Examines effects of three writing tasks on students' writing, writing processes, and later understanding of short stories. Results indicate that personal analytic and formal analytic writing were associated with significantly higher posttest scores on literature than restricted writing in the form of short answer questions. (SRT)