| Curriculum-Based Measurement (CBM) is designed to be a valid, time efficient instrument used to monitor an individual's learning growth in regards to core school subjects (e.g.: math and reading; (Foegen & Deno, 2001; Deno, 2003; Fuches, 2003). Despite a huge increase in the use of CBM in schools, the psychometric properties of CBM are often limited. The purpose of this study is to illustrate a comprehensive, state-of-the-art assessment of the psychometric properties of CBM. As an illustrative example, this study examined psychometric properties of a set of four basic-math multiple skills assessments used with 283 sixth grade students in a public middle school. Students completed a different CBM assessment in their math class every 7 to 14 days during the first half of the fall semester, and then repeated the cycle of assessments in the second half of the semester. After examining test-retest and alternate form reliability, this study examined the tests' ecological validity (i.e., covariance with grades and state standardized tests) and precision (e.g., specificity and sensitivity for scoring in the lowest category in the annual state standardized test). Consistent with other studies of CBM, results found (a) marginal reliability that could be increased to acceptable levels through aggregating scores; and (b) acceptable predictive validity. However, investigation of discriminant validity and efficiency of cutting scores, a rare practice in CBM studies, revealed significant limitations of the CBM measure. The implication of these analyses is that CBM needs to be examined with more sophisticated and comprehensive psychometric methods. |
