Dataset supporting the paper: Anxiety Biases Audiovisual Processing of Social Signals

Data collection method:

This dataset contains data for behavioural outcomes from two audiovisual emotion recognition tasks involving three different types of emotional stimuli. Participants were asked to categorise emotion for three types of stimuli: (1) dynamic faces and voices, (2) point-light displays showing the body motion of two interacting agents and corresponding dialogues, and (3) circles and tones varying in brightness and pitch. For each of the three stimulus types participants completed two separate tasks where they had to identify whether stimuli were happy or angry from unimodal (audio- and visual-only), congruent and incongruent audiovisual displays. The ‘Directed Attention Task’ aimed at examining the impact of anxiety on crossmodal interference by requiring participants to base their judgements exclusively on information from one modality (either audio or visual) while ignoring any concurrently presented information from the other modality. The ‘Undirected Attention Task’ did not require participants to focus their judgements on information from a particular modality, as they responded based on their ‘overall impression’ of the emotional content of the stimuli.

Data processing and preparation activities:

All the data in this dataset has been anonymised prior to sharing. Univariate outliers were identified as datapoints lying more than +/- 3 median absolute deviations from the median value. Outliers have been winsorized prior to sharing (i.e. replaced with the next most extreme non-outlier value in the dataset).

Additional information:

Information on the layout of the data has been provided in the README sheet in each Excel spreadsheet file.