Leveraging (dynamically growing) Individual Participant Data Meta-Analyses to Investigate Individual Differences, Learning Dynamics, and Key Outcome Measures in conditioned Responding across a large number of individual studies

• Applicant: Professorin Dr. Tina Barbara Lonsdorf
• Subject Area: Biological Psychology and Cognitive Neuroscience ; General, Cognitive and Mathematical Psychology
• Term: since 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 451328127

This proposal aims to advance the field of fear conditioning research through an innovative and sustainable re-use of existing data - the FEAR-BASE corpus which is a large, harmonized database of individual participant-level data in the research field. Rather than relying on traditional aggregated-level meta-analyses (AL-MAs), this project implements individual participant data meta-analyses (IPD-MAs), which retain study-level detail and allow for the rigorous assessment of effect modifiers, between-study variability. This methodological approach substantially enhances the interpretability and generalizability of findings in a field marked by high heterogeneity in design, population, and outcome measures. Most importantly, the IPD framework allows to investigate research questions beyond those addressed in the original publications (from which the individual datasets are derived). For instance, we identified a substantial share of datasets that include measures of trait negative affect but have not reported on this association in the respective publications. Altogether, this proposal offers a scalable and impactful strategy for maximizing the value of existing research investments. Building on prior preparatory work, the project identifies key gaps in the literature that cannot be resolved using individual studies or conventional meta-analytic approaches. Using the IPD-MA framework, we will synthesize trial-level data from numerous studies to address three central aims: (1) assess convergence and divergence of typical outcome measures in fear conditioning paradigms; (2) examine the role of trait negative affect in fear learning and extinction; and (3) model the temporal dynamics of fear acquisition and extinction via computational reinforcement learning models. The IPD-MA framework will allow us to comprehensively and systematically evaluating the impact of moderators such as age, sex, and experimental variables at an unpreceded level of detail across a large number of individual datasets. Furthermore, a unique feature of this project is its focus on sustainable knowledge generation: This project uses IPD meta-analyses of existing datasets to promote ecological and economic sustainability and ensure long-term accessibility of publicly funded research.

• DFG Programme: Research Grants