An evaluation of presentation methods for CAL (abstract only)

Abstract

Computer-Assisted Learning (CAL) has been used as an educational tool for nearly twenty years. Its potential as a valuable supplement or replacement for traditional methods of instruction has been widely acknowledged; the realization of its potential has remained less than complete. CAL has been used with varying degrees of success in many areas of education, including public school education, university courses, distance education, vocational training in college environments, and in-house industrial training. The research described in this presentation focuses on the use of CAL in industrial training. An initial investigation into the use of CAL in industry, where course objectives are usually short term and measurable, shows that many drawbacks exist. It is felt that CAL in industry will not reach its full potential until more effective products are available. At the moment, cost is not the limiting factor so much as mere availability. Work needs to be done on software portability, innovative methods of presentation, alternative methods of presentation for various learning styles, and for individual companies the ability to customize software for their particular needs.This phase of our CAL research concentrates on the effectiveness of various presentation methods. Guidelines for effective screen design exist, but empirical work is lacking to verify all of the criteria and in particular to determine the relative importance of the many screen design considerations. Some of the areas of concern in presentation design include [1,2,3]: consistent use of functional areas on the screenuse of scrollingamount of material per screenuse of graphicsuse of colorbest placement of text, graphics, and questionsuse of highlightingsignificance of line justificationentry and exit points from a CAL packageIt is difficult to set up objective tests for these criteria because of; the subjective nature of the type of learner, age of learner, and especially the nature of the learning material. Our investigation has started by testing those parameters which we identified as being most likely to cause measurable differences in learner response. The initial broad parameter chosen was the use of scrolling, e.g. scrolling versus paging versus windowing. The challenge to be met to ensure objective evaluation is to keep the educational material consistent from test package to test package and to keep the secondary presentation parameter differences to a minimum. Even so, designing a valid evaluation procedure to analyze the results of each test vehicle is not a clear cut process. We are interested in measuring the software's ability to motivate, stimulate, and promote long term retention of material and concepts, whereas the only effect that is straightforward to measure is short term retention.Experimental software has been developed to evaluate the relative effectiveness of presenting the same educational material using the three different broad presentation methods; scrolling, paging, and windowing. The software is used by first year computer science students in the first week or two of term. We evaluate the effectiveness of the packages through two feedback mechanisms: first of all by questionnaires, and secondly by the analysis of a usage log that is produced by the software. The effectiveness of each package as far as fulfilling its educational objective is concerned is measured. In this discussion, the presentation methods used in designing the courseware, the evaluation techniques chosen, and some results and observations are examined. Preliminary work in designing “dynamic” courseware, i.e. self-monitoring and self-modifying, is discussed.