Three Applications of Customized Testing in Local School Districts

Abstract

Teachers in three local school districts helped customize two standardized norm-referenced tests (NRTs). The primary purpose of the investigation was to consider the effects of deleting items from the NRTs and of adding locally constructed items to the NRTs. Specifically, the normative data, percentile ranks (PRs), that would be provided for these customized tests were of interest. The results indicated that the PRs provided for customized tests may be very different from those for the complete test. This was true regardless of how the customized test was constructed, whether it consisted only of items selected from the NRT or of locally constructed items added to those selected from the NRT. Unlike previous studies, however, these results provide little evidence to suggest that the PRs associated with the customized tests were systematically higher than the PRs associated with the full-length test.