Computer-Based Versus Paper-Pencil Modes of Administration United States Government End of Course Exams: Student Cumulative Grade Point Averages as Predictors of Success

Abstract

Beginning with the No Child Left Behind federal legislation, states were required to use data to monitor and improve student achievement. For high schools, the Missouri Department of Elementary and Secondary Education chose End of Course Exams (EOC) to demonstrate levels of student achievement. The policy changed from school choice of paper-pencil or computer-based testing to mandated computer-based testing. This study examined whether this decision best demonstrates the level of student mastery. Using high school EOC test scores for United States Government exams as the independent variable and high school cumulative grade point average (CGPA) as the dependent variable, the study examined the correlation between CGPA and computer-based (CBT) versus paper-pencil (PPT) modes of testing.

Random samples from two comparable school districts were used to provide data. School A tested using computers, while school B used paper-pencil testing. Data presented in this study demonstrate there is little relationship between CGPA and EOC scores depending upon the mode of test administration. For the most part, the null hypotheses were not rejected. Results indicated limited support in some subgroups for the alternative hypotheses that students with a 2.5 or higher cumulative grade point average will score higher on end of course paper-pencil tests, while students with less than a 2.5 cumulative grade point average will score higher on end of course computer-based tests.

Results of this study call into question whether the state and school districts should allow students choice of test mode or perhaps even require students to take the test using the mode of administration their cumulative grade point average indicates would demonstrate their actual level of achievement. This study also questions whether other high stakes tests such as the ACT, SAT, TOEFL, and LSAT, should determine mode of administration based on students' CGPA. Finally with the push for data driven classroom curriculum assessment, should the results of this study be applied to the need for differentiation in the classroom with regard to assessments.