R-Forge Logo

Welcome to klausuR: Multiple Choice Test Evaluation project!

A package to generate results of multiple choice tests: Detailed global results, anonymous/personalised feedback (LaTeX or PDF), item statistics like Cronbach's alpha or disciminatory power. Supports parallel test forms and partially answered items.

What can you do with klausuR?

This package is intended for persons or institutions who/which apply Multiple Choice (MC) tests. Work on it started off in 2009, as a collection of functions we used in our own department (University of Düsseldorf, Institute of Experimental Psychology, Department of Diagnostics and Differential Psychology). Its main purpose is to evaluate results for each test subject, assign marks accordingly, and produce printable summaries (e.g., to be archived with the test, to be able to verify the correctness of results any time).

Over time the package gained a lot of features we don't actually needed ourselves, but which might be interesting for research. For instance it cannot only apply the typical Number Right scoring (one point for each correctly marked alternative), but also Coombs Elimination Testing and Number Right Elimination Testing (development version).

Core features

Graphical user interface

There is also a GUI to work with the package, implemented as a plugin for RKWard. After installation the pluginmap can be found as $R_LIBS_USER/klausuR/rkward/klausuR.pluginmap

Development snapshots

If you want to try the bleeding edge unstable development version (i.e. SVN snapshots), add http://R.reaktanz.de to your repository list. But be aware that "unstable" means unstable.

Limitations

Some of the advanced features, especially in the development version, are not extensively tested in all possible combinations. If you should run into a bug, please don't hesitate to send a report! But even if everything runs smoothly, please do always check at least a sample of the results to make sure no errors occurred.

The project summary page you can find here.