Observer bias in the hearing testing of profoundly involved multiply handicapped children.

Observer bias was investigated using 27 profoundly involved multiply handicapped children and behavioral observation audiometry. Two observers were in the room with the child, two others viewed via a video monitor, and each of the four observers had full knowledge of stimulus-type and intensity. Two additional observers viewed a videotape without being provided information about the stimuli. Results showed clear bias effects in 85% of the children. At low intensities, observers aware of stimulus events tended to score fewer responses than those judges unaware of stimulus intensity. At high levels, this trend was reversed. It is recommended that the estimation of hearing threshold in handicapped children be made by observers who do not have access to the sequence of stimulus presentations.