Performance tests are increasingly used as alternatives to, or in connection with, standardized multiple-choice tests as a means of assessing student learning and school accountability. Besides their proposed equity advantages over multiple-choice tests in measuring student learning across groups of students, performance assessments have also been viewed as having greater utility for monitoring school progress because of their proposed closer correspondence to the curriculum that is actually taught. We examined these assumptions by comparing third-grade student performance on a performance-based writing test and a multiple-choice test of language skills. We observed smaller differences in achievement on the writing performance assessment for some groups of students (e.g., low socioeconomic status, various ethnic backgrounds) than are commonly observed on multiple-choice tests. Girls, however, had higher mean scores than boys on both types of assessments. Moreover, the school's identification and commitment over time to improving its students' writing skills positively related to its students' outcomes on the writing performance test. Overall, our examination of performance-based writing assessment is encouraging with respect to providing a relatively fair assessment and measuring learning tasks that are related to the school's curricular practices.
Post-graduate student performance in ‘supervised in-class’ vs. ‘unsupervised online’ multiple choice tests: implications for cheating and test security
