The development of holistic face processing in children: a variation of the composite face effect

Thorniley, Sarah and Davis, Josh P. (2012) The development of holistic face processing in children: a variation of the composite face effect. In: British Psychological Society Annual Conference 2012, 18-20 Apr 2012, London, UK.

PDF - Submitted Version Restricted to Repository staff only Download (592kB)

Abstract

Objectives: Holistic face processing is considered a developmental face recognition marker, and is often assessed using the Composite Face Effect. This research, comparing adults and young children, employed chimeric (vertically-split) hemi-faces as a paradigm variation. Horizontally separated (gap) and vertically separated (misaligned) conditions measured release from holistic binding. A simultaneous matching condition, examined whole face processing. Images were constructed from photographs of a single target (Target present - TP), or of two different targets (Target absent – TA). Overall, children’s performance was expected to be inferior to adults. However, with less developed holistic binding processes, their performance at TA hemi-face discrimination might be comparatively superior.

Design: A 2 (age: adult, children) x 4 (image condition: matching, chimeric, gap, misaligned) x 5 (image identity: TP-Same, TP-Mirrored, TP-Different, TA-Similar (physical appearance), TA-Dissimilar) mixed design was employed. Accuracy and response time data were collected.

Methods: Participants (21 adults; 27 children) provided speeded ‘same’ or ‘different’ identity responses to randomly presented images across conditions.

Results: No response time effects across age groups were significant. Adults were more accurate in the gap, matching and the chimeric conditions, but not misaligned trials. TA-Similar accuracy, particularly in children was worse than TA-Dissimilar.

Conclusions: In matching and chimeric trials children displayed equivalent but inferior patterns of performance to adults, suggesting that expertise develops with age. There were no age group differences in the misaligned trials. Children’s gap condition accuracy was considerably worse suggesting that they release less readily from holistic binding with horizontally, but not vertically separated hemi-face images.

Actions (login required)