Abstract

Arising from the so-called 'replication crisis' in the experimental psychology literature, there has been a growing call to reassess whether specific analytic practices might lead to the improved accuracy and precision of given findings. This issue is explored here, through a case-study examination of two previously collected datasets (Dual Mechanisms of Cognitive Control [DMCC] task battery), used to illustrate some of the unique advantages afforded by Hierarchical Bayesian Regression (HBR) models as a potentially more rigorous analytic approach to statistical inference. The goal is not to provide a definitive argument for the HBR approach relative to other available computational methods, but rather to highlight the flexibility and power of HBR statistical models versus more conventional tests. In particular, these models provide the means to: 1) generate cumulative posterior distributions that represent the current degree of certainty on the most probable values of relevant parameter estimates, 2) make more direct statements regarding replication as well as evidence in favor of specific alternative and null hypotheses, and 3) more precisely model complex features that underlie the data structure. In the DMCC datasets, two sets of HBR models are presented, with the estimates of the former used as priors for the latter. The results not only demonstrate quantitatively precise estimates for each effect of theoretical interest, but also enable more refined qualitative conclusions to be drawn regarding evidence in favor of specific null and alternative hypotheses.

Committee Chair

Todd Braver

Committee Members

Joshua Jackson, Wouter Kool

Degree

Master of Arts (AM/MA)

Author's Department

Psychology

Author's School

Graduate School of Arts and Sciences

Document Type

Thesis

Date of Award

Winter 12-2024

Language

English (en)

Download

Included in

Psychology Commons

Share

COinS