Test performance of language-handicapped children with familiar and unfamiliar examiners

The importance of examiner familiarity to children's performance on tasks requiring high or low levels of symbolic mediation was investigated. Thirty-four handicapped preschool children were examined within a repeated measures crossover design, once by one of two familiar classroom teachers and once by one of four strange teachers. Subjects performed significantly better with familiar than with unfamiliar examiners on high symbolic mediation tasks; no such differential performance was obtained on low symbolic mediation items. Findings are related to current efforts to identify procedural and situational variables in assessment, uncontrolled by present standard test administrations, that may preclude children's optimal performance.