In pre-test 1, all stimuli were rated in an online questionnaire (N = 22, mean age =31, SD = 10, 13 women) on their level of typicality in a procedure similar to Veryzerand Hutchinson (1998) (1: not typical – 5: typical). In addition, participants rated theproducts’ functionality (1: not functional – 5: functional) and ease of use (1: not easy touse – 5: easy to use) on 5-point scales1. Functionality and ease of use were combinedinto an overall functionality variable by averaging them (rteapots = .665, rtoasters = .790,rhand-juicers = .448, rwashingmachines = .778, all ps < .01). The lowest correlation betweentypicality and overall functionality was found for toasters (rteapots = .424, rtoasters = .308,rhand-juicers = .415, rwashingmachines = .341, all ps < .01). In addition, a repeated measuresAnalysis of Variance (ANOVA) revealed a significant product category × prototype deviation interaction for overall functionality, F(12, 252) = 4.0, p < .001. Teapots, handjuicers,and washing machines differed significantly in overall functionality, F(4, 84) =11.0, F(4, 84) = 6.066, F(4, 84) = 18.3, respectively; all ps < .01. Toasters did notdiffer in overall functionality, F(4, 84) = 2.4, p > .10. Furthermore, the toasters differedsignificantly in perceived typicality, F(4, 84) = 10.5, p < .01 (mean ranged from 4.14to 2.64), with the most deviant shape being rated the least typical, suggesting that ourmanipulation of typicality was successful for the toaster stimuli.