Using policy capturing to examine tipping decisions

Participants (n = 15) made tipping decisions for 80 restaurant situations. A policy-capturing analysis was then conducted for each participant to quantitatively describe relations between his or her judgments and the information used to make those judgments. Participants possessed reliable, simple, and nonconfigural models. The majority of these individual models heavily weighted bill-size information. In addition, service-quality, server-friendliness, or food-quality information affected tipping decisions, to a lesser extent; for a number of individuals. Atmosphere, server gender, and restaurant cleanliness information were not considered in any tipping model. Unlike affect, social desirability, and gender, participants' dining-out frequency was related to the types of information used when tipping. Finally, cluster analysis of the models revealed 11 general approaches to tipping.