ERIC Number: ED163070
Record Type: RIE
Publication Date: 1978-Mar
Reference Count: 0
Some Methodological Issues in Using Standardized Test Scores to Evaluate Large Scale Educational Programs. CSE Report No. 111.
It is unsafe to estimate achievement test norms for fall testing by linear interpolation between norms for the previous spring testing and norms for the following spring. Second grade students were tested in reading and arithmetic in the fall and in the spring to provide pre-test and post-test data for evaluation of academic growth. Three procedures were used for computing standard scores for the fall testing. One procedure used linear interpolation without adjusting for the month in which the fall testing was administered. A second procedure used linear interpolation and also adjusted for the month of test administration. The third procedure estimated fall norms on the basis of growth curves given for first grade children on the same test. The use of linear interpolation resulted in greater apparent gains than were shown when the estimated fall norms were adjusted to match first grade growth patterns. (CTM)
Publication Type: Reports - Research
Education Level: N/A
Sponsor: National Inst. of Education (DHEW), Washington, DC.
Authoring Institution: California Univ., Los Angeles. Center for the Study of Evaluation.
Identifiers: Comprehensive Tests of Basic Skills
Note: Paper presented at the Annual Meeting of the California Educational Research Association (Marina Del Ray, California, November 17, 1977)