Artefact: Reflective Piece

Reflection on Research Methods and Professional Practice Module

Introduction

My experience of the Research Methods and Professional Practice (RMPP) module has been an excellent opportunity to develop my critical and research skills. Reflecting on the many artefacts generated during the module, the stand-out experiences were the literature review and research proposal presentation. I will also discuss my skill development and evidence of learning. I will use the reflective method of Rolfe et al. as a framework (Rolfe et al., 2001).

Outcomes and Key Moments (‘What?’)

In my "Medical Implant Risk Analysis" case study, in CLDF 1, I used critical thinking skills to map between the ethical principles of the British Computer Society (BCS) and the Association for Computing Machinery (ACM) (British Computer Society, ND; Association for Computing Machinery, ND, 2018). My summary post included valuable feedback from peers, and I highlighted ethical design and patient trust as key factors.

Building on learning around AI ethics, the unit 1 reflective activity covered the link between ethical principles and AI regulation, as discussed by Corrêa et al. (2023), and included a strategy for AI ethics education (Deckard, 2023). I have included in my personal development ‘Action Plan’ (see e-portfolio) an action to learn more about AI ethics and in particular the developing AI regulation as I believe this will be critical to the field in the future.

The literature review was the first summative component. Initially, this was daunting due to the wide-open choice of topic, and the breadth of the literature, but a series of guiding questions were useful, such as who the audience is, and what is the specific focus. A personal interest is chest X-ray (CXR) deep learning (DL) systems, so I chose this, hoping to use the knowledge I gain again in the future. A key step was developing the research questions and conducting the search. A key learning for me was the value of submission of an outline to a supervisor for early review, thereby shaping the final piece of work. Feedback gained from the course tutor was very helpful in my choice to narrow the scope of the work.

The final major task was the research proposal presentation. Initially, I sought to jump directly from the gap identified by the literature review and seek to fill that gap with a research proposal. From my previous academic work, this is the logical approach. However, based on invaluable feedback from the course tutor, I stuck to the same topic (CXR DL) but pivoted to a totally new angle: improving model performance through experimentation with multimodal models.

In the later units and seminars, I completed statistical exercises. These were relatively simple analyses, but it is always useful to refresh on the fundamentals of summary statistics, inference and hypothesis testing, and making charts.

Analysis and Interpretation (‘So What?’)

The "Medical Implant Risk Analysis" case study (CLDF 1) was a good example of improving the strength and depth of academic analysis based on feedback from peers. In this case, my initial one-dimensional analysis was improved by additional dimensions including legal issues and risk disclosure. This led me to see the topic in a different way.

The literature review was a useful exercise in academic skills and critical analysis. Having developed my search query and searched a literature database, there were many articles. I was pleased with the final feedback on my work, noting the criticism of structure and adherence to presentation guidelines. I was surprised by criticism of reviewing too many articles and not focusing on a comparison of a small number, as my conception of the goal of a literature review is to survey the literature and identify gaps, which would not be met unless the review is complete. However, I do see that if the goal is instead to showcase a deep understanding of the literature, then focus on a few articles would be preferable.

The key moment in the research proposal task was my pivot from focusing on real-world performance to my final proposal of experimenting with multimodal data integration. I think this pivot demonstrated my adaptability and responsiveness to feedback. It improved the learning experience as the new angle built on my growing domain knowledge in CXR DL systems but led to totally new research around multimodality. Combining the two learnings of current CXR DL systems and a future technical direction improved the overall experience.

Completing the statistical exercises, it was interesting to use excel to do inferential statistics and hypothesis testing, as I have for a long time used R for statistical programming. It did confirm to me the advantages of R/Python over excel, being that the storage of the data and analytic code are not mixed, the arbitrary choice of cells spatially in excel, and the limitations of customisation of charts, and downstream/output formats of statistical results.

To the future (‘Now What?’)

For future written projects, I will be keen to submit an outline to supervisors in good time to incorporate feedback in the final piece. This is a great opportunity to improve the direction of the work before starting on the deep research. It is clear to me that just as much time and effort should be spent on developing the scope and direction of the research question, as to undertaking the research itself. Being able to pivot based on feedback on a project is a key skill that I am pleased to be developing.

Furthermore, based on the feedback from my summative, and teaching on the course, I have gained an appreciation of the importance of careful and considered citation of others’ work and ideas. It is critical in an academic setting to highlight the novel contribution of your work.

Skills and Personal Development

In addition to my SWOT analysis and skill matrix, available in my e-portfolio, I will consider my skill development. Deckard (2023) argues for the importance of inter-disciplinary skills, and I would argue that my work on this module is evidence of such skills: from artefacts involving reflection on AI ethics, to technical research proposals on multimodal AI models.

Conclusion

The RMPP module has been an excellent learning experience, and I will benefit from the academic skills developed, and my personal development actions generated, in my future professional and academic career.

References

Association for Computing Machinery (2018) Case: Medical Implant Risk Analysis - ACM Ethics Available from: https://ethics.acm.org/code-of-ethics/using-the-code/case-medical-implant-risk-analysis/ [Accessed 6 May 2024].

Association for Computing Machinery (no date) The Code affirms an obligation of computing professionals to use their skills for the benefit of society. Available from: https://www.acm.org/code-of-ethics [Accessed 6 May 2024].

British Computer Society (no date) BCS Code of Conduct for members - Ethics for IT professionals | BCS. Available from: https://www.bcs.org/membership-and-registrations/become-a-member/bcs-code-of-conduct [Accessed 6 May 2024].

Corrêa, N.K., Galvão, C., Santos, J.W., Del Pino, C., Pinto, E.P., Barbosa, C., Massmann, D., Mambrini, R., Galvão, L., Terem, E. & de Oliveira, N. (2023) Worldwide AI ethics: A review of 200 guidelines and recommendations for AI governance. Patterns 4 (10): 100857.

Deckard, R. (2023) What are ethics in AI? | BCS. Available from: https://www.bcs.org/articles-opinion-and-research/what-are-ethics-in-ai/ [Accessed 10 July 2024].

Rolfe, G., Freshwater, D. & Jasper, M. (2001) Critical Reflection for Nursing and the Helping Professions: A User’s Guide Palgrave MacMillan.