Making faces: Developmental trends in the construction and recognition of photofit face composites

Abstract

Two experiments explored children's ability to create facial composites with the Photofit system. In Experiment 1, children ( n= 48) aged 8–9 and 11–12 years made Photofit pictures of two male faces from photographs, one from memory and a second with the face present. Irrespective of condition, older children produced superior initial verbal descriptions and more identifiable composites than did the younger groups. Children's performance was inferior to that of a comparable adult sample. Experiment 2 examined the ability of 5-to 11-year-old children ( n = 80) to identify a target Photofit face from an array containing alternative faces differing in only a single feature. Younger children made more errors than older children and the pattern of errors was slightly different for the 5-year-olds. No age × condition (face present vs. absent) effect was found. These results implicate the initial encoding stage of face processing as a potential locus of the observed age differences. Children are technically capable of acting as Photofit witnesses, but it is suggested that the composites produced by young children should be treated with a degree of caution.