According to the authors stated that for psychometricians the term bias “is a factor inherent in a test that systematically prevents accurate, impartial measurement” (Cohen & Swerdlik, 2018, p. 192). Furthermore, the authors stated that the term bias suggests systematic variation (Cohen & Swerdlik, 2018). According to the text, a test is deemed bias if a portion of its variances derives from factors that are unrelated to the performance on the criterion being measured; subsequently, a group of test takers will have performed differently than another (Cohen & Swerdlik, 2018). Thus, the others stated that some tests are found to be bias based on the design research study instead of the design of the test (Cohen & Swerdlik, 2018).
Moreover, the textbook mentioned that the best way to prevent a test from being bias is during test development one should be aware of how to prevent a test from being bias (Cohen & Swerdlik, 2018). The authors mentioned that “a procedure called estimated true score transformations represent one of many available post hoc remedies” for test bias (Cohen & Swerdlik, 2018, p. 193).
According to the textbook, a test fairness is rooted in issues involving values (Cohen & Swerdlik, 2018). The book defines fairness “in a psychometric context as the extent to which a test is used in an impartial, just, and equitable way (Cohen & Swerdlik, 2018, p. 196). The text stated that for some reasonable person judgment, a test might blatantly view as being unfair (Cohen & Swerdlik, 2018, p. 196). A test may be deemed unfair due to the fact that they discriminate against a certain group of people (Cohen & Swerdlik, 2018). Nonetheless, the authors stated that “The test user strives for fairness in the way the test is used. Society strives for fairness in test use by means of legislation, judicial decisions, and administrative regulations” (Cohen & Swerdlik, 2018, p. 196).
According to the text, CTT is widely used and accepted as the model in the psychometric literature of today (Cohen & Swerdlik, 2018). Although CTT is said to be an easier model to understand and that it has its advantages and disadvantages, in my opinion I believe the IRT is preferential for responding to questions about a test’s fairness. The reason why is because as per the authors the assumptions may be characterized as being weak for the CTT; whereas an IRT method may be viewed as a strong, robust, rigorous and hard (Cohen & Swerdlik, 2018). Some believe that the IRT is a worthy successor to CTT (Cohen & Swerdlik, 2018).
Moreover, the IRT refers to a family of methods and theories which has a variety of models that are designed to handle certain assumptions concerning data characteristics (Cohen & Swerdlik, 2018). Furthermore, The IRT differs a lot from the CTT, and an IRT makes no assumption concerning the frequency of a test scores distribution (Cohen & Swerdlik, 2018). The text also stated that ” Some IRT models have very specific and stringent assumptions about the underlying distribution. In one group of IRT models developed by the Danish mathematician Georg Rasch, each item on the test is assumed to have an equivalent relationship with the construct being measured by the test” (Cohen & Swerdlik, 2018, P. 167). Lastly, the text states that the IRT model has some psychometric advantages especially for large-scale test publishers, academic and commercial test developers (Cohen & Swerdlik, 2018). According to the authors, this model is found to have an increase in as it pertains to the application of standardized test questionnaires used and professional licensing examinations (Cohen & Swerdlik, 2018).
Cohen, R. J., Swerdlik, M. (2018). Psychological Testing and Assessment. [Capella]. Retrieved from https://capella.vitalsource.com/#/books/1260303195/