| Summary: | Most medical interaction studies have been conducted on audiotaped recordings of physician-patient encounters. Empirical studies have not previously demonstrated whether coders' scores differ on audio-only versus videotaped data. Data from a convenience sample of forty-seven physician-patient interactions were analyzed using the same coding systems to judge audio-only versus video-based data formats. All coding conditions demonstrated acceptable reliability, using intraclass correlation coefficients. However, MANOVA analyses show that ratings of audiotaped physician-patient interactions are not equivalent to ratings of videotaped encounters. Exploratory factor analyses show differences in the underlying structures of the data derived from the audio-only versus the video information. The differences in the video-based factor solutions account for more total variance and are more consistent with theoretical expectations. Reprinted by permission of the publisher.
|
|---|