implicitMeasures

R package for computing different Implicit Measures scores

Aim and Overview

The goal of implicitMeasures package is to provide a tool for easily cleaning and computing the scores for the Implicit Association Test (IAT; Greenwald, McGhee, and Schwartz 1998) and the Single Category Implicit Association Test (SC-IAT; Karpinski and Steinman 2006).

Six different algorithms for the computation of the IAT effect, the so-called D-score, are available (Greenwald, Nosek, and Banaji 2003). The six algorithms differentiate themselves for the treatment of the extreme fast responses and for the strategy used for replacing the error responses. Different mistakes can be made during the computation of the D-score, which can raise replicability issues of the results. On top of that, many researchers fail to report the exact algorithm they have used for computing the D-score, raising again replicability issues (Ellithorpe, Ewoldsen, and Velez 2015).

implicitMeasures includes the following functions:

• clean_iat(): Prepare and clean IAT data.
• clean_sciat(): Prepare and clean SC-IAT data.
• compute_iat(): Compute IAT D-score.
• compute_sciat(): Compute SC-IAT D-score.
• descript_d(): Descriptive table of D-scores (also in LaTeX format).
• d_density(): Plot IAT or SC-IAT scores (distribution).
• d_point(): Plot IAT or SC-IAT scores (points).
• multi_dscore(): Compute and plot multiple IAT D-scores.
• multi_dsciat(): Plot SC-IAT D-scores.
• IAT_rel(): computes the IAT reliability (Gawronski et al. 2017)

All the functions for graphically representing the results are based on ggplot2 (Wickham 2016), and can be further customized by the users.

Installation

You can install the released version of implicitMeasures from CRAN with:

install.packages("implicitMeasures")

and the development version from GitHub with:

# install.packages("devtools") # un-comment to install devtools
devtools::install_github("OttaviaE/implicitMeasures")

Example

This is a basic example which shows you how to compute the IAT D-score. More detailed examples are given in the package vignettes.

library(implicitMeasures)
# load the raw_data dataframe
data("raw_data")

# prepare the dataset for the computation
iat_data <- clean_iat(raw_data, 
                          sbj_id = "Participant",
                          block_id = "blockcode",
                          mapA_practice = "practice.iat.Milkbad",
                          mapA_test = "test.iat.Milkbad",
                          mapB_practice = "practice.iat.Milkgood",
                          mapB_test = "test.iat.Milkgood",
                          latency_id = "latency",
                          accuracy_id = "correct",
                          trial_id = "trialcode",
                          trial_eliminate = c("reminder", "reminder1"))

# store the dataset for computing the D-score
iat <- iat_data[[1]]

# Compute the D-score
dscore <- compute_iat(iat, D = "d3")

compute_iat() results in a data frame with class dscore. This data frame can be passed to other functions, for example for plotting the results, either at the individual level:

or at the sample level:

Data import

You can import your data in any format you want. If you are importing data sets from SPSS, please use either haven::read_sav("~/path/to/mydata.sav") or foreign::read.spss("~/path/to/mydata.sav") without changing the default options of the functions. implicitMeasures will recognize that the data frame is coming from SPSS and will handle that.

Bugs and problems

If you find any bugs or encounter any problems in using this package, please post a minimal reproducible example on github. For questions and other discussions, you can contact the author and maintainer of the package at otta.epifania@gmail.com.

Contributing to implicitMeasures

If you want to contribute to implicitMeasures, by all means! You can open a new branch on https://github.com/OttaviaE/implicitMeasures, modify the code, and submit your pull request for added features.

References