A study designed to develop a selection model for the prediction of Signalman performance in sending and receiving Morse code and to evaluate training strategies was conducted with 180 Navy and Coast Guard enlisted men. Trainees were taught to send Morse code using innovative training materials (mnemonics and guided practice). High and average aptitude groups were compared on code learning and performance factors in relationship to traditional vs. innovative materials. The Armed Services Aptitude Battery (ASVAB) and nonverbal tests (Visual Pattern Discrimination--VPD--and Visual Reaction Time--VRT) were used to predict performance. Mnemonics and guided practice proved superior to traditional training materials for learning Morse code, but the type of training materials had no significant effect on ability to send and receive messages. The ASVAB tests were good predictors for high aptitude individuals, while the VPD proved the best predictor for the lowest quartile. As a result of the study, the following training model was proposed: (1) VPD as well as other tests for screening low and average aptitude applicants; (2) self-paced instruction; (3) VPD exercise used as orientation for code reception; (4) examination of validity of course entry requirements. (Author/KC)