Social Intelligence and Decoding Skills in Nonverbal Communication

Abstract

The present study sought to ascertain whether ability to decode nonverbal messages can be legitimately viewed as a part of a generalized construct of “social intelligence.” Fifty-two townspeople viewed two sets of 70 black-and-white photographs. For one set of photos, subjects had to judge whether or not a mixed-sex couple shown in each photo represented a couple in a genuine relationship (as opposed to a pair of strangers posing just for the experiment). For the other set of photos, subjects had to judge which of two individuals (same- or mixed-sex) was the other's supervisor. Models of stimulus variation (used to assess internal validity) provided quite good accounts of the cues used in making these judgments and of the weights assigned to the cues. Subject variation on the tasks (used to assess external validity) was unrelated to any but one of the external measures (a cognitive one), however, and performances on the two kinds of judgments were uncorrelated. Thus, internal validation procedures succeeded, but external validation procedures did not. It was concluded from these (and other) data that claims that the measurement of decoding skills assesses an aspect of social intelligence may be at best premature, and at worst incorrect.