The effects of using a word processor containing grammar and spell checkers on the composition writing of sixth graders
Improvement of written communication skill is a major goal of many American educators today. This quasi-experimental study, conducted in six sixth-grade reading classes, evaluated and compared the effectiveness of three approaches to improving writing in the reading classroom and examined the students' attitudes toward writing. Independent variables were method of treatment, language level of students, and home language of students. Dependent variables were writing achievement, scored holistically and analytically, and attitude toward writing. All classes were taught by the same teacher, received the same instruction, and completed two writing assignments each week for six weeks. The six classes were divided into three treatment groups: WP+ (using a spell and grammar checker-enhanced word processing software program), WP (using the same software program without the checker options), and P&P (using pen and paper). Prior to implementation of the study, each student completed an attitude survey and wrote a descriptive paragraph that was used as a pretest. After six weeks of treatment, students wrote two posttest essays-one with the treatment used in the study, and the other handwritten-and completed the attitude survey. Interviews with students explored their attitudes, with emphasis on whether students preferred writing with paper or computer, and why. Student comments are included in the Appendix. All student writing was evaluated after the conclusion of the study. Handwritten pretests and posttests were typed into the computer, reproducing all student work exactly, to assure that all papers would be treated equally. Results of the writing and the attitude surveys were analyzed by analysis of covariance, using pretest writing scores as covariate (to compensate for any differences resulting from the need to use intact classes). Five null hypotheses were investigated; two were rejected and three were retained. Significant differences attributable to treatment were determined between students using enhanced word processors and those using pen and paper, and English home language students and those from bilingual home backgrounds. Areas where there was no indication of treatment effects were writing achievement of low-language ability students and those who were not low, and student attitudes toward writing.