Adult Social Implicit Learning

Predictive coding, especially in a social context, appears to be challenging for autistic individuals (Keysers et al., 2024; Van de Cruys et al., 2014; von der Lühe et al., 2016). Specifically, autistic individuals may overlearn about the volatility in their environment and make overly precise predictions that are unable to account for the range, nuance, and variation that exists in social interactions (Keysers et al., 2024; Van de Cruys et al., 2014). Therefore we expect social implicit learning and prediction to be more challenging for adults with a higher rate of autistic traits. The Adult Social Implicit Learning research project has two purposes. The first objective is to examine whether performance on the tasks decreases as autistic traits increase. The second purpose is to test a novel social and nonsocial implicit learning paradigm adapted from the Weather Prediction Task (WPT). We want to see whether participants are able to implicitly learn the underlying rules and patterns of the task and improve their accuracy. Before beginning the experimental task, participants will complete the shortened Social Responsiveness Scale (shortened SRS) to probe autistic traits. Next, they will complete adaptations of the WPT. The WPT is an implicit learning paradigm in which participants are presented with one to three cards out of four possible cues cards with geometric patterns and are asked to say whether the weather will be rain or shine based on those cards (Gluck et al., 2002). Over the course of the experiment, participants improve their abilities to make these predictions (Gluck et al., 2002; Rustemeier et al., 2013). Non-declarative strategies have been shown to be the most common and most effective in this task (Rustemeier et al., 2013). The adapted social task has four cue cards with social actions and 12 outcome cards with socioemotional responses. The adapted nonsocial task has four cue cards with geometric patterns and 12 outcome cards with different types of weather. In an online asynchronous study, participants first see two cue cards that are associated with outcome cards with different probabilities. Then they will see two outcome cards and will be asked to determine what the outcome will be based on the two cue cards presented in each trial. There will be three acquisition blocks with feedback and one test block without feedback. At the end of the task, a questionnaire probing declarative knowledge of the relationships between the cue cards and the outcome cards will be administered. If participants are able to answer the questions correctly, that indicates that they used explicit knowledge to improve their task performance, but if they are unable to, that indicates using implicit knowledge. The social stimuli will include social patterns that are atypical to real life to disallow participants from using prior social knowledge to complete the task. Changes in accuracy over time and accuracy in the test phase will be measured. Additionally, we will examine whether accuracy changes as one’s level of autistic traits changes. We expect to find that participants will improve their accuracy over time, and will perform above chance on the test block. We also expect participants with higher autistic traits to perform worse on the social task, but similarly on the nonsocial task.