User: Guest  Login
Title:

Analysis of Extreme Response Style in Vocational Interests

Document type:
Konferenzbeitrag
Author(s):
Heine, Jörg-Henrik; Tarnai, Christian; Hartmann, Florian G.
Abstract:
Individual response styles in polytomous response scales are a frequently observed phenomenon. One type response style is a distortion in terms of a tendency towards middle (MRS) or extreme (ERS) response categories. There is evidence from research that this distortion is independent of the content of the measured construct and the offered items. In the present study, the six scales of the General Interest Structure Test (AIST-R, Bergmann & Eder, 2005), which is based on the hexagonal model (RIASEC) of Holland (1997), are tested for their scalability using Rasch and mixed-Rasch models. It is examined, whether for the six sub-scales different, content-independent response sets (ERS vs. MRS) could be found. The calculations were performed with the R-package “mixRasch” (Willse, 2009) using jML estimation and the windows program “WinMira” (Davier, 2001) using cML estimation. Data basis for this study is a sample of University students (N = 734). The findings of this study suggest that three of the six scales are Rasch-scalable, in terms of item and person homogeneity. A two-class solution has proven best for the other three sub-scales. In each of the two classes different item category difficulties are observed, which can be considered as an indication of different response styles, in terms of a tendency to medium (MRS) or extreme (ERS) response categories. Both methods for model estimation (jML vs. cML) returned similar results. The findings are discussed in the context of the hexagonal model of interests (Holland, 1997). References: Bergmann, C. & Eder, F. (2005). AIST-R - Allgemeiner Interessen-Struktur-Test mit Umwelt-Struktur-Test (UST-R). Göttingen: Beltz. Davier, M. v. (2001). Winmira 2001. Groningen, The Netherlands: ASC-Assessment Systems Corporation USA and Science Plus Group. Holland, J. L. (1997). Making vocational choices. A theory of vocational personalities and work environments (3rd ed.). Lutz, FL: Psychological Assessment Resources. Willse, J. T. (2009). mixRasch: Mixture Rasch Models with JMLE (Version 0.1). Retrieved from http://CRAN.R-project.org/package 1⁄4 mixRasch.
Book / Congress title:
11th European Conference on Psychological Assessment (ECPA)
Congress (additional information):
31. 08 - 03. 09. 2011
Publisher address:
Riga, Latvia
Year:
2011
Month:
August
Language:
en
 BibTeX