Abstract
Accurate memory for faces is important for numerous social interactions, but is modulated by physiognomic characteristics. Recently, it was shown that attractive faces are remembered less accurately than unattractive faces when distinctiveness is controlled for (Wiese, Altmann, & Schweinberger, 2014). It is still unclear, however, if this effect is constituted by a disadvantage for attractive faces, a benefit for unattractive faces, or both. To clarify, we conducted a recognition memory experiment including attractive, mid-attractive and unattractive faces, matched for deviation-based distinctiveness (Wickham & Morris, 2003). In addition to behavioural measures, event-related potentials were analysed. Memory was most accurate for unattractive faces, followed by attractive faces, with poorest performance for mid-attractive faces. Item analyses confirmed that this effect remained significant when emotional valence was taken into account. Furthermore, participants responded less conservatively to attractive compared to both mid- and unattractive faces. Replicating earlier findings, an attractiveness effect in the early posterior negativity (EPN) during learning, with larger amplitudes for attractive than unattractive faces, correlated significantly with the memory advantage for unattractive faces. Higher amplitudes with increasing attractiveness were found both in the EPN at test and in the late positive component (LPC) during both experimental phases. Finally, an old/new-effect with larger amplitudes for hits than correct rejections (Rugg & Curran, 2007) was found between 500-700 ms at test, but did not interact with attractiveness. Our findings show an effect of attractiveness on face memory beyond distinctiveness and emotional valence. As the EPN is typically enhanced for affective stimuli, processing of emotionally relevant attractive faces during learning could hamper their encoding into memory. This, however, cannot explain poor performance for emotionally neutral mid-attractive faces. We suggest an explanation based on a modified face-space model, in which mid-attractive faces are more frequent, and hence less well distinguishable, than both attractive and unattractive faces.
Meeting abstract presented at VSS 2015