Achievement Tests

Common Forms of Achievement Tests

The most common form of achievement test that students encounter is the “teacher-made” test. These tests are devised and administered by classroom instructors, and their content is closely linked to the disciplinary knowledge and skills recently emphasized within the class. Teacher-made tests often take the form of paper-and-pencil examinations of mathematics, language arts, social studies, and science. However, because achievement testing happens across the curriculum, teacher-made tests provide a window on the variety of its possible forms. “Performance tests” are often used in music, art, or physical education classes. These require students to perform in order to demonstrate the knowledge and skills that they have acquired. For example, students may be asked to enact a routine on a balance beam or to produce a webpage. Such achievement tests are sometimes also called “authentic assessments” because they resemble performances that take place outside of school.

Teacher-made tests are often contrasted with large-scale, standardized tests of achievement. These are developed by organizations external to the school, administered to large groups of test takers dispersed across many locations, and typically scored by machine. They rely on standardized instructions, time limits, test materials, and scoring procedures. Standardization serves at least three important functions: It contributes to the fairness of an examination; it enables inferences about whether test scores reflect students' knowledge and skills, rather than unknown or unexpected events; and it facilitates comparisons of scores across disparate test takers, classrooms, schools, or other entities.

Large-scale, standardized achievement tests are often categorized according to their scoring systems. Norm-referenced tests are designed to yield scores that form a bell-shaped, or normal, distribution. The normal distribution facilitates comparisons of scores across test takers and also readily enables a percentile ranking of test takers. The College Board's SAT is one such test. However, scores from norm-referenced tests reveal the relative performance of students but are not clear indicators of students' acquisition of instructional content. In contrast, scores from criterion-referenced tests help shed light on the extent to which students have acquired disciplinary knowledge and skills. Results are often reported according to performance standards, such as basic, proficient, and advanced. At the margins between performance standards, only very minor differences in the number of correct answers may yield a categorical score shift. Thus, categorical differences between neighboring performance standards, such as basic and proficient, may not represent markedly different levels of knowledge and skill.

...