This study assessed whether word processing can be taught conceptually so that students can decontextualize concepts from a specific program setting and thus achieve a greater level of learning and far transfer. The sample for this study consisted of 38 students at Central Wyoming College enrolled in word processing classes--two sections during Spring 1991 and two sections during Fall 1991. In a pilot study 18 college students were given approximately 6 weeks training in basic word processing. Students in the pilot treatment group were taught using the conceptual abstracting instructional method and were exposed to word processing with two software programs. Students in the pilot control group were only exposed to one word processing program. Both pilot groups were asked to take a transfer posttest in a new program, as well as a test of the program studied. Pilot study results led the researcher to modify the conceptual abstracting instructional method to help students guide their thinking. The study was then conducted with 38 college students as the pilot study had been. Results support the conclusion that the conceptual abstracting instructional method increases both learning and transfer by a small amount. Results also support the use of conceptual frameworks to enhance transfer of computer skills. One figure illustrates an analogy and generic framework for merge operations. (SLD)