Examining the Validity of ACT Composite Score and High School Grade Point Average for Predicting First-Year College GPA of Special-Tested Students

ACT Research Report Series 2016 (7)Examining the Validity of ACT® Composite Score and High School Grade Point Average for Predicting First-Year College GPA of Special-Tested Students NooRee Huh, PhDChi-Yu Huang, PhD © 2016 by ACT, Inc. All rights reserved. ACT® is a registered trademark of ACT, Inc. 5931NooRee Huh is a senior psychometrician in Measurement Research at ACT specializing in research and analysis in educational psychometrics, data forensics, and accommodations. Chi-Yu Huang is a principal psychometrician in Measurement Research at ACT specializing in research and analysis in educational psychometrics, data forensics, and accommodations. AcknowledgmentsThe authors extend a special thanks to Deborah J. Harris for suggesting we undertake this research. We also wish to thank Xiaohong Gao, Justine Radunzel, and Richard Sawyer for their constructive comments and suggestions. ContentsAbstract .................................................................................. ivIntroduction .............................................................................. 1Data and Analyses ....................................................................... 3Results ................................................................................... 4Analyses by Extended Time and Diagnosis, Jointly ....................................... 5Summary and Discussion ............................................................... 5References .............................................................................. 7Appendix ................................................................................ 9 AbstractThe current study is a replication of the study by Ziomek and Andrews (1996), which, for both regular-tested and special-tested students, examined the predictive validity of using a student’s Composite score on the ACT® test and high school grade point average (GPA) jointly to predict the student’s first-year college GPA. In addition to the joint prediction model, the current study examines two other prediction models for comparison purposes: a model using high school GPA alone, and a model using ACT Composite scores alone.Two diagnosis subgroups (attention deficit and reading disability) and two extended time subgroups (up to time-and-a-half time and up to triple-time) of special-tested students had a sufficient sample size to be examined in the current analysis. The results for these subgroups were compared to those for regular-tested students enrolled in the same postsecondary institutions. The results indicate that college GPA predictions jointly based on ACT Composite score and high school GPA are more accurate than predictions based on either variable alone. In contrast to using both predictors, using only one of the predictors (high school GPA or ACT Composite score) tends to result in bigger over-prediction of first-year college GPA for the special-tested subgroups. In a future study, the validity of the prediction models could be examined for several other accommodations provided to students, if sufficient sample sizes for different accommodations groups are available. Moreover, additional research is needed to evaluate how the number and types of services received in college by students with disabilities would affect study results. 1IntroductionThe state-led effort to develop the Common Core State Standards was launched in 2009. During the development process, the standards were divided into two categories: the college and career readiness standards and the K–12 standards (www.corestandards.org/about-the-standards/development-process/). The college and career readiness standards address what students are expected to know and understand by the time they graduate high school. From an academic perspective, college and career readiness means that high school graduates have the English and mathematics knowledge and skills needed to qualify for and succeed in the postsecondary job training and/or postsecondary education necessary for their chosen careers (www.achieve.org/college-and-career-readiness). Measuring whether students acquired requisite skills and knowledge in their high school years provides data that helps set students on a successful career path. The ACT test is designed to measure the skills and knowledge important for success in postsecondary education that are acquired in secondary education. ACT provides each state with a report that details the college readiness of students who took the ACT (ACT, 2007; ACT, 2013). According to Clinedinst (2015), admission test scores such as ACT scores and high school GPA are reported as two of the most important factors to evaluate student entrance applications by many college and university admissions officers. Several studies have documented a positive correlation between ACT scores and/or high school GPA and students’ success in the first year of college (Noble & Sawyer, 2002; Noble & Sawyer, 2004; Allen & Sconing, 2005; Allen & Robbins