Regression versus fast and frugal models of decision-making: the case of prescribing for depression

The present study compared the relative fits of two different models to general practitioners' (GPs') decisions regarding possible prescriptions of medication in response to vignettes of patients with or without each of eight symptoms (cues) relating to depression. The two models were (1) a non-compensatory ‘fast and frugal’ (F&F) model called the Matching Heuristic (MH) model and the compensatory integration model of linear (logistic) regression. The MH model described the GPs' prescribing responses almost as well as the logistic regression model but with a more economical use of cues. Both models indicated that a relatively small number of cues were used in reaching decisions. However, the MH model was more psychologically plausible and proposed a more flexible and cognitively simple strategy. Overall, the results support the utility of F&F approaches to real life decision-making.