chapter  Seven
Scores and Statistics
ByGavin T. L. Brown
Pages 17

All tests use scores to report performance. Teachers can determine some of these scores for themselves using simple tools, but they need some familiarity with the technical tools used by test publishers to communicate essential information about test quality and student performance. This requires a brief primer on educational test statistics. This chapter addresses issues involved in determining test quality i.e., psychometrics and understanding a variety of scores used to describe performance. A challenge to score calculation occurs when tasks are rated across a multi-point scale for quality rather than simply for correctness. Accuracy is the other aspect of reliability, in that it focuses more on how truthful the observed score is. The Standard Error of Measurement (SEM) indicates the number of marks a score could vary by chance, without any substantive change in the student's ability. Criterion-referenced scores attempt to describe performance relative to objective standards for a learning outcome domain.