of our findings to current practice, model interpretation and future research needs.