Process and analyse IAT data from gorilla

Runs a full analysis on data from an Implicit Association Test (IAT) that was run in gorilla using the standard template. It picks out the trial data, calculates each person's mean reaction time in the congruent and incongruent blocks, and tests the difference.

Usage

gorilla_iatanalysis(
  data = NULL,
  outp = "analysis",
  exclude_pids = NULL,
  exclude_items = NULL,
  ...
)

Arguments

data

The list of data from a gorilla import. If missing, you will be asked to choose a folder of gorilla downloads.

outp

Name of the folder to save results in (it is created if needed). Set to NULL or "" to not save anything, and just return the results.

exclude_pids

Vector of participant ids to leave out.

exclude_items

Vector of items (stimuli) to leave out.

...

Further arguments passed to plot_model(), and from there to pirateye(), to change the plots.

Value

A list. The first (unnamed) item is the output of plot_model(), then ParticipantMeans (one row per person, with the IAT score) and rawdata (the selected trial data).

Details

The analysis follows these steps:

1. Trials are selected where metadata is "congruent" or "incongruent", at the first attempt, in the display called "trials". Incorrect trials and those with reaction times over 3000 ms are removed.

2. Mean reaction times are calculated for each person, and participants with a missing condition are dropped. The IAT score is incongruent minus congruent.

3. A paired t-test is run and saved in IAT_ttest.txt in outp.

4. A mixed ANOVA is run with block order (firstblock) between participants and condition within participants. It is plotted and printed using plot_model().

5. The participant means are saved as ParticipantMeans.csv in outp.

See also

plot_model(), data_collator_gorilla()

Examples

if (FALSE) { # \dontrun{
data <- data_collator_gorilla("gorilla_data")
iat <- gorilla_iatanalysis(data, exclude_pids = c(1234567))
iat$ParticipantMeans
} # }