A study analyzed the revision practices of 25 children in a third grade classroom, half of whom received LOGO instruction, to determine if the revision practices inherent in LOGO programing sponsor revision practices in the composing of autonomous texts of young children. It was hypothesized that the revision strategies in LOGO, both engendered and supported by the building of graphic models, would be a stepping stone between highly contextual revisions (drawing) and the more abstract revision required in writing words alone. The design of the study was three-tiered: a field experiment, a field study, and case studies. Revisions were analyzed according to observed changes in composing, features of the written products themselves, and the writers' thoughts about revision. The findings revealed an influence of LOGO revision practices in the paper-and-pencil composing behaviors of the students. The LOGO group increased its percentage of textual revisions by 20%, while the control group increased its percentage by only 8%. In addition, the frequency of rereading behaviors in the LOGO group increased more than 300%, while there was a 48% decline in the frequency of this behavior for the control group. Graphs and tables of data and a two-page list of references are included in the report. (HOD)