Skip to content
ImperialApply
Checking sign-in…Workspace →

Writing examples: from evidence to review

Explore 39 original teaching cases across 13 programmes. Choose a question and follow the evidence, paragraph plan, full teaching draft and detailed review. Free to explore, with no sign-in or credits required.

Original fictional teaching case · Free to explore

A prototype is a starting point

Follow a developed example from evidence to paragraph planning, a full teaching draft and five review comments. The draft deliberately retains issues to discuss. These are fictional teaching cases, not real admission results; viewing them does not change your work.

1 · Evidence notes

English sample notes: separate observed work, future intentions and missing evidence. The situation is invented for teaching.

  • Computing student; academic credit-screening prototype using a teaching dataset.
  • Compared errors across represented groups and documented a sparsely represented segment.
  • Interested in junior financial data analysis, later product evaluation; no deployed lending system.
  • Preparation and fallback plan: I would look for junior work that includes data checks, error analysis and discussions with the people using a model's output. Before applying, I plan to document the prototype's evaluation in a short note that a non-technical reader can follow, including where the sample is too small. A broader data-quality role in financial services would be a useful alternative if a lending-product position is unavailable. I would judge progress by whether I can connect an error pattern to a decision and propose a useful next check, rather than by adding more algorithms to a list.
  • Progress check: I would keep a record of the checks I miss and the questions raised by non-technical colleagues, using both to guide my next learning steps.
  • Evaluation role: compared results, added group sample counts, found a small segment’s percentage sensitive to a few cases. Could explain why to retain the row but struggled with the business costs of different errors.