Mechanismen der Reizdiskrimination: Reihenfolgeeffekte und das Interne-Referenz-Modell

01.11.2015 bis 31.10.2018

A central aim of psychophysical research is to quantify the human ability to compare and discriminate among stimuli. In classical psychophysical tasks designed to this purpose, participants are often asked to discriminate between a constant standard stimulus s and a variable comparison stimulus c. Intriguingly, discrimination sensitivity in such tasks depends on the temporal order of these stimuli, such that participants typically exhibit a higher discrimination sensitivity on trials with stimulus order "sc" than on trials with order "cs". This so-called Type-B effect is hard to reconcile with classical difference models of stimulus discrimination. It can be accounted for, however, by assuming that participants not only rely on the stimuli presented during a given experimental trial, but also take stimulation from previous trials into account. For example, according to the Internal Reference Model (IRM), participants build up an internal reference, which is dynamically updated from trial to trial and which is then compared to the second stimulus presented on the current trial.
During the first funding period, several basic predictions of IRM, such as Type-B effects and sequential trial-by-trial effects on perceived magnitude, were derived analytically as well as by means of Monte Carlo simulation. Empirical results for different modalities, task types, and stimulus magnitudes were generally closely in line with these predictions, thereby supporting the assumption of a dynamically updated internal reference. Moreover, we investigated the role of cognitive influences on the magnitude of the Type-B effect and designed program routines that allow fitting of separate psychometric functions to the different stimulus orders, thus facilitating the investigation of Type-B effects.
In the second funding period we aim to continue and expand our previous research in several aspects. First, while our previous studies mostly employed duration discrimination tasks, we now aim at generalizing the Type-B effect, and thus the basic mechanism of IRM, to other task domains and across modalities. Second, IRM predicts a distinctive pattern of results regarding discrimination sensitivity and perceived magnitude for the roving standard task, in which different standard stimuli are randomly intermixed. Therefore, a series of experiments will investigate this crucial set of predictions. Third, we plan to assess the stability of the internal reference over time as well as its robustness to interfering stimulation. Finally, a series of experiments is dedicated to investigate which stimulus features are represented in the internal reference. Taken together, in this second funding period, we seek to further enhance our understanding of the processes giving rise to order effects in stimulus discrimination and which compose a fundamental aspect of human discrimination performance.