Russell and Haney (1997) reported that open-ended test items administered on paper may underestimate the achievement of students accustomed to writing on computers. This study builds on Russell and Haney's work by examining the effect of taking open-ended tests on computers and on paper for students with different levels of computer skill. Using items from the Massachusetts Comprehensive Assessment System (MCAS) and the National Assessment of Educational Progress (NAEP), this study focuses on

more »

... nguage arts, science and math tests administered to eighth grade students. In addition, information on students' prior computer use and keyboarding speed was collected. Unlike the previous study that found large effects for open-ended writing and science items, this study reports mixed results. For the science test, performance on computers had a positive group effect. For the two language arts tests, an overall group effect was not found. However, for students whose keyboarding speed is at least 0.5 or one-half of a standard deviation above the mean, performing the language arts test on computer had a moderate positive effect. Conversely, for students whose keyboarding speed was 0.5 standard deviations below the mean, performing the tests on computer had a substantial negative effect. For the math test, performing the test on computer had an overall negative effect, but this effect became less pronounced as keyboarding speed increased. Implications are discussed in terms of testing policies and future research.