Merging Data with R

[Howquez]

Hi,

Using DICE, you will end up with different datasets that you may want to merge for your analyses. Typically, you have data such as:

1. raw data, e.g., all_apps_wide_2024-08-15.csv
2. processed data you retrieved after uploading the raw data to the shiny app for pre-processing, e.g., DICE-processed-2024-08-15.csv. The shiny app helps you to transform the json data into a long csv file with participant x post level observations
3. your qualtrics data, in my case qualtrics_August+15,+2024_23.12.csv
4. the stimuli you created and uploaded to Github to create the feeds in the first place, I called the file brazil.csv for our brand safety case study
5. so-called page times data you can use to learn how much time participants spent on each of DICE's pages.

Below, I created an R script that shows you how I would merge 2, 3, and 4 (the page times require some pre-processing, so I will write an additional script for that one at some point--but you would merge it via the participant_code instead of the participant_label.)

In this example, I recruited participants via prolific and stored the respective PROLIFIC_PID in qualtrics (see embedded data fields. One can then merge this variable with DICE's participant_label.

# Packages (consider using groundhog) -----
# install.packages("data.table") # install this package first, if you don't have it
library(data.table)

# Read data -----
raw       <- fread(input = "data/raw/all_apps_wide_2024-08-15.csv")         # read only, never edit these
processed <- fread(input = "data/processed/DICE-processed-2024-08-15.csv")  # the data you get after you uploaded it to the shiny app: https://dice-app.shinyapps.io/DICE-Preprocessing/
times     <- fread(input = "data/raw/PageTimes-2024-08-15.csv")             # page times, available but not too relevant here. read only, never edit these
qualtrics <- fread(input = "data/raw/qualtrics_August+15,+2024_23.12.csv")[Finished == "TRUE"] # read only, never edit these
stimuli   <- fread(input = "data/stimuli/brazil.csv")                       # the csv file you created to design the stimuli

# Merge data sequentially -----
merge_1 <- merge.data.table(x = processed, y = qualtrics, by.x = "participant_label", by.y = "PROLIFIC_PID")
DICE    <- merge.data.table(x = merge_1, y = stimuli, by = "doc_id")

# Rename & remove some vars -----
DICE[, condition.y := NULL]
setnames(DICE,
         old = c("condition.x", "replies.x", "replies.y"),
         new = c("condition", "comments", "n_replies"))
rm(merge_1)

# View merged data -----
View(DICE)

I hope this helps.

[Howquez]

The data I analyzed are relatively old. The script should be compatible with updated versions of DICE + the Shiny app, though 🤞.