Alumni talks: Felix Ball (Magdeburg) and Luke Tudge (Berlin)
Hosted by Thomas Christophel (hayneslab)
Ball Abstract: In real life, we are exposed to a rich environment, a complex and continuous stream of multisensory information. This information needs to be integrated to generate a reliable mental model of our world. There is converging evidence for several optimization mechanisms to integrate incoming information, among them are multisensory interplay (MSI) and temporal attention (TE). Both these mechanisms are known to enhance perception: whenever stimuli are presented at expected moments in time or when stimuli are multisensory rather than unisensory, response times are shortened and discrimination sensitivity is enhanced. Previous research focused on the influence of temporal expectation on perceptual processing mainly in unisensory auditory, visual, and tactile contexts. We used a different approach and tested – in a series of experiments – whether temporal expectations can enhance perception in multisensory contexts and whether this enhancement differs from enhancements in unisensory contexts. First, I will present data from a series of behavioral experiments indicating that visual, auditory and audiovisual temporal expectation effects differ in strength and are not necessarily identical. Next, I will present data indicating that the aforementioned effects are not necessarily modulated by top-down control and are most likely driven by automatic stimulus processing. Finally, I will present some preliminary EEG results with which we try to pin down the temporal profile and the differences of multisensory temporal expectation effects. Together, our results suggest that participants benefit from multisensory stimulation (relative to unisensory stimulation) and that this enhanced informational content enables the robust extraction of temporal regularities, highlighting the need for of multisensory paradigms in future studies investigating temporal expectations and other phenomena.
Tudge Abstract: My job is to teach statistics to students from a variety of cognitive-science-related disciplines. Among my colleagues in the profession, there is fairly wide agreement that this job is sometimes quite difficult, but there is considerable dispute as to why. I will present the case for two broad answers to the question ‘Why is stats hard?’. The cognitive mismatch view holds that our minds are just inherently ill-equipped for explicit quantitative inference; we shouldn’t expect to find any ‘magic bullet’ method of teaching that makes it easier. The content mismatch view holds instead that it is not statistics in general that is hard, but rather the particular idiosyncratic approach to statistics that has historically been taught in certain disciplines: Hypothesis testing. There are alternative approaches to statistical inference that may be more intuitive and therefore easier to teach. I will briefly review some arguments and evidence for each of these two views, and focus in particular on two competing proposals from the content mismatch camp about how to frame statistical inference more intuitively: statistics as updating degrees of belief, and statistics as estimation and prediction.
The talks are followed by drinks at Café Flora Soft, Haus 19, HU Campus Nord, 10115 Berlin
