1st September

Picking Up Patterns: An Individual-Differences Perspective on Statistical Learning

Louisa Bogaerts

Department of Experimental Psychology, Ghent University

Humans show remarkable sensitivity to the statistical structure of their environments, suggesting that we are less dependent on effortful learning and explicit instruction than once assumed. As a result, statistical learning has been proposed as a core mechanism supporting a wide range of higher-level cognitive functions, including language. My lab’s research has focused on individual differences in visual and auditory statistical learning. Performance varies substantially and meaningfully across individuals: while some learners readily extract repeating sensory patterns (demonstrating reliable recognition and processing advantages for predictable stimuli), others show little sensitivity. Recent advances in the measurement of statistical learning have made it possible to characterize this variability among learners more reliably.

In this talk, I will share what our individual-differences approach has thus far revealed about the structure of the human statistical learning system (or rather, systems?) and its predictive relations with language skills. Overall, the findings provide evidence for a multi-componential ability. Whereas a meta-analysis indicated a significant but modest association between statistical learning ability and language, our latest work using latent variable methods provides evidence for a much stronger interconnection.

2nd September

Working memory and episodic long-term memory

Klaus Oberauer

University of Zurich

The relation between working memory (WM) and episodic long-term memory (LTM) has been a matter of intense debate for several decades: Are they different at all, and if so, how do they differ, and how can we distinguish them empirically? I propose to approach these questions through a functional analysis. The function of WM is to hold the currently most relevant information available for processing. This requires that the contents of WM can be rapidly updated, and that WM protects its contents against proactive interference from outdated information. Episodic LTM, by contrast, has the function of maintaining a record of past events that are potentially relevant in the future. These records should be robust against updating or loss, and as a consequence, are vulnerable to proactive interference. Moreover, whereas WM is vulnerable to disruption by unrelated processing, LTM should be immune to such disruption. Experiments confirm these predictions, constituting a double dissociation of WM and episodic LTM. Nevertheless, keeping them separate in practice is challenging: On some tests of WM participants draw on episodic LTM when WM is overloaded; with other tests they rely purely on WM. Evidence so far suggests that LTM is recruited only when doing so yields a better chance of success than relying on WM alone.

3rd September

Using robots to understand human cognition

Agnieszka Wykowska

Italian Institute of Technology

As robots are believed to soon populate human environments, they have received enthusiastic support in the scientific community. Most research aims at designing robots for assisting humans in daily lives, healthcare, or elderly care. However, there is also a less explored way of using robots - robots as tools to understand human cognition. In my lab, we take this approach, examining human cognitive mechanisms in social interaction. In this talk, I will present the work from my lab where we have examined how attentional orienting, sensorimotor processes, sense of agency and cognitive control unfold in interaction or joint action with others – with natural and artificial partners. Our results showed, both at the behavioural and neural level, that interaction modulates attention, sensorimotor processes, sense of individual and joint agency, as well as cognitive control. I will discuss these results in a broader context of using robots at the service of psychological research.