| The purpose of this study was to compare instruction via computer to traditional teacher-directed instruction in computational estimation.;A pretest of 29 items, the Assessing Computational Estimation Test, was administered to 88 seventh-grade mathematics students of middle ability who attended a junior high school in rural Indiana. The study covered an eight-week period in the spring of 1988.;The subjects, who were from five different mathematics classes with two different teachers, were divided into two groups for instruction in computational estimation strategies. To minimize the differences due to teaching style approximately the same number of students from each of the two teachers was assigned to each of the treatment groups. The computer instruction group (CI) used computer programs on computational estimation which were designed and coded by the researcher for use on an Apple II computer. Members of the traditional instruction group (TI) were taught computational estimation strategies by their mathematics teacher in the regular mathematics classroom.;Following the instruction the pretest was readministered as a posttest. Three weeks after the posttest a two-part transfer test was given to the students; the first part was timed and the second part was not timed. A questionnaire, asking the students about their use of estimation, followed the transfer test. The students' California Achievement Test (CAT) scores were also used by the researcher.;Descriptive statistics, t-tests, and Pearson correlations were used to test the hypotheses of the study at the.05 level of significance. Computer statistical analysis yielded the following results: (1) Students in the computer group did not significantly improve their scores on the ACE Test following instruction in estimation strategies, and student scores in the traditional instruction group declined from pretest to posttest. (2) Boys performed significantly better than girls on computational estimation tasks in both the CI and TI groups. (3) A signficant positive relationship existed between total mathematics scores on the CAT and scores on the estimation posttest. (4) Students did not appear to transfer estimation skills to tasks which did not specifically direct them to use estimation. |
