The Multi-Attribute Attentional Drift Diffusion Model of Consumer Choice

Many choices consumers face are over outcomes that consist of multiple attributes. For instance, when deciding to purchase a particular product consumers must account for both its price and its brand. Although these choices are quite common, we know relatively little about how changes in consumer attention to specific attributes impacts decisions. We propose a new computational model, the multi-attribute attentional drift diffusion model (maDDM), that describes how consumers weight a product’s attributes when making a decision. The model makes predictions about how changes in the amount of attention deployed to different attributes affects choices. In a laboratory experiment that makes use of eye tracking, we test and find evidence for these predictions. The model also predicts the existence of an attentional bias in multi-attribute choice: consumers increase the weight of the currently attended attribute and decrease the weight of unattended attributes. This bias affects decisions and has important implications for consumer science.