This is the research compendium that accompanies the paper Developing and Applying IR-Tree Models: Guidelines, Caveats, and an Extension to Multiple Groups (Plieninger, 2020). Herein, documentation and code is provided for the empirical example. Furthermore, a simulation study is reported that was removed from the final paper.
In the empirical example, data from the International Social Survey Programme were used, namely, from ISSP 2015 – Work Orientations IV (ISSP Research Group, 2017). The data are not part of this compendium in order to prevent copyright issues. To do further analyses with these data, the following steps are needed:
- Please download the SPSS (*.sav) data file version 2.1.0 from doi:
10.4232/1.12848 and save it as
data-raw/issp15/ZA6770_v2-1-0.sav. - Run the two R scripts in
/data-raw/issp15/. - Further documentation is provided in the README file in
/code/issp15/.
A simulation study was conducted to demonstrate that improper models (i.e., using incorrect pseudoitems) can lead to serious consequences and invalid conclusions.
- The data were simulated using the R script
/code/simulation/01_generate-data.R. - A graphical overview of the generated items and their parameters can
also be found in
/figures/sim-study_items.pdf. - Further details can be found in the README file in
/code/simulation/. - The results are reported in
/simulation-study/.
Mplus code for single-group models as used in the simulation study can
be found in /code/simulation/mplus/. Mplus
code for multiple-group models as used in the empirical example can be
found in /code/issp15/mplus/.