Here we provide a shiny app that can be used by other researchers to code qualitative ESM data based on a given coding scheme. It contains several columns showing the raw observation (see screenshots below “activity_open”) and the given code (“Code”). To accommodate information from different coding dimensions (e.g., company mentioned alongside activities), our app includes two additional columns for capturing such information. The first column, “Additional Information,” is intended for details about participants' feelings and the social context during activities. The second column, “Other/Comments,” can be used for any other relevant information or observations. Our codes have a hierarchical structure and are therefore displayed in different colors.
This app was build as part of the project “Assessing Daily Life Activities: Comparing Predefined Categories with the Qualitative Analysis of Open-Ended Responses in Experience Sampling Methodology (ESM)” Marie Stadel, Anna M. Langener, Katie Hoemann, Laura Bringmann.
To facilitate the coding process, a list of codes appears when clicking on the “Code” field, and matching codes are suggested interactively while typing.
To use this app, please download and open the app.R file in R studio (https://posit.co/download/rstudio-desktop/). If desired, you can also download a codebook. We currently provide two options, one based on Stadel et al. 2024 (Codebook_Stadeletal.csv: https://annalangener.github.io/QualitativeVis/) and one based on Skimina et al. 2020 (Codebook_Skiminiaetal.csv: https://annalangener.github.io/QualitativeVis/Skimina.html). It is also possible to use your own codebook.
The app runs locally, so there are no privacy concerns when using it.
The following sections explain how to set up this app. You only need to change the first few lines of code in the app (Six lines in total). Here is a little more explanation of what you need to change.
First, you need to select your project folder where you want to save your coding results. Ideally, this folder will also contain the file where your codebook is stored. This can be any folder on your computer or in a cloud.
# 1. Select the folder in which the codebook is stored and the results will be saved
Projectwd <- "/Users/annalangener/Nextcloud/Shared/Testing Methods to Capture Social Context/Qualitative context/3. Coding/QualitativeCoding_Activies/"Next you need to read in the codebook. In this example, we use the Codebook_Stadeletal.csv file, but in theory you can use any codebook. Below we also provide sample code to read a codebook stored in an Excel file. The codebook should look similar to this:
Important: The codebook must contain a column named “Code” containing the codes. If desired, the codebook can contain a second column called “Level”. This can be used if the codes have a hierarchical structure and you want to display them in different colors.
# 2. Select the coding scheme that you want to use.
# You can choose our proposed coding scheme, the coding scheme proposed by Skimina et al., or you own by specify the path where the codebook is stored
# IMPORTANT: The codes need to be in a column named "Code" and if levels are included, those need to be in a column called "Level"
#Codebook_Act <- read_excel(paste(Projectwd,"Codebook_shared_activities.xlsx",sep =""), sheet = 1)
Codebook <- "Codebook_Stadeletal.csv" # Needed
Codebook_Act <- read.csv(paste(Projectwd,Codebook,sep ="")) Specify where your data is stored in your project folder. In our example, we created a subfolder called “Data” that contains our data file. See the image below for an example of how our project folder is structured.
To be able to code “context” (e.g. someone is on vacation), the dataframe should be sorted by date. Below we provide sample code on how to do this.
# 3. Select the path where the data is stored
Data <- read.csv(paste(Projectwd,"Data/act_coding_ALL.csv",sep = ""))[,-1]
# The dataframe should be sorted by Date, to allow for context coding
#Data <- Data %>% arrange(ppID, timeStampStart)
#write.csv(Data,paste(Projectwd,"Data/act_coding_ALL.csv",sep = ""))The next step is to select who will be coding (this can be useful if more than one researcher is coding on a project). You can choose any name you like. If you choose a new username, a subfolder will automatically be created for that user (see image below), if an “old” user is selected, the previous codes will be read into the app and the user can continue coding where they left off.
# 4. Select who is coding (a folder will be created if this is a new person)
User <- "Example_newuser" # "Marie_FullCoding", "Marie", "Anna"In this example, a new folder named “Example_newuser” automatically be created in our project folder to store the results for this user.
5) Indicate how your column is named that includes the participant IDs and select the participant that you want to code
To reduce loading time and to be able to code the context of an observation, we load each participant’s data separately. The first time we load a participant’s data, a csv file is automatically created to store the coding results (until there are results to store, it will contain “NA”).
# 5. Indicate how you column is named that includes the participant IDs and select the participant of interest
id_column = "ppID" # Change the name of the column here
ppID <- 106 
It is possible to use a hierarchical coding system with codes at multiple levels. Each level then is displayed in a different font color. You can set Levels to FALSE if your codebook does not have a hierarchical structure or if you do not want to display them in different colors.
# 6. Indicate whether your codebook contains different levels?
Levels = TRUE #FALSELastly, you only have to click “Run App”.
This repository also contains some simple R code to start analyzing the data: Results_Analysis.Rmd. For example, we provide some code to read the results from each participant and calculate the frequency of how often a particular code was used.
Write an email to anna.m.langener[at]dartmouth.edu, langener95[at]gmail.com, or m.stadel[at].rug.nl