The Application of G-Theory on MSRT (MCHE) Score Dependability; Variability Due to Persons, Gender, Subject Fields, Sections, and Items

The purpose of this study was to determine the relative contributions to MSRT (MCHE) score dependability (that is analogous to classical theory reliability) of persons, gender, subject fields, sections, and items as well as the interactions among them. The study sampled 0011 test takers, 011 female and 011 male participants in the MSRT (MCHE) Test from theAugust, 4102 administration. The test was administered under normal operational conditionsand included three sections: (0) Listening Comprehension, (4) Structure and Written Expression, and (3) Reading Comprehension. The analyses included descriptive statistics, classical theory reliability estimates, and a series of generalizability studies conducted toisolate the variance components due to persons, gender, items, sections and subject fields, and their effects on the dependability of the test. Classical reliability analyses found theMSRT (MCHE) test scores of the examinees in question to be highly reliable. The items within a section varied greatly in difficulty. An examinee that scored relatively high in one section was likely to score high in the other section. Gender and its interactions with itemsand sections were not very important relative to all the other sources of variance. In fact, the participants’ proficiency differed considerably across items and somehow across sections, butnot much across gender. The variance component due to subject fields was relatively smalland all of the participants (regardless of their subject fields) performed about the same. These results could be useful to test developers and researchers interested in the relative effects of such factors on test design

Keywords:

CTS , G Theory , Dependability , G-Study , D-Study