Action Prediction Laboratory

Action Prediction Laboratory

How are we able to use our knowledge of others from prior interactions to predict their future behaviour?

This research is based on the concept that we implicitly accumulate knowledge about how other people respond in certain situations. Specifically, whether there is reactivation of information for the typical actions of others during current social interactions. It may be that this occurs by mapping the behaviour of others onto our own motor system to enable understanding and prediction of their behaviour.

In a first series of studies participants are presented with photographs of actors acting towards or away from objects to produce covariations between actors and their 'preferred' objects.  This enabled the analysis of differences in response when they behave as expected (i.e., by acting towards their preferred object and turning away from another object) and when their behaviour goes against expectation in oddball trials. The findings suggest that irrespective of task (person or action identification), responses are faster and less erroneous when actors acted towards their preferred object and turned away from another object, but slower and with more errors when the actors acted towards a different object and turned away from their preferred object.  This suggests not only that the participants had formed associations between actors and how they tend to interact with certain objects, but that these associations influenced their responses. What is really crucial is that this seems to be a highly implicit process because none of these participants were able to explicitly state the covariations when asked in a funnel debrief.

Based on this working paradigm, the investigation will move on to look at the potential for mapping this activity onto our own motor systems and will test the generalisability of actions with similar stimuli.  These further investigations will continue in both rigid and controlled lab settings and in more 'real-life' situations.  This research will not only inform us of how we act during social interactions, but it may also highlight how and where potential impairments in underlying structures might impinge upon social interactions.