Abstract:
Computer-aided instruction is the focus of much contemporary research; this thesis is a contribution to that research in the area of composition. After examining previous scholarship in the areas of sentence-combining as a means for improving composition, different methods for teaching relative clauses, and finally, CAL in general, the researcher created a five-exercise computer program on sentence-combining with relative clauses. With this program, she could then examine the feasibility of CAl from the vantage points of programmer, instructor, and student. She selected 86 subjects from freshman composition classes by means of a diagnostic pre-test and divided them into three groups; CAL, mini course, and control. She kept a log of programming experiences and a log of instructional observation during the study. To obtain statistical data, she classified the exercise, post-test, and post-post-test scores by sex, class level, and type of instruction. the researcher discovered from the programming phase that, with a patient teacher, one can learn to use a computer in a short period of time. From the instructional observations, she found that most students were enthusiastic about CAT and that it was n good aid for teachers. Finally, from the statistical data, she discovered that in both the mini-course and CAl groups, female scores were higher than male scores, EN 08 students did better than the other two class levels, and the mini-course group did better overall than the CAl group. However, because the difference in test scores was a matter of only one or two points with quite a few of them not statistically significant and because the CAl group's main deterrent may have been that they were not accustomed to reproducing answers exactly, perhaps none of the variables is as important as giving the students more exposure to CAl to determine if they are suitable candidates in the first place.
