Analisi dell’associazione tra generi cinematografici e resoconti in tempo reale dell’esperienza emotiva

Movies are ecological stimuli eliciting multifaceted emotional experiences. Yet, studies focusing on movie excerpts have possibly underestimated the role of contextual information and the importance of emotion dynamics. Moreover, a detailed analysis of the relationship between movie genres and the emotion they bring about is lacking. Here, we evaluated whether and to what extent genres explain real-time affective reports during movie watching. A total of 116 participants (56F; mean age = 26.25; SD = 3.55) watched 60 full-length movies covering 18 different genres and continuously rated their emotional experience using 20 categories (e.g., joy, contempt). We estimated the average occurrence of the 20 emotions in each film and measured the cosine similarity between movies based on affective reports. Also, we computed films’ similarity on a movie-by-genre matrix. Kendall’s τ assessed the correlation between the two matrices. Movie similarity based on affect significantly correlates with genre (τ = .38, p < .001). Valence explains 46.71% of the variance in affective ratings (e.g., coef: Anguish = –.54; Amusement = .63) and 24.25% in genres (e.g., Drama = –.56; Comedy = .63). Valence scores obtained from affective reports correlate with the first principal component derived from movie categories (τ = .53; p < .001). Graph analysis reveals 8 clusters of movies based on affect and 9 on genre, with a Jaccard similarity index between the two partitions of 0.124 (only 45 movie pairings belonging to the same cluster between graphs). Overall, our findings show that emotions elicited by movies relate to genres. Yet, genres alone do not capture emotional nuances in time-varying reports of experience, as testified by the moderate correlation. Contextual information is crucial in eliciting complex states and should not be disregarded by emotion research. Affective reports could complement the taxonomy based on movie genres and used to refine film recommendation systems.