Context

Author

Jon Reades

Published

October 1, 2026

The overall assessment package is intended to test students’ comprehension of, and ability to integrate, technical skills with a broader understanding of, and reflection upon, computational approaches to urban research and spatial data science.

Assessment Elements

The assessments are grounded in a mixture of critical reflection and group work that map on to real-world data science challenges, including:

  1. Collaboratively evaluating and analysing a data set (two parts, worth 35% and 50%) as part of a small group you will be determining the suitability of a data set for tackling an analytical ‘problem’ using a mix of coding, analysis, and presentation skills in a reproducible format (due Wednesday, 16 December 2026 @ 10:00);
  2. Reflecting on the process (15%) to better-understand why a project succeeded/failed so as to improve future outcomes and recognise the contributions of individual members of the group to the success of the project (due Tuesday, 15 December 2026 @ 15:00).

Rationale

Collectively, these assessments seek to provide opportunities to ‘shine’ both individually and as part of a group. You do not need to be the best programmer in the class in order to do well on these assessments. Indeed, focussing only on the programming is likely to result in a low mark because it missed the context in which data science and data analysis ‘work’. As a budding data scientist/analyst your job is just as much to understand your audience and their needs: you will work with clients who can’t really articulate what they want or why, so good project management often involves putting yourself in your client’s shoes and working out how to translate what they say they want into what they actually need.

You will therefore do poorly on the assessments if you do not do the readings, watch the pre-recorded lectures, or participate in discussions (both online and in-person during practicals and classes). These provide you with context for the work that is being done when you start typing and running code in a Jupyter Notebook. Code is the how. Context is the why.

Use of AI

Both group assessments fall under Category 2: AI tools can be used in an assistive role. You may use generative AI tools (e.g. ChatGPT/Codex, Claude, Copilot) to support your work. You may not use them to do the work for you.

Permitted (assistive) uses include:

  • Explaining an error message, a concept, or someone else’s code.
  • Suggesting fixes to code that your group wrote, provided you understand and can explain the result.
  • Feedback on the clarity, structure, or grammar of text that your group drafted.
  • Help finding literature. You must read and check every source yourself.

Not permitted:

  • Submitting AI-generated text as your group’s analysis or argument. The briefing must be your own thinking and writing. Using an LLM as a co-author rather than a co-pilot would be considered academic misconduct.
  • Citing sources you have not read or verified. LLMs routinely invent references and DOIs. A fabricated reference will be considered academic misconduct.
  • Using AI to write or score your self- and peer-evaluation.

You must acknowledge any use. The declaration in the template requires you to state how AI tools contributed to the submission. Keep a short record (tool, purpose, what you changed) as you go. It takes a minute and makes the declaration easy to write.

Why we take this position: you will use these tools in your work, but the people who get the most out of them are those who can already code and write. That is what this module is trying to help you develop (see On Writing). Tools that write the code or the argument for you skip the part where you learn; it’s like using a forklift at the gym! They are also confidently wrong often enough that you will make serious errors in your academic and professional life if you try to rely on this short-cut. This is also why we don’t provide an AI subscription for the module.