Abstract
A key extra-perceptual deficit in developmental prosopagnosia (DP) is face recollection memory, the all-or-none retrieval of face information, including semantic and contextual details. However, it remains unclear whether cognitive training can improve face recollection in DPs and what predicts training success. In the current study, we used a novel 8-week perceptual (holistic processing) + repetition-lag training program targeting face recollection in 23 DPs (Age: M=42.13, SD=14.88). Before, immediately after, and 6 weeks post-training, DPs completed assessments of face perception (e.g., computerized Benton, CFPT), face memory (e.g., CFMT), and face recollection/familiarity. DPs improved in overall face perception (Pre: M z=-1.44, SD=1.06, Post: M z=-1.00, SD=1.06; p<.001) and overall face memory (Pre: M z=-1.56, SD=.64, Post: M z=-1.21, SD=.78; p=.003). Notably, we observed differential improvement on an independent measure of recollection (Pre: M z=-0.75, SD=0.60, Post: M z=-0.53, SD=.62; p=.008) compared to familiarity (Pre: M z=-0.39, SD=0.53, Post:M z=-0.27, SD=.39; p=.12), with significant improvement only observed in recollection. DPs with better baseline face perception and eye discrimination ability were more accurate on the repetition-lag training, reaching higher levels of difficulty (Perceptual: r(21) = .54, p=.007; Eyes: r(21) = .46, p=.03). When dividing trainees into “better trainee” and “worse trainee” subgroups based on level-of-training reached, better trainees had significantly greater improvements across perception (Better: M z=.67, SD=.52, Worse: M z=.22, SD=.60; p=.04) and recollection (Better: M z=.50, SD=.36, Worse: M z=.04, SD=.47; p=.008) tasks. Together, this suggests that repetition-lag training can successfully improve face recollection and face perception/memory, particularly for DPs with less perceptual impairment who can reach the more challenging training stages.