Understanding face matching

There is tentative evidence that the match-mismatch dissociation in face matching might be driven by similar biases to those that attenuate the mirror effect.

“There is tentative evidence that the match-mismatch dissociation in face matching might be driven by similar biases to those that attenuate the mirror effect.”

Biases at the stimulus level of matching tasks might merely reflect that some face pairs are easier than others, reflecting similarity of faces depicted.

“Thus, biases at the stimulus level of matching tasks, might merely reflect that some face pairs are just easier than others, reflecting the similarity of the faces depicted (e.g., Fysh & Bindemann, 2023; for similar effects shown by somewhat increasing the distinctiveness of faces see McIntyre et al., 2013).”

Performance for match and mismatch trials is typically unrelated or correlated negatively.

“Contrary to this reasoning, performance for match and mismatch trials is typically unrelated or correlated negatively, whereby observers who are good at making one type of identification (e.g., that a face pair is an identity match) are poor at another (e.g., that a face pair is a mismatch; Fysh & Bindemann, 2023;Megreya & Burton, 2007;Sauerland et al., 2016).”

Variability in criterion can explain why match and mismatch performance and old/new performance are often uncorrelated in behavioural experiments.

“Collectively, this can explain why match and mismatch performance and old/new performance is often uncorrelated in behavioural experiments (e.g., Fysh & Bindemann, 2023;Megreya & Burton, 2007;Megreya et al., 2011;Sauerland et al., 2016).”