• Blümke Matthias

• Blümke Matthias

Many instruments for assessing psychological constructs or health-related quality of life are applied to different groups (e.g., different ethnic background of participants, multi-center studies, cross-national samples). This two-day course provides a quick introduction to the standard procedures of constructing a reliable and valid measurement instrument on the basis of (quantitatively analyzed) questionnaires (so-called “scales”). Without going needlessly into the statistical details of classical test theory or factor analysis, current state-of-the art recommendations for scale construction will nonetheless be reviewed for practical applications (content validity, item selection, dimensionality, factorial validity, reliability estimates). Short scales can be particularly useful for individual screening or research purposes, but they are also heavily critized, so their utility needs to be discussed against the distinction of clinical testing and personality assessment.

As a first objective, the course will outline how to implement procedures for developing shortened scales that meet reliability and validity criteria. A second objective is to present procedures that test whether instruments are fair and unbiased for different groups (e.g., gender, age, cohorts, education, ethnicity, culture/language). To this end, the basics of testing measurement equivalence within the multi-group confirmatory factor analytical framework will be introduced, and the fundamental statistical testing procedures will be presented. All focal procedures will be demonstrated in R, so that course participants can adapt them to their own needs in their later practice. Prior experience with R as well fundamental knowledge of regression analysis, factor analysis and structural equation modeling are essential for participation. Course credit is due for a successful assignment that demonstrates the ability to shorten a questionnaire suitably on the basis of existing data (e.g., public databases) and evaluate its utility in line with the test criteria learned in the course.

Tutorial parts during the course will demonstrate how to implement statistical procedures in R/RStudio. Students are invited to use the PC-environment in Fribourg, alternatively to bring their own laptop with them. Details on the required software and package versions will be provided to enrolled participants about 1 week in advance.

Assignment: Hand in a seminar paper (Hausarbeit) that demonstrates empirically the student’s own shortening and testing of an instrument. The assignment will be marked anonymously.

Assignment: Hand in a seminar paper (Hausarbeit) that demonstrates empirically the student’s own shortening and testing of an instrument. The assignment will be marked anonymously.

Assignment: Hand in a seminar paper (Hausarbeit) that demonstrates empirically the student’s own shortening and testing of an instrument. The assignment will be marked anonymously.

Assignment: Hand in a seminar paper (Hausarbeit) that demonstrates empirically the student’s own shortening and testing of an instrument. The assignment will be marked anonymously.

Assignment: Hand in a seminar paper (Hausarbeit) that demonstrates empirically the student’s own shortening and testing of an instrument. The assignment will be marked anonymously.