Evaluating computer-based teaching and learning situations : theoretical approaches to TILT-E's work
A methodology and a framework for evaluating computer-based learning was produced by the TLTP Teaching with Independent Learning Technologies (TILT) Project's dedicated Evaluation Group, known as TILT-E, of which the author was a member. TILT-E's evaluation work was a result of over twenty evaluation case studies conducted by the group from 1993 to 1998, the majority of which were performed during 1994 and 1995. The TILT-E studies can be viewed as examples of good practice in the evaluation of a range of computer-based teaching situations. While TILT-E's method and measures provide a useful framework for evaluating computer-based learning they did not fully exploit the strengths and history of the theoretical approaches underpinning the TILT-E work. In fact, research traditions are seldom mentioned in the TILT-E literature yet are central not only to understanding the value of the TILT-E work, but also to assisting future research in the search for a model of the computer-based teaching and learning situation. By examining the evolution of the TILT-E methodologies through three of the first TILT-E studies, referred to in this thesis as the Pilot Studies, and then assessing several later case studies much is learnt about the need for a pluralist approach to evaluation in the computer-based teaching and learning context. While TILT-E advocates the use of different methods, the group failed to justify this approach and to recommend when and why such a mix would be appropriate. This thesis aims to rectify this imbalance through the detailed examination of eight evaluation episodes covering three different computer-based teaching and learning situations, all of which had been carried out by the author. Firstly, the Fast Frac case study is considered, which involved evaluating the replacement of a lecture with the Fast Frac software. The Fast Frac study consisted of three evaluation episodes over a period of four years. The study found that the package could replace the lecture, and noted not only that a comparative design does not necessarily constrain the researcher to empirical methods alone, but also that such an approach need not disadvantage the students in either the control or the experimental groups.