General Mental Ability over and above Domain-Specific Measures: Predicting Academic Achievement in a 3 Year Longitudinal Study on College Admission

<p>General mental ability (GMA) prevails as an undeniable predictor for occupational and educational achievement. Due to this predictive validity, GMA is often used in selection processes such as college admission. However, for a specific occupational or educational setting, domain-specific measures also have predictive validity. Thus, college admission processes are faced with a problem comparable to the bandwidth-fidelity problem: Is it more feasible to select students due to GMA or domain-specific measures or even both? The aim of this study was to compare the predictive validity for academic achievement of GMA and domain-specific measures as used in college admission.</p> <p>Data were collected from one cohort of 237 students in three year bachelor’s degrees at the University College of Teacher Education Styria. The grades of their 1st, 2nd, and 3rd year and their final grade on their bachelor’s thesis were collected. For college admission, all students faced a selection process comprised, among others, of GMA and domain-specific measures. GMA was assessed with five subtests of the INSBAT (Arendasy et al., 2012): verbal fluency, long-term memory, arithmetical competence, numerical and figural inductive reasoning. There were two domain-specific measures: the IPS (Schaarschmidt & Fischer, 2013), a personality test for traits relevant for the teacher profession, and a structured interview which was designed specifically for college admission in teacher education.<br>Path analyses were computed to analyze the effect of GMA and the domain-specific measures on the GPAs. GMA and the structured interview predicted the 1st year GPAs, which in turn predicted the 2nd year GPAs, which in turn predicted the 3rd year GPAs. While the 3rd year GPAs did not predict the final grades, GMA and the structured interview did. Thus, GMA and the structured interview had a mediated effect up to the 3rd year GPAs and a direct effect on the final grades. GMA and the structured interview were uncorrelated and their effects were nearly identical in size. </p> <p>The longitudinal results demonstrate the importance of GMA for academic achievement over and above domain-specific measures. The structured interview was specifically designed for college admission in teacher education, and thus it can be expected that it has predictive validity for academic achievement in this domain. However GMA, as the name states, is a general measure with no specificity for the given domain, and still retained its predictive validity for academic achievement over the course of an entire bachelor’s degree.</p> <p> </p> <p>