Video analysis of pre-verbal communication behaviours: use and reliability

Abstract

A technique for assessing pre-verbal communication behaviours in young children with hearing impairment has previously been developed, based on rigorous analysis of video recordings of the child interacting with a known adult in a controlled setting. The technique has been found to be valuable for monitoring progress of children with hearing aids, and more recently with cochlear implants. The originators have shown the resultant measures to have important predictive validity, and the technique is gaining widespread acceptance. This study examined the reliability of the method across different observers. Twelve video samples were selected to represent the range of ages and stages of auditory development encountered in the implant programme. Three observers analysed the samples independently, according to the established protocol. There was a very high level of accord amongst observers, with 498 agreements and only 12 differences in all possible comparisons between the three observers on the various items scored. Statistical analysis revealed that the correlation coefficients between the scores of the observers ranged from 0.86 to 1, all being statistically significant (p<0.01). It is concluded that the video analysis technique in young children has a very high reliability among different observers. This finding supports the development of training methods for more widespread use of the technique in assessing pre-verbal communication behaviours in young deaf children with cochlear implants or hearing aids.