CHDH Seminar Series 2020: The Death of Accumulator Models? — Nathan Evans

Image for  CHDH Seminar Series 2020: The Death of Accumulator Models? — Nathan Evans

Presenter: Dr Nathan Evans

Date: Monday 21 September

Time: 12 - 1 PM

Location: Via Zoom. Please use this link at the time of the event to join: https://unimelb.zoom.us/j/93193522071?pwd=QkNIYk03SlpNU1pWem5iYmxYcmJwUT09

Seminar Description

Evidence accumulation models (EAMs) have become the dominant framework for understanding rapid decision-making, where noisy samples of stimulus information accumulate until a threshold is reached for one alternative, triggering a decision. The success of EAMs has resulted in the development of many models within the framework, ranging from the simple independent accumulator models designed to provide tools for measuring latent variables, to more complex models with time-varying processes designed to provide accurate representations of how the decision-making process operates. Here, Nathan will discuss some of the major issues with the simple independent accumulator models, and why they provide an unreliable measurement tool when further constraints are not imposed. Nathan will also discuss some of the major issues with more complex time-varying models, and how their lack of constraint leads to unreliable measurement properties. Most importantly, Nathan will stress the need for more robust assessments of the measurement properties of models when they are used in the context of being a measurement tool, and provide some solutions for researchers unsure of how to best perform these assessments.

Presenter Biography

Nathan Evans is an ARC DECRA Fellow at the University of Queensland’s School of Psychology, who completed his PhD at the University of Newcastle in 2017 and previously worked at Vanderbilt University and the University of Amsterdam. His research focuses on developing and testing models of cognitive processes, with a major focus on models of human response time, as well as the development of methods for model-based inference, such as Bayesian inference and pseudolikelihood approximations.

Nathan currently has a PhD position open. For information on the position: https://graduate-school.uq.edu.au/phd-scholarships-humanities