Integration of Evaluation Processes into e-learning Environments: Developing the learning of Practical Science with the Bradford Robotic Telescope

Abstract

This thesis presents a novel framework for the collection and evaluation of data around e-learning. It shows how e-learning can play a positive role in empowering teachers in reflective practice through accessible statistical methods, as part of an evidence-based approach. Within this new framework data generated by pupils’ actions in three levels of pedagogical activity: declarative content based, functional tools based and social functional are aligned with three levels of evaluation: satisfaction, learning and behavioural changes. The framework is evaluated using the e-learning system for the Bradford Robotic Telescope. Analysis and assessment of the data by using the tools and concepts of statistics are performed. The first group of 78 clusters, mean size of 25.6 pupils, are examined to determine the effect of contextual factors. Pupils responded consistently to the same learning design across a range of contexts. National indicators for social/economic, academic achievement and group size are examined for bias. Suitable measurements for interpretation by simple Gaussian distributions are identified: satisfaction through the frequency of use with a probability of P>0.05, learning through formative assessment (P>0.3) and behavioural changes through engagement with higher order activities (P>0.2). The second group of 168 clusters, mean cluster size 25.9, demonstrates a meaningful effect size for a change in approach within the e-learning system in the areas of satisfaction, learning and behavioural changes with a probability of P<0.01.