The Preliminary SAT/National Merit Scholarship Qualifying Test (PSAT/NMSQT) is a test jointly administrated by the College Board and National Merit Scholarship Corporation, which is used to prepare high school students for the SAT, and as initial screening for numerous scholarships (Marini, Mattern, & Shaw, 2011). Previous research has acknowledged that the PSAT/NMSQT measures student's critical reading skills, quantitative reasoning, and writing skills related to academic success in college (Camara, 1998). Therefore, there is a substantial number of students who choose to retake the test in the subsequent year. The purpose of this study was to provide information about how test scores change in each component of the test (i.e. the PSAT-CR the PSAT-M, and the PSAT-W) when students retake the PSAT as well as to explore what factor influence the change in the test scores.

Two cohorts of data were provided by the College Board--- students who graduated from high school in 2008, and in 2009. For these students, up to four testing occasions were available in the data along with numerous demographic and educational variables. Latent growth modeling (LGM) was employed to examine the trends in the individual trajectories of the PSAT-CR, the PSAT-M, and the PSAT-W. A series of unconditional models, models conditional on policy covariates, and models conditional on policy and educational covariates were fit to the data. Models assuming linear growth fit well in the PSAT-CR and the PSAT-M components of the test. The fit was less adequate for the PSAT-W. Gender and ethnicity emerged as significant predictors of the score obtained in the freshman year as well as the rate of change. The strongest educational predictor was writing ability.

Results obtained from this longitudinal research are theoretically and practically informative for understanding the magnitudes of score change in the PSAT and the processes that are behind this change. Information on score change when the test is taken repeatedly may help students, teachers, test developers and test administrators to make better-informed decisions about testing students. Moreover, this information may also help parents and counselors who advise students on how to prepare for the test and whether to retake it.