Poster #7 - Associations between environmental unpredictability and cognitive learning processes in early adolescence

• In Event: Poster Session 11

Abstract

Introduction: Early experiences are posited to shape youth’s cognition in contextually-adaptive ways to prepare them to meet the demands of their specific environments (Frankenhuis et al., 2020). This often involves trade-offs–experiences that promote the development of abilities that are advantageous in one context (e.g., unpredictable conditions), but disadvantageous in another (e.g., stable conditions). Recent work suggests that environmental unpredictability may affect youths’ associative learning in this contextually-contingent manner. For instance, some studies have shown associations between unpredictability and broad impairments in learning and higher order cognition (e.g., Harms et al., 2018). However, consistent with the idea of developmental adaptation to context, unpredictable environments have also been linked to enhancements in keeping track of, and storing information about, rapidly changing conditions (e.g., Fields et al., 2021; Young et al., 2018)–learning that is advantageous in these contexts. Here we aim to tease apart these relations by considering the role of unpredictability in multiple interrelated cognitive processes. Specifically, we will examine associations between unpredictability in the home environment and individual differences in reinforcement learning and cognitive flexibility in adolescence.

Methods: Participants include 56 12- to 15-year-olds (51.8% female, 94.6% White, 19.6% Asian, 3.6% Hispanic, 1.8% Other) living in a metropolitan city in the Midwest with an average household income-to-needs ratio of 6.1 (sd = 3.5). Youth completed a computerized learning paradigm that begins with an acquisition phase where youth learn stimulus-outcome associations between a stimulus (i.e., emotional and non-emotional pictures) and a reinforcement via reward (where points are gained) or punishment (where points are lost). About halfway through the task, outcomes associated with half of the stimuli switch to the opposite reinforcement value, which forms the reversal learning phase. Each phase consists of 8 blocks (96 trials) and combined make up one run. The task contains four runs with a new set of 12 stimuli presented in random order. Cognitive flexibility (DV) will be indexed via accuracy on reversal learning trials. Reinforcement learning (DV) will be indexed via learning rate (i.e., valuation and sensitivity to rewards and losses) generated from Bayesian computational models. Environmental unpredictability (IV) will be indexed via a standardized cumulative score combining the parent-reported Chaos Hubbard and Order Scale (CHAOS; Matheny et al., 1995) and adolescent-reported Questionnaire of Unpredictability in Childhood (QUIC; Glynn et al., 2019).

Planned analyses and hypotheses: We will use linear regression to test the hypothesis that higher levels of unpredictability will be associated with greater cognitive flexibility required for adapting to changing environmental demands. Further, informed by recent work with rodents (e.g., Lin et al., 2022), we expect that environmental unpredictability will be associated with lower learning rates in the context of punishment/loss specifically, due to the relative importance of tracking potentially negative unpredictable outcomes. Sociodemographic covariates including family socioeconomic status, race/ethnicity, sex, and age will be included in the models.

Conclusion: Anticipated findings have the potential to provide insight into the adaptive cognitive skills and abilities that adolescents can develop in response to adverse, unpredictable conditions.

Authors

• Amelia Mary Rose Spaeth, University of Minnesota - Twin Cities
  Presenting Author

• Cielo DeCastro, University of Minnesota - Twin Cities

• Meriah L. DeJoseph, University of Minnesota - Twin Cities

• Willem E Frankenhuis

• Daniel Berry, University of Minnesota - Twin Cities