Shortly after the commencement of the study, we found that some changes would have to be made in the original design. In particular, the planned provision of test results to parents no longer seemed feasible. We took the opportunity to introduce a number of new treatments into the altered design. In one of them ability testing only was carried out (Treatment 4B). In a second, achievement testing only was carried out (Treatment 5B). In both treatments results on pupil performance were given to teachers. In a third group, a criterion-referenced test in mathematics was administered to pupils and results returned to teachers (Treatment 5A). In a fourth group, teachers were provided with diagnostic information in addition to norm-referenced information (Treatment 4A). This treatment was envisaged in the original design; however, implementing it was possible only at grade levels for which machine-scorable answer sheets were used (grades 4 to 6).
KeywordsCovariance Glean Marmer
Unable to display preview. Download preview PDF.