JosueMA / epe

Data and code for Emotion Prediction Errors

epe

Behavioral data and analysis script for "Emotion Prediction Errors Guide Socially Adaptive Behavior"

The manuscript is published in Nature Human Behavior at the following doi: [to add]

Setup

Clone repository or simply download in a zip file.

The R Markdown file in the Analysis folder contains all the analysis code for the paper, as well as a html version, and the supplement. In order to run the code you'll need the following packages up-to-date:

library(here)         # relative paths
library(tidyverse)    # tidy functions
library(knitr)        # knit functions
library(kableExtra)   # extra markdown functions
library(lsr)          # cohenD()
library(lme4)         # mixed-effects regressions
library(lmerTest)     # mixed-effects regressions
library(broom.mixed)  # tidy()
library(AICcmodavg)   # predictSE()
library(cowplot)      # plot_grid()
library(ggrepel)      # geom_text_repel
library(gridExtra)    # aesthetics 
library(permutes)     # permutation testing
library(doParallel)   # parallel computing 
library(purrr)        # map functions
library(sjPlot)       # clean tables
library(ggstance)     # vertical position dodge
library(mousetrap)    # mouse tracking analysis
library(clusterperm)  # cluster-based permutation
library(exchangr)     # shuffle function

Folders

This repo contains the following folders: analysis, data, and graphs.

1. analysis

   Main analysis script can be found here alongside an html Markdown file showing all results and code. Supplementary analysis script can be found here as well.

2. data

   Cleaned data can be found here for all experiments separately. Note Data files for Experiment 3, involving mouse tracking, are too large to host in Github but can be found in an OSF repository here. You'll want to place these in the data/study3_jg/ folder or change the path locations in the analysis script.

3. graphs

   Graphs produced for manuscript from R. Note that some style changes have been added to the final graphs in the manuscript using Illustrator, but in general I try to keep the graph output from R as close as possible to the final graph.

Contact

If you have any questions or suggestions please feel free to open an issue on this repo or email me directly at [e-mail].