A comparison of the effects of traditional instruction, tutoring, and software tutorial in the Latin classroom.

Abstract

Benjamin Bloom has challenged researchers to identify an instructional methodology which will raise achievement gains to a level equaling or surpassing that produced by one-to-one tutoring. Software tutorials, because they maximize locus of control, require attention/participation, and minimize stimulus variation of instructional components, may provide a response to Bloom's challenge. If tutorials achieve this goal, they will provide a justification for computer use and a description of settings for which computer instruction is best-suited. This report describes the differences in instruction among traditional group instruction, tutoring, and software tutorials received by 80 secondary Latin students in 5 classrooms at 2 suburban schools, and measures the impact of each on (1) achievement of meta-linguistic information, (2) locus of control, and (3) attention/participation. A 2 x 3 multivariate factorial design (gender-by-treatment) including school as a nesting feature was developed and applied to results of short- and long-term achievement. Three major hypotheses were tested by analysis of covariance and a multivariate analysis of variance. Gender differences and gender-treatment interactions were explored. Results revealed achievement differences among the groups, in some cases contrary to predictions. Treatment did not account for variation among groups in locus of control or attention/participation. Variations in the pattern of time-on-task observations for the tutorial group were noted. Furthermore, (1) female achievement surpassed male achievement under all methodologies in the long term, and (2) male superiority predicted under software tutorials was not found. Software tutorials result in achievement gains surpassing or equaling those available through tutoring in the short term; in the long term, the effect of all three treatments compared is equivalent. Contradictory findings may stem from the effects of gender and school and from confounding effects in the study and software designs. Study findings fail to confirm the ability of software tutorials to respond to Bloom's challenge; however, results may be used to generate pragmatic recommendations for effective classroom use of current software tutorials.