Basic Skills Testing

Test Design and Scoring

The design of any standardized achievement test ideally begins with an understanding of the purpose or purposes for the test's use. The knowledge that a test will be used to assess basic skills will inform the topics that are included in a test and the items, or questions, that are used to sample these topics. To illustrate, a test meant to assess adults' basic literacy skills will include sentence structures that are less complex than a reading comprehension test intended to inform admission decisions for graduate school.

Yet, basic skills testing may employ achievement tests whose purpose is to evaluate the span of achievement in a discipline at a given level of schooling, for example eighth-grade mathematics. When such broader-range achievement tests are used, determinations must be made about what sort of student performance or score reflects the acquisition of basic skills. Standardized achievement tests are commonly scored in two ways. Norm-referenced scoring facilitates comparisons among test takers by producing scores that fall within a normal distribution or bell-shaped curve. This scoring is frequently reported in percentile ranks. By design, therefore, norm-referenced scoring yields a distribution in which half the students are above average and half are below. There are no set procedures for deciding on a cut-off score for students' minimum competence, or other levels of achievement, within that distribution.

Criterion-referenced scoring is intended to facilitate comparisons between students' scores and a criterion of knowledge and skill within a given discipline. Criterion-referenced scoring yields reporting categories, such as pass, fail, basic, proficient, or advanced. Deciding on cut scores between criterion-referenced categories entails decision making by panels of diverse and well-trained judges in a process called standard setting.

Purposes of Basic Skills Testing

Scoring and standard setting are important because they support the interpretation of test results and, thus, how results may be used. Basic skills testing commonly serves two broad functions. It may be used formatively, that is, to shape classroom instruction to improve learning. Basic skills testing may also be used summatively, to inform decisions made at the conclusion of instruction. Summative uses often carry high-stakes consequences. For example, the scores may be used to place students in low-level or remedial classes. Researchers in the [Page 207]United States have commonly found that such classes are not conducive to improving students' academic achievement. Basic skills testing may be used to assign students to particular programs, such as special education services. Results from basic skills testing may be used to retain students in grade. For example, a policy in Chicago begun in the late 1990s required that scores on the Iowa Test of Basic Skills determine whether a child would be promoted or retained in grade. However, studies of retention show that it undermines students' long-term achievement and rates of high school graduation. Scores from basic skills testing may also be used in decisions to withhold students' high school diplomas, the absence of which is known to have long-term consequences for students' income, health, and future children.

...