CAN CARICATURES REALLY PRODUCE DISTINCTIVENESS EFFECTS

The effects of distinctiveness have previously been assessed using one set of distinctive faces and a separate set of typical faces. However, a much more controlled experimental design is to vary distinctiveness within a single set of faces by using caricatures and veridical line drawings. Four experiments which manipulated distinctiveness in this way are described. The first experiment examined the effect of distinctiveness enhancement on the performance on a face recognition task while the second experiment examined the same effect on a face classification task. The results mirrored those found when comparing faces that vary naturally in distinctiveness. Experiment 3 used a learned identification task to show that unfamiliar faces were also susceptible to distinctiveness-enhancement effects. Furthermore, Experiment 4 showed that when given a limited amount of training, recognition of the real (photographed) faces was significantly better following training with caricatures as opposed to veridical line-drawings. These results support the predictions arising from Valentine's multi-dimensional space model (1991). Furthermore, they suggest that caricatures may be ideal materials to examine distinctiveness effects.