Week 42 — revision and project scoping

A 2½-hour revision task and a one-page group project brief for 21 October.

There is no teaching on 14 October. During calendar week 42, 12–18 October, revise the design and simulation ideas covered so far and prepare one page for your group project. Allow at most 2½ hours per student, including the usual Unit 6 consolidation. No new model fitting is required.

1. Individual revision — 30 minutes

Use the Unit 6 slides selectively, your saved baseline curve and the supplied reference results. Divide this block as follows; do not add a second consolidation assignment.

  • 10 minutes — reconstruct simulation: identify the experimental unit and explain design, assumed expected response, random observation and fitted effect. Explain what changes when you repeat simulation or increase the number of pens.
  • 10 minutes — sensitivity comparison: revisit your in-class prediction. In reference/power-summary.csv, select your group and null == FALSE. Compare Baseline with Sensitivity at the same total pens. Record detection, Monte Carlo interval and unresolved rate. Does the change support your prediction, and does your recommendation need qualification? No new simulation is required.
  • 10 minutes — model map: annotate response ~ diet + room with your response, family/link, independent unit and the role of room. Bring the map and one unresolved question on 21 October.

Detailed separation, overdispersion and precision examples in the Unit 6 consolidation are optional reference material. Use a relevant excerpt if helpful, rather than completing the whole worksheet or recording in addition to this task.

2. Group scoping discussion — 40 minutes

Open the project brief and choose a provisional case. Read the case description together. Agree one biological question, one primary outcome and one comparison. Sketch the design, marking allocation or sampling, independent units, grouping and repeated measurements.

Divide the information search between group members. Identify a small number of useful sources; do not attempt a literature review.

3. Information search — 45 minutes each

Group member Search focus
A Biological context and a useful magnitude of change, with a reason and source.
B Outcome definitions, units, measurement schedule and dataset documentation.
C Practical design constraints and plausible variability for a follow-up experiment.

The linked paper and case documentation are enough to start. Separate published study details from the hypothetical teaching dataset. If a quantity is not documented, label it unknown or provisional assumption. Stop at the time limit and record the unanswered question.

4. Assemble one page — 35 minutes

Use the project-scoping template. Include:

  1. Your provisional question, outcome and main comparison.
  2. A design sketch and the independent units.
  3. A feasible dataset route, with the supplied case as fallback.
  4. A short source list and what each source contributes.
  5. A useful effect-size assumption and a variability assumption, each labelled as supported, provisional or unknown.
  6. Three unresolved questions and your next action.

Bring the brief to 21 October for discussion and feedback. It can contain uncertainties and unfinished decisions. Keep the individual revision answers separately. You do not need a fitted model, power calculation, polished report or slide presentation yet. In teaching Week 10, submit the analysis plan from your case folder before fitting; the October brief may still leave method choices open.