This repository contains all scripts used to clean, wrangle, visualise, and prepare the data for subsequent analysis by Mihaela Taranu.
In the experiment, 7 children played with 14 LEGO bricks across 3 consecutive iterations. Eye tracking data were recorded with Pupil Labs. The videos extracted from Pupil Labs were subsequently manually annotated in Boris. Each time a new fixation was made on a different brick a new manual code was annotated. Thus, the raw data partly originates from Pupil Lab and partly from Boris. Both Pupil Lab and Boris generated a csv-file for every iteration (block) resulting in a total of 21 Boris files and 21 Pupil Lab files. As information from both the Pupil Lab and the Boris files are crucial for the study, all the files are being gathered into a single csv-file.
This repository has the following directory structure:
| Column | Description |
|---|---|
data |
The folder containing all data including the output from Pupil Lab and Boris as well as the cleaned and gathered files. |
documents |
The folder with documents related to the project including the first explorative report, project description, plan etc. |
scripts |
The folder with all R-markdown files used to pre-process, clean, and gather the data. |
LICENSE |
An MIT license |
README.md |
This readme file. |
There are several scripts and they should be executed in an appropriate order.
Clean and gather the Boris files
The script 'clean_boris_files.Rmd' will import all files in the folder Boris_cleaned, gather them into a single dataframe and drop redundant columns. The single output file will be stored in the folder gathered_files as 'boris_cleaned.csv'.
Clean and gather the Pupil Lab files
The script 'clean_pupil_files.Rmd' will import all files in the folder Pupil_Labs_raw_data, gather them into a single dataframe and drop redundant columns. The single output file will be stored in the folder gathered_files as 'pupil_cleaned.csv'.
Merge the Pupil Lab and the Boris data
The script 'merge_data.Rmd' will take the two files, 'boris_cleaned.csv' and 'pupil_cleaned.csv', and merge them according to approximate timestamps. The timestamps in 'boris_cleaned.csv' are relative to each block of the experiment while the timestamps in 'pupil_cleaned.csv' are not. By subtracting the start recording timestamp of each block, the timestamps corresponding to the timestamps in 'boris_cleaned.csv' can be obtained. A more elaborate explanation of this method can be found here. The start recording timestamps are provided in the metadata of Pupil lab in the folder export_info under Pupil_Labs_raw_data.
Prepare the data for analysis
The script 'prepare_analysis.Rmd' will import the merged file 'dataAll.csv' containing both the Pupil Lab and the Boris data and prepare it for analysis and visulisations. A predefined function will be employed to filter out all rows that are not actual fixations (other behaviors).
The data pre-processing as well as visualisations have been conducted by Aske Svane Qvist, research assistant on the project. If you have any comments or questions as to how the data cleaning, tidying, and visualisation have been done, feel free to contact me:
Aske Svane Qvist
Mail: askesvane@gmail.com
Tlf. +45 52430411