The effectiveness of the use of self-reported imagery strategies on children's subsequent memory performance was studied, and the coding redundancy hypothesis that memory is facilitated by using an encoding procedure in both words and images was tested. The two levels of reported memory strategy (imagize, verbalize) were crossed with "think picture,""think word" and "no induced strategy" levels. A 40-item recognition test of words and pictures was the criterion measure administered to second grade students in three Phoenix schools. The main effect, induced memory strategy was statistically significant, reflecting a higher mean score for the "think word" group than for the "no induced strategy" group. Further results are discussed, including the superiority of the "think word" group over subjects who presumably encountered stimuli in both forms, calling into question the redundancy hypothesis. Although no significant differences were associated with reported strategies' effectiveness, it was indicated that most second grade level children do report using a particular memory strategy. Further study of picture stimuli on memory strategies and strategy effects at different age levels is suggested. (CM)