The PDS group focuses on understanding and predicting the psychology and behaviour of individuals. We bring together computational, theoretical and empirical methods to achieve new insights and to create technologies with social value.
While some of our work involves standard techniques such as test-based psychometrics and statistical analysis, our distinctive focus is on what we call “Implicit Psychometrics”: using non-traditional sources of information, such as gameplay or long-term patterns of behaviour, to make inferences about people’s abilities, preferences and/or future behaviour.
This approach is exemplified by the MoCHA system, a tool for monitoring the cognitive health of elders at risk for Alzheimer’s Disease (AD) and providing early warning of cognitive decline. Traditional diagnostic tests for AD are stressful, expensive, and require that the elder make an appointment with a clinician. By building psychometrics into a suite of (genuinely fun!) tablet-based games, MoCHA makes it possible to conduct assessments more frequently and conveniently in the patient’s own home. This system was developed by our team at IHPC, and is currently in an 18-month experimental trial conducted in collaboration with the Gerontology Research Programme at National University Hospital.
MoCHA System Demostration
In collaboration with industry and public-sector partners, we are using related techniques to analyze high-volume behavioural datasets and create predictive models of habitual behaviour and day-to-day urban decision making. As needed for particular projects we also exploit a wide variety of empirical methods, including experiments, in-person and crowdsourced surveys and field studies.
Our other current and recent research topics include:
- Computational modeling of “first impression” formation
- Study of urban transit wayfinding and service satisfaction
- Study of early-career scientists’ formative experiences and career ambitions
- Experiments investigating the effects of social division on attitudes to discussion of controversial issues
Figure 1. Model of "first impression" formation
Dr. Ilya Farber
Senior Scientist and Capability Group Manager