Boris Penaloza¹, Marcello Maniglia², Jaap Munneke¹, C. Shawn Green³, Aaron Seitz¹; ¹Northeastern University, ²Rochester Institute of Technology, ³University of Wisconsin-Madison

Perceptual learning (PL), defined as experience-based improvements on perceptual tasks, is an important field addressing both mechanisms of neural plasticity and methodologies to improve perceptual function. However, the field faces major limitations in terms of reproducibility and accessibility. Inconsistencies across studies may stem from methodological variations or differences between research participants. Moreover, despite its potential as a cost-effective intervention for visual disorders, PL research is often restricted to university laboratories or specialized clinics. To address these challenges, we developed PLFest, a Unity-powered application that enables data collection across different sites and platforms (e.g., computers, smartphones, and tablets). PLFest currently supports multiple vision training paradigms, as well as hearing, attentional control, and cognitive assessments. Here, we evaluate PLFest’s reliability using data from a large group of participants (N=241). We demonstrate that PLFest is a reliable platform for a variety of assessments including visual acuity, contrast sensitivity functions, visual search, and reading, among others, with data replicating well-established findings in the field. Further we show that the platform is appropriate to collect high-quality perceptual learning data using a variety of approaches to train contrast sensitivity (e.g., standard perceptual learning with adaptive staircases, training with external noise, training with collinear flankers, stimulus variety across multiple spatial frequencies and locations, and multisensory facilitation with auditory cues). These preliminary data position PLFest as a shareable and reproducible research platform capable of reliably measuring and training visual functions in large groups of participants using accessible tools such as computer tablets. This suggests that PLFest can support big-data approaches to vision science, as well as providing promising results for increased outreach and translational opportunities.