The fusiform face area (FFA) responds to visual expertise; neuropsychological data says not only to process faces, it also contributes to non-face expertise recognition (cars, birds, radiographs)

P-curving the Fusiform Face Area: Meta-Analyses Support the Expertise Hypothesis. Edwin J. Burns, Taylor Arnold, Cindy M. Bukach. Neuroscience & Biobehavioral Reviews, July 2 2019. https://doi.org/10.1016/j.neubiorev.2019.07.003

Highlights
•    Meta-analyses show the fusiform face area (FFA) responds to visual expertise
•    These effects are not due to publication bias or p-hacking
•    The failed replications do not negate the evidence supporting the expertise account
•    Neuropsychological data confirms the FFA contributes to non-face expertise recognition

Abstract: Psychologists have debated whether the right fusiform face area’s (FFA) responses are domain specific to faces, or domain general for certain object categories that we have visual expertise with. This latter domain general expertise account has been criticised for basing its assumptions upon studies that suffer from small participant numbers, small effects, and statistically significant p-values that are close to .05. An additional criticism is that these findings are difficult to replicate. A modern reader familiar with the replication crisis may therefore question whether the FFA’s expertise effect is real. The p-curve is a relatively new form of meta-analysis that enables researchers to identify whether there is evidential value for any given effect in the literature. We put the literature to the test by running p-curve analyses on all published expertise studies. Contrary to aforementioned criticisms, our meta-analyses confirm the right FFA’s expertise effect is based upon evidential value. We therefore review the broader literature to address additional criticisms of the expertise account and propose ways to improve replicability.