Final Thesis

Modulating Serial Recall Performance By Audiovisual Target Discrepancies

Key Info

Basic Information

Research Area:
Audio-visual Interaction
Type of Thesis:


Master Thesis of Burger, Julian

People perceive most events with several senses at the same time. In everyday life, these different sensations coincide in time and space to form one coherent percept. However, in online conferencing, which has become an increasingly popular mean of communication, there may be a temporal and/or spacial mismatch between visual and auditory input, e.g., due to delays in transmission. In this work, the effect of such temporal stimulus onset asynchronies (SOAs) on short-term memory (STM) is investigated. For this purpose, a test based on an audiovisual serial recall (aVSR) paradigm is developed. The digits from one to nine are presented to the participants auditorily, visually, or audiovisually with SOAs between -400 ms and 400 ms. Subsequently, the digits had to be recalled in the same order in which they were presented. Firstly, it is tested how recall performance is affected by the presentation modes, i.e., visual only, audio only, or audiovisual (AV) presentation. Secondly, it is investigated how the performance in the serial recall is influenced by the SOAs mentioned above. No effect was observed for the experimental variations, neither for the modality of presentation nor for the SOAs. The most likely reason for this is that the digits are too abstract as stimuli and thus no AV benefit, i.e., an improvement in performance compared to unimodal presentation, could be observed. Additionally, AV integration is not necessary to remember the digits, as both unisensory modalities seem to evoke complete representations in memory. Suggestions for changes to the stimuli and to the paradigm are presented and discussed.