Measurement and reliability of nonverbal behavior

This review synthesizes data on the reliability of nonverbal behaviors, classified according to kinesics, proxemics, haptics, vocalics, turn-taking behaviors, and global assessments, across a wide range of studies. The purpose of the review was (1) to determine which nonverbal behaviors within each classification can be measured reliably and (2) to determine which codes/classifications can be reliably measured. It was found that: (1) most nonverbal behaviors within each classification, except vocalics, can be measured reliably (.80+), (2) the codes themselves, when considered in aggregate form, can be measured reliably (.80+) when using the median as the best estimate. Several qualifications and recommendations for future research are suggested.