Research Associate in Physics Education
Applications are invited for a Research Associate (postdoc) to work on an Imperial’s Digital Innovation Fund project Module Evaluation in the age of Artificial Intelligence. The post will be based in the Department of Physics while liaising with collaborators at Imperial as well as external collaborators.Job summary
Applications are invited for a Research Associate to work on an Imperial’s Digital Innovation Fund project Module Evaluation in the age of Artificial Intelligence. The post will be based in the Department of Physics while liaising with collaborators at Imperial as well as external collaborators.
Duties and responsibilities
Evaluating modules (courses) to determine whether students are achieving the intended learning outcomes is a challenge when dealing with large class sizes. Traditionally, in Physics, research-based assessment instruments such as the force concept inventory have been used to evaluate the effectiveness of courses in achieving their learning outcomes. These require students to complete additional work and are not designed for one specific course. Furthermore, these multiple-choice assessments naturally lack some of the richness that students are capable of demonstrating. In lab courses, students often produce written work in the form of lab reports and lab notebooks. The purpose of this project is to use those artefacts to evaluate student learning by training large language models to systematically identify skills students demonstrate related to the learning outcomes of the course.
The Research Associate will take the lead on building the training data set for this project before evaluating the possibility of large language models to identify skills such as argumentation and iteration in student work. As part of the project, an instructional intervention is envisioned that we will attempt to measure the impact of using the developed artificial intelligence lab-module evaluation tool.
The role will begin with qualitative work to build the training data set and will draw on experience of teaching or demonstrating in undergraduate physical science laboratories to interpret student work when applying qualitative education methods to analyse that work. The next step is to then select, train, and evaluate large-language models to automate the analysis drawing on knowledge of supervised machine-learning methodologies. We recognise that these are a diverse set of activities and, therefore, are interested in hearing from applicants who have the experience in at least one of these areas and would take the opportunity of this role to gain expertise in the other areas. Hence, we emphasise that we welcome candidates with an interest in diversifying their skillset and will support those candidates in acquiring the needed skills to successfully complete the project.
More information and how to apply can be found on the Imperial Jobs page:
https://www.imperial.ac.uk/
