Abstract We introduce NEVi ( N egative E motional Vi deo dataset), a validated and standardized video dataset designed to evoke negative emotional responses. A total of 112 videos eliciting negative emotions were selected from established emotional video datasets. NEVi stands out by offering matched videos in two durations: short (1-sec) and long (5-sec), both individually extracted from the emotional video stimuli, taking into account the point of the highest intensity and the comprehensibility of the content. A total of 650 international, English-speaking volunteers evaluated the dataset by rating the stimuli on the dimensions of valence and arousal. Videos were categorized by intensity (low and high) based on these ratings. Particular care was taken to ensure the suitability of the video content for younger audiences, making it appropriate for use with adolescents. Results are provided as CSV files, containing participant demographics and ratings (overall and by gender) for valence and arousal. The data (CC-BY) are fully ‘open access’. Because stimuli come from third-party sources, we do not redistribute videos; instead, information is provided for reconstruction.
Schurig et al. (Tue,) studied this question.