Readspeaker menu

Workshop 6: Prof. Dr. Frank Goldhammer - Writing (Computer-Based) Test Items

The major goal of the workshop "Writing (computer-based) test items" is to make participants familiar with fundamental concepts and steps of effective test development.  Special emphasis will be given to the test specification for the construct to be measured, and the following item development process. Moreover, (explanatory) item response models will be used to empirically test assumptions that governed the item writing process.

Content and method of instruction

The development of a valid test requires a systematic approach based on a good understanding of both the cognitive domain and educational measurement principles. The workshop will introduce into this process which starts with the definition of the construct to be assessed. Next, the test specification is developed by delineating the content and cognitive skills to be covered, and the various item attributes to be considered, e.g., response mode and psychometric characteristics. The test blueprint outlines how many items have to be written for each possible combination of content areas and attributes included in the test specification. The workshop will make participants familiar with important principles for writing traditional selected-response items (e.g., multiple-choice items). Regarding computer-based testing, the possibilities of innovative item and response formats will be presented and discussed. Finally, the workshop will show how explanatory item response models can be used to validate items empirically by testing the hypothesized influence of item attributes on difficulty.

The workshop will consist of lectures and hands-on-exercises on developing items for a given construct and on analyzing illustrative data using R.
Item writing examples from international computer-based large-scale assessments, e.g., PISA, PIAAC, and from own research work will be presented.

Intended audience

Participants should have a basic understanding of psychological testing, regression analysis and item response modeling (IRT).

Materials

Participants should bring their own laptop with R (http://cran.r-project.org/) and the package lme4 including associated packages pre-installed.



Biography              




 

Goldhammer





Frank Goldhammer received his doctoral degree in Psychology from Goethe University Frankfurt am Main (Germany) in 2006. He is professor for educational and psychological assessment and head of the Technology Based Assessment (TBA) group at the department for Educational Quality and Evaluation of the German Institute for International Educational Research (DIPF). His professorship is part of the Centre for International Student Assessment (ZIB) that was established by the Federal States and the Federal Government. His research interests include the psychological assessment of competencies and abilities (computer-related skills, intelligence, attention); modeling of responses, response times and process data; as well as computer-based testing in large scale assessment.

For further information please click here.