All Scores Are Not Equal: Evaluating Score Reliability to Improve the Interpretation of Results From Tests of Intelligence in Clinical, Forensic, and School Settings

Abstract

Tests of intelligence are administered in clinical, forensic, and school settings. The score most practitioners are familiar with is the Full-Scale Intelligence Quotient (FSIQ). An individual’s FSIQ score is often used for program eligibility and is the most reliable score. In addition to the FSIQ, practitioners often evaluate an examinee’s performance on factor/composite scores. These scores may be used to evaluate an individual’s personal strengths and weaknesses but are less reliable than the FSIQ score. The least reliable scores on an intelligence test are the test’s scaled scores, which are derived from raw scores on individual subtests. It is important for clinicians and practitioners to be aware of the interpretation of scaled scores, but also of their relativly low reliability and higher variability when compared to the FSIQ and factor/composite scores. Because of their lower reliability and variability, clinicians are urged to use extreme caution when interpretating an examinee’s performance on an intelligence test at the scaled score level.