Fred IngramPhD student
Most of my career has been in the IT industry, with over thirty year’s experience in Business Analysis, Project Management, Systems Analysis & Design, Data Warehousing, Programming, etc. However, the last 15 years were spent as a teacher of Mathematics and ICT at a secondary school in the UK.
I am currently a full-time, fully-funded PhD student endeavouring to build an agent-based simulation of classroom interactions between students and teacher (and teaching assistant) in order to investigate the long-term consequences of common lesson scenarios and interventions.
The PhD project is to build an agent-based simulation of classroom interactions between students and teacher (and teaching assistant) in order to investigate the long-term consequences of common lesson scenarios and interventions.
The project has three components. The first is a custom-built, highly efficient data collection program which so far has been used to collect approximately 40 hours' worth of lessons, comprising approximately 20,000 classroom events which cover 17 different student states/activities and 12 for the teacher. This means that I now have quantitative measures (distributions) of how long students and teachers spend on different activities (e.g. working alone, working with others, listening to the teacher, chatting, being told off, being helped individually, helping others). The intention is to combine this data with the student survey data and school records also collected. The university Ethics Committee has retrospectively approved this data. The aim is to produce student profiles. The plan is to collect even more data.
The second component of the project is an animated visualization of the collected data. Using NetLogo (a popular, free, open-source agent-based simulation development environment), a program has been developed to 'play back' the collected event data to give a visual overview of what happened in a lesson. For each lesson, a room layout is created with the students positioned at their desks and the teacher and teaching assistant moving around. From the playback (with adjustable speed) one can clearly recognize patterns of behaviour. The program also produces graphs and statistics (e.g. a state transition matrix for each student) all aimed at giving insights into student and teacher behaviour.
The third component is the agent-based simulation itself. The development of the student and teacher attributes and rules is the primary focus of the project. The current agent rules embody sensible behaviour predictions but need further refinement. The agent behaviour needs to be calibrated according to the real data collected. One of the many features of the simulation is the ability to run a visualization, pause, alter something (e.g. a teacher's response to a student or change the seating arrangements) and now simulate the consequences. The results of the simulation can then be compared to the actual outcomes. This part of the project is where the research is most intense.
BSc (Hons) Computer Science, MSc Information Systems, Cert. Ed. Post Compulsory, MA Education, BSc (Hons) Mathematics