ISSID-Graz, 2003 abstract; Dr. J.M. Müller
Comparison of the NEO-FFI, EPP, 16PF-R, EPQ-R and TIPI Extraversion-Scale:
A new approach in proofing content validity by comparing the rasch-scaled variance of person parameters
Content validity is a central concept of personality assessment. A traditional way to find out what a test is measuring, is to perform a validity study in the sense of Multi-Trait-Multi-Method (Campell & Fiske, 1959; Eid, 2000). Item-Response-Theory open the opportunity to proof content validity by assuring the homogeneity of items. A new approach to proof content validity of several tests, when data comes from several person sample, was given by Müller (2002a,b,c). Referring to the argumentation that the rasch-scaled variance of the person parameters should be equal, this gives the opportunity to proof this hypothesis, and stricter than by correlation based inferences. Therefore the norm-data of five personality questionnaires, namely the NEO-FFI (Borkenau & Ostendorf, 1991), EPP-D (Bulheller & Häcker, 1998), FPI-R (Fahrenberg, Hampel & Selg, 2001), EPQ-R (Ruch, 2001) and TIPI (Becker, 2001), are reanalysed and a supplementary item selection process was done to yield comparable questionnaires in item number and item difficulty. The results are discussed recommending several influencing aspect in estimating the rasch-scaled variance of the person parameter. The approach should illustrate a new methodology and first empirical result, if – or if not – the questionnaires are measuring the same intended extraversion dimension.
Müller, J. M. (2002a). Unterschiedliche Variationen in psychologischen Eigenschaften - eine Interpretation der Erstreckung einer Raschskalierung. Zeitschrift für Differentielle und Diagnostische Psychologie, 23, 261-271.
Müller, J. M. (2002b). Die testübergreifende Bedeutung von Raschvarianzen als Maß der Unterschiedlichkeit von Personen. 43. Kongress der Deutschen Gesellschaft für Psychologie vom 22.-26. September in Berlin.