Extreme Response style: A construct-independent and consistent meta-trait?

Several studies point to the fact that on measuring dispositions using multi-level response formats a distortion in the sense of a tendency towards medium (MRS) or extreme (ERS) response categories can be discovered. The present study will explore, whether these different response styles persist as an individual, consistent person variable independently from the measured contents or construct areas, respectively, of the corresponding scales. Data basis is the response behavior of N = 655 students on self-description scales of five different construct areas: Personality (Big-Five), Occupational Interests, Music-Preference, Social Dominance Orientation and Right-Wing Authoritarianism. For some of the analyzed scales the results of Rasch- and Mixed-Rasch analyses advert to different response styles (ERS vs. MRS). The consistency in the sense of a meta-trait as well as the influence of the discovered response styles upon the scale inter-correlations as well as the correlations between the constructs are being discussed.     «

Several studies point to the fact that on measuring dispositions using multi-level response formats a distortion in the sense of a tendency towards medium (MRS) or extreme (ERS) response categories can be discovered. The present study will explore, whether these different response styles persist as an individual, consistent person variable independently from the measured contents or construct areas, respectively, of the corresponding scales. Data basis is the response behavior of N = 655 studen...     »