In mental imagery of faces, we thus expected to ALSO find that the "typical person" is visualized more as male To our surprise there was no evidence for this If anything, participants' mental image of the face of a "typical person" overlapped more with their image of a "woman"

What could explain our unexpected result? Maybe there's some methodological flaw we didn't catch? Maybe there's a meaningful diff in bias in representing human concepts and visualizing human faces? For some of my related thoughts see also www.cell.com/trends/cogni...