Eye Placement Bias Is Remarkably Robust

Abstract

When drawing faces, people show a systematic bias of placing the eyes higher up the head than they are placed in reality. This study investigated the development of this phenomenon while removing the potential confound of drawing ability. Participants ( N = 124) in three age groups (3–5 yo, 10–11 yo, and adults) reconstructed two foam faces: one from observation and one from memory. The high eye placement bias was remarkably robust with mean eye placement in every condition significantly higher than the original faces. The same bias was not shown for mouth placement. Eye placement was highest for the youngest participants and for the memory conditions. The results suggest that an eye placement bias is not caused by the motor skill demands required for drawing and lend evidence to the suggestion that an eye placement bias is caused by perceptual and decision-making processes.

It’s the Great Eigen-pumpkin, Charlie Brown!

“…Simpler versions of such a production task have also been used with developmental populations to estimate children's changing configural template for face detection and recognition. Both Balas et al (2023) and Smith et al (2021) found evidence of this bias in childhood, with younger children making especially large errors in eye height when producing images, but Balas et al additionally report that children in the target age range reliably recognized that faces with more canonical feature placement were better examples of typical face images. With regard to face recognition (rather than the overall face-ness of face portraits), production tasks also provide unique insights.…”