Table 1.
Outcome variables list and measurement.
Table 2.
Spotify auditory features definitions.
Table 3.
Descriptive statistics for AM Likert ratings.
Fig 1.
Predictor variables included in PCA include nine auditory features: acousticness, danceability, energy, instrumentalness, liveness, loudness, speechiness, tempo and valence.
Fig 2.
Scatterplots of E-A and each auditory feature.
Correlation values are indicated in the top left corner of each scatterplot.
Table 4.
Estimate, standard error, t-value and p-values of linear mixed effects models of E-A for each outcome variable.
Table 5.
Estimate, standard error, z-value and p-values of mixed effects logistic regressions of E-A on reporting each outcome variable.
Fig 3.
Plots showing E-A and outcome variables.
A. Scatterplots and sigmoid function plots of E-A and MEAM emotional content outcomes B. Scatterplots of E-A and non-LIWC phenomenological characteristics. C. Scatterplots and sigmoid function plot of E-A and cue efficiency outcomes. D. Sigmoid function plots of E-A and LIWC phenomenological characteristic outcomes. Significance indicated as follows: *** p < 0.001; ** p < 0.01; * p < 0.05.
Fig 4.
Occurrence of reported AM-associated categorical emotions.
Values reflect sum of times each emotion was reported as associated with an AM. The top fourteen most commonly reported emotions used in analyses are indicated in dark blue.
Fig 5.
Sigmoid function plots of top emotion categories.
Significant models are indicated with asterisks as follows: * = significant at α = 0.0035 (0.05/14); ** = significant at α = 0.0007 (.01/14); *** = significant at α = 0.00007 (.001/14).
Fig 6.
P-value heatmaps of mixed-effects models including individual auditory features as independent predictors of MEAM outcome variables.
Note that E-A was not included as a predictor in these models but is added to these plots to allow the reader to compare the two analysis approaches. A. P-value heatmap for outcome variables excluding LIWC and categorical emotions. B. P-value heatmap for categorical emotions. C. P-value heatmap for LIWC categories. For plots A and C, * = significant at α = 0.05, ** = significant at α = 0.01, ** = significant at α = 0.001. For plot B, * = significant at α = 0.0035, ** = significant at α = 0.0007, ** = significant at α = 0.00007 reflecting Bonferroni corrections.