GAUDIE: Development, validation, and exploration of a naturalistic German AUDItory Emotional database
Abstract
Despite a great need in the German-speaking area, thoroughly validated naturalistic speech stimulus databases for emotion induction are rare. Therefore, we present GAUDIE (German AUDItory Emotional Database) - a validated, richly annotated and online accessible stimulus database of German speech sequences for emotion induction. GAUDIE comprises 37 audio speech sequences with a total length of 92 minutes. The audio sequences last between 1 and 4 minutes and induce positive using comedian shows, neutral using weather forecasts, and negative emotions using arguments between couples and relatives. GAUDIE was validated by 26 native German speakers. In naturalistic scenarios, experienced emotions are highly context-dependent and reveal a high variability over the course of time. Therefore, participants were asked to provide continuous ratings during the listening experience tracking the time courses and variability of valence and arousal. These ratings are key features of the database. These continuous ratings allow future researchers to link the behavioural time series to evoked (neuro-)physiological reactions. Additionally, post-presentation ratings examined discrete emotion classification and potential moderators. For the assessment of stimulus quality, we quantify how well audio sequences differentiate on the valence-arousal-dominance system and generalize regarding the perceived emotional strength and other ratings across participants. To find an optimal stimulus selection, we used Monte Carlo Simulation based subtractive comparisons. The database GAUDIE annotated with multiple emotion ratings fills a gap as a German emotion inducing speech database. All stimuli, along with their annotations, can be accessed online through the OSF project repository GAUDIE.
M.Sc. Katharina Lingelbach
PhD student, Neuroscientist, and Psychologist (she/her/hers)
My research focuses on interacting neuronal dynamics of emotional and cognitive processes, decoding approaches of neuronal dynamics using machine learning using electrophysiology and neuroimaging methods