The effect of retention interval and target-decoy similarity on facial recognition : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in Psychology at Massey University

Abstract

This research was an attempt to resolve the inconsistent results for the effect of delay interval on facial recognition. The theory tested was that the degree of target/decoy similarity may act either to enhance or to diminish the effect of delay primarily by influencing false alarm rates. The first experiment used a novel method to scale the 80 faces along the dimension of similarity. The results showed that the method used was reasonably successful in ordering the faces along the similarity dimension. It enabled the use of four sets of 20 faces as either low or high similarity decoy and target sets in a second experiment aimed at testing the proposed theory. It was predicted that high target/ decoy similarity would result in a greater effect of delay than low target/ decoy similarity. Six groups of 15 subjects completed a standard face recognition experiment which crossed 0, 1 and 21 days delay with high and low similarity target/ decoy sets. The results showed a main effect for similarity, but, surprisingly, no main effect for delay. Nor was there the predicted interaction between similarity and delay for false alarms. The failure of the second experiment to test adequately the theory, and reasons for failure are discussed, along with the importance of the link between similarity and delay.