Identifying trends concerning computer science students with dyslexia: A data mining approach

With dyslexia becoming more widely accepted as a learning disability, many questions come to light about what is being done to help those affected by it. Are those diagnosed given enough support in early education and are they being encouraged to continue with their studies at a higher level? With dyslexia being a learning disability normally specific to the use of language, it is thought that those with the disability, often excel at subjects that are less language dependent, one of these subjects being Computer Science. This paper documents the study of data collated from students on a Computer Science related degree at Salford University and University of Manchester (UMIST) with a the goal of identifying any trends, specifically focusing on those concerning students with dyslexia. To identify the trends we, in the end, decided to use Decision Trees and Association Rules and applied them to the dataset derived from the collected questionnaire answers. Statistical techniques were also applied, including graphs, to the some dataset. The statistical results we obtained met our expectations while parts did not which we thought to be interesting. The results from the mined data that we obtained gave us quite a few rules but we selected those that had a reasonably high support and/or confidence and if the rule(s) make sense of which was also a choosing factor even if the support and confidence were not that good.