Despite vigorous research on risk communication, little is known about the social forces that drive these choices. Erev, Wallsten, & Neal (1991) showed that forecasters learn to select vague or precise risk estimations as a function of what best serves the group's collective interests. This study extends the notion and further investigates whether evaluation apprehension contributes to the selection of the risk expressions. We hypothesize that group size and public feedback can engender apprehension and affect the learning of risk communication. Experiment 1 reproduced Erev et al.'s (1991) results and in addition showed that forecasters in small groups learned the optimal mode faster than those in larger groups. Experiment 2 contrasted social versus personal feedback and showed learning was faster in the personal feedback condition.