In the comprehension phase of the study (Lesch, 2008) (pre- and post-tests), participants viewed symbols paired with labels and were asked whether the labels matched the meaning of the symbol. On separate trials, the symbol was paired with correct and incorrect (but plausible) labels. A symbol was considered to be understood only if the participant accepted the correct label and rejected the incorrect label. Participants also rated their confidence in their decisions. Both training conditions benefited performance, with significantly greater benefits associated with the accident scenario training. Training which paired warning symbols with accident scenarios improved comprehension from 40 to 84% correct, reduced reaction times from 4.9 to 2.6 s, and increased confidence in responses by 27%.