FB Kolloq WS 23/24: Heinrich Liesefeld

„Systematic Search: A new paradigm to measure search”


14. Februar 2024 16:15 – 14. Februar 2024 17:45
Termin herunterladen (.ics)

Gutenbergstr.18, Dekanatssaal

Dr. Heinrich Liesefeld, Universität Bremen in Kooperation mit dem Graduiertenkolleg
„Systematic Search: A new paradigm to measure search”

A beautiful finding has catapulted research on visual search to the forefront of cognitive sciences where it has since remained for over 40 years: when people look for a single feature (e.g., “red”), the number of stimuli in the display does not affect the time it takes to find an object possessing that feature. However, when people look for a combination of features (e.g., “red” and “horizontal”), search times increase as a function of the number of stimuli. This pattern has been accounted for by a stunningly simple model in which all stimuli are either processed in parallel or via a stimulus-by-stimulus serial search. Many straight-forward predictions from this model have been disproven and appar-ently simpler models assuming just a single (parallel or serial) search strategy have been proposed. Likely due to the initial focus on parallel vs. serial search, these subsequent models have focused on explaining variation in search slopes. And they were successful, too successful maybe: it appears that any of a dozen mechanisms can easily predict a continuous variation in search slopes and other signature patterns of search data. Thus, I shall argue, traditional search slopes are not suited as a criterion to distinguish search modes and models. Consequently, we have developed an alternative approach to measur-ing search slopes. Our new Systematic-Search paradigm reliably induces a spe-cific search order so that search slopes can be derived from the effect of target position. This paradigm also produces several empirical signature patterns that might serve to distinguish between search modes.


Dr. Heinrich Liesefeld