Pre-conference workshop: Rasch analysis using open source software for reproducible science: RISEkbmRasch R package and Quarto

This workshops walks participants through the process of importing data into R and making use of a simple code template for the different components of a Rasch analysis. The content will focus on various aspects of analysis of polytomous questionnaire data using the partial credit model. These include analyses of item fit, relative residual correlations, dimensionality, targeting, item hierarchy, differential item functioning, and reliability. Commonly used modifications of data, such as removing items and merging response categories, will be demonstrated. 

The workshop will both use the online package vignette as an example as well as live coding and analysis of a dataset provided by a workshop participant (or downloaded from an open data repository).

Participants are encouraged to bring their own datasets to do work on during the workshop. The second part of the workshop focuses on participants analyzing their own data, and the workshop leader will provide hands-on support.

The source code used and produced during the workshop will be shared openly on GitHub.com

By using Rstudio and Quarto, the process of analysis and its documentation can be kept in the same text file, which can be rendered into different output formats such as HTML, PDF and Word files. This ensures traceability of decisions and actions made during the analysis, as well as reproducibility of results, provided one has access to the datafile analyzed.

Participants will need to have installed Rstudio (version v2023.12 or later) and R (version 4.1 or later), instructions are available here: https://github.com/pgmj/RstudioQuartoIntro

The R package should also be installed prior to the workshop, see https://github.com/pgmj/RISEkbmRasch for instructions.

Program:

• 09.00-10.15 Session1: Using a template to set up a project and import data.
• 10.15-10.45 Break
• 10.45-12.00 Session 2: Live coding & analysis of a new dataset, and participant can work with their own data ande get hands-on support.