Research Projects

Photo by Danièle Martinie

Feature Processing in Perceptual Categorization

A hallmark of decision making is that information from multiple sources must be combined to achieve specific goals.Simple decisions are best thought of as the accumulation of information over time from a single source. For complex decisions, accumulation from multiple sources might occur simultaneously, making the decision process surprisingly simple, or sequentially, making the decision process complicated.

This project focuses on revealing the processes and representations that underlie decision making in perceptual categorization by focusing on detailed analyses of the time course of information processing. Our computational approach combines parametric model fitting with non-parametric analyses to strengthen inferences and avoid problems associated with model mimicry. The computational models synthesize the information accumulation approach used to understand simple decisions with mental architecture models of serial and parallel processing, enabling predictions at the level of full RT distributions. By incorporating mental architectures, the models address fundamental questions about whether multiple dimensions are processed sequentially in serial fashion or simultaneously in parallel or pooled into a common processing channel.

Funded by

Australian Research Council Grant DP120103120

Australian Research Council Grant DP160102360

Photo by Dan Gold on Unsplash

Understanding Human Information Processing

The goal of this project is to promote advances and applications in Systems Factorial Technology (SFT), a theory-driven methodology aimed at identifying fundamental characteristics of human information processing.  SFT was developed by James T. Townsend and colleagues. The Knowlab has contributed to the development of SFT over the past 10 years most notably with the publcation of Systems Factorial Technology: A Theory Driven Methodology for the Identification of Perceptual and Cognitive Mechanisms.

Further applications include the examination of information processing in speeded cued detection tasks, the extension of methods of workload capacity to deal with distractors and conflicting information, and the extension of SFT to deal with errors.

Funded by

Australian Research Council Grant DP160102360

Photo by Jeremy Galliani on Unsplash

Human detection of fake news stories and popular misconceptions

The emergence of fake news and misinformation online highlights the difficulty of determining the quality of information distributed through the networked society.

The claims of hackers plotting to influence elections, the rise of anti-vaccination groups and various other potentially dangerous phenomena can all be traced back to the way that information is spread and interpreted online. In many of these cases, the misinformation has been designed to exploit cognitive biases and weaknesses in human cognitive architecture to convince people of an untruth or misconception.

The rise of this phenomenon is linked to societal, economic, technical and psychological factors. Combating the spread of fake news and misinformation online has become a critical issue and one not easily addressed through any individual discipline or solution.

This project seeks to better understand how people come to believe in misinformation and misconceptions in the networked society. We will then use this information to develop a proof of concept approach using data and analytics to predict when and how to introduce interventions to help people develop skill in critically evaluating the information they are exposed to online.

Funded by

Melbourne Networked Society Seed Funding Grant 2017

Publications

Searson, R., Lodge, J., Fidler, F., Bailey, J., Nolan, D. & Little, D. R. (2018). Human Detection of Fake News. Preregistration Document and Results. https://osf.io/u2qjb/