This study examines a math journal writing assignment comparing how high school (grades 10 through 12) algebra students who performed well and students who performed poorly on traditional mathematics tests constructed their corresponding journal entries. Statistically significant differences found indicated that students who performed well on the tests were more likely to have originally composed the text and examples in their journal entries, and students who performed poorly were more likely to have copied much of the mathematical language and examples in their journal entries from their textbooks. Students who performed well on the test were also more likely to include examples accompanied by explanation for each step toward a solution. An assignment involving several such explanatory examples could perform a same or similar function as the longer journal assignment examined in this study. (Contains 4 tables.) [Master of Education Thesis, Cedarville University.]