README

This is the reproducibility archive for the paper "Multimodality and Skewness in Emotion Time Series" (preprint: https://psyarxiv.com/qudr6). It allows one to reproduce all analyses and results in the paper.

Folders

• /Data contains the preprocessed data of the seven studies we re-analyzed in our paper. The scripts used to preprocess the data based on the data files provided by the authors can be found at http://github.com/jmbh/EmotionTimeSeries.

• /Files contains outputs created by the below scripts.

• /Figures contains the figures created by the below scripts.

Scripts

• aux_Functions.R contains auxiliary functions for plotting, estimating modes with various methods we compare in the appendix, and ways to generate data which is used for the validation studies in the appendix.

• EstimateModalitySkew.R loads the seven datasets, estimates the modality, and computes the skewness of each individual distribution, and outputs the files with file names EMS_... (see below).

• Evaluation_Figures.R takes the files A_DS_info.RDS, A_Modality.RDS, A_Densities.RDS and A_Skewness.RDS (see below) as input, and plots all the Figures in the main text.

• Illustration_Figure1.R plots Figure 1 with the 8 example distributions of the emotion sad.

• Additional_Analyses.R contains the code to produce Appendix F (looking into the relationship between skewness and the location of the distribution) and Appendix G (looking into the absolute model fit of the VAR model).

• Method_Validation.R contains the code to reproduce the validation of our Mode-estimation method in Appendices A.2 and A.3.

• Evaluation_ML_Skew.R loads estimates of skewness, data-set characteristics and between-person variables. Performs all multilevel analyses on skewness described in the main text Section 2.3.2 and Appendix E (Tables 3 and 4, Figure 13).

• Evaluation_ML_MModality.R loads estimates of modality, data-set characteristics and between-person variables. Performs the study-specific multilevel analyses on modality described in the main text Section 2.3.1; prepares data for 3-level analysis done in Julia (see file julia_analysis.jl).

• julia_analysis.jl contains the Julia code to perform the 3-level multilevel model on modality in Section 2.3.1 and Appendix D.

Files

• EMS_DS_info.RDS contains a list with various information on the dataset level (which items are included, which scale has been used, etc.).

• EMS_Modality.RDS contains an array with the modality estimates for all individual distributions; contains an NA if the distribution was excluded.

• EMS_Densities.RDS contains a list with the density estimates for all individual distributions, on which the modality estimates are based on; empty, if the distribution was excluded.

• EMS_Skewness.RDS contains an array with the skewness for each individual distribution.

• EMS_Exclusion.RDS contains a matrix indicating the counts for each exclusions separately for exclusion type and dataset.

• EMS_TSlength.RDS contains a list which contains matrices for each dataset, indicating the time series length for each subject.

• EMS_TSlength_IDs.RDS contains IDs for each individual in each study, including distributions that have been excluded.

• SimResults_Validation.RDS contain the results of the validation simulation in Appendix A.2.

• SimResults_Validation_Skew.RDS contain the results of the validation simulation in Appendix A.3.

• SimResults_Validation_Skew.RDS contain the results of the validation simulation in Appendix A.3.

• ModalityML_Data.csv dataset created by Evaluation_ML_MModality.R to run the 3-level multilevel model on modality in the file julia_analysis.jl.

• BetweenData.RDS contains between-person data for all datasets. Obtained from: http://github.com/jmbh/EmotionTimeSeries.