The issue of fuzzy concepts in test construction and possible remedies

Abstract

Many test constructions show a lack of awareness concerning the comprehensibility of items, specifically in personality tests. However, this lack of awareness can have implications for psychometric quality. Frequently, it is simply assumed that respondents’ understanding of a test item matches the meaning implied by the test developer. It is, however, rarely ever tested whether all respondents of the targeted population actually understand the test items correctly and in a similar way. As pointed out in the last editorial (Ziegler, 2014), item content should be precisely tuned to the needs of every potential respondent in a population targeted by a newly developed test. More specifically, person variables such as age, gender, and education (Rammstedt & Kemper, 2011) have to be taken into account when constructing test items in order to assure that each potential respondent fully understands the meaning and may respond accordingly. If these person variables are neglected in the process of test construction, the psychometric quality of a test may be substantially affected by fuzzy concepts. The goal of this editorial is to raise awareness for the detrimental effects fuzzy concepts can have in test development and possible remedies. (author's abstract)