Two studies investigated the effects of training on how accurately observers (college students) decoded a complex nonverbal stimulus. In the first experiment, observers viewed silent videotapes of 16 third and fourth-grade school children who were listening to an easy or a difficult lesson. Half of the children were responding spontaneously, while the other half had been instructed to pretend that they either understood or did not understand the lesson. Observers were asked to make judgments about the type of behavior they had witnessed (spontaneous or deliberate, and easy or difficult). Some observers were given feedback about their decoding accuracy after each trial, others were given feedback plus an opportunity to discuss the decision among themselves, and a third group served as the control. Results showed that all observers demonstrated a consistent practice effect, but neither feedback nor feedback plus discussion improved decoding accuracy above the level of the control group. The second study examined the effect of practice alone (10 practice trials plus six test trials). Results indicated that when compared to a no-experience control group (who just decoded the six test trials), practice had no effect on decoding accuracy. (Author/HOD)