Analysis plan — submit in teaching Week 10 before fitting

Pig feeding and diarrhoea

Use this sheet or the equivalent Class 10 workshop sheet; submit one initial plan. Keep it with your report and record later changes. You do not need a completed simulation at this checkpoint.

Group, case, members and date:

Your biological question: What do you want to learn about these animals or groups?

The quantity you want to estimate: Name the outcome, units, comparison or relationship, time period and animals or groups the answer concerns. If you average predictions, state which groups or conditions receive weight in that average and why.

What researchers did: Draw how researchers assigned treatments to individual animals or whole groups. For an observational study, draw how they selected farms or pasture groups instead. Explain what each row represents and which rows describe the same animal or group. Mark missing design information.

A model for the recorded outcome: What values can the outcome take? Explain why your proposed distribution can describe them. For a proportion, state the number affected and the total number observed. For a rate, explain the amount of observation time. A distribution should let you generate new example outcomes for a simulation.

What the model allows to differ: Explain the predictors, any change over time, and how the model accounts for animals sharing a pen or measurements coming from the same animal. A formula may help, but explain every term in words. State the assumptions needed to answer your question.

Checks that could change your mind: Name checks and explain their purpose. What pattern, fitting problem or unsupported assumption would make you revise the model? Do not list plots without saying what you would look for.

One plausible alternative: Describe one scientifically reasonable change to the analysis. Explain why you would compare it and how both analyses can answer the same question in the same units.

A future experiment: Describe the proposed treatment assignments, measurements, useful difference, feasible sizes and two sets of assumptions. Use earlier evidence or mark assumptions as provisional; do not select the useful difference merely because today’s estimated difference looks favourable.

Information still needed: What needs clarification, a useful source or a narrower question?

Sources and assistance: Record class examples, documentation, people or tools that helped. Say whether you saw earlier complete case solutions. Agree with the instructor how to demonstrate your own reasoning if you did.

After feedback: Record dated changes and the reason for each one.