# Final design challenge: change one aspect of the next study

Work on your group's proposed experiment. Choose its case card or a useful peer question agreed with the instructor. These are hypothetical constraints: no new animal measurements or fitting are required.

## Write five lines

State the original question, new constraint, proposed change, consequence for uncertainty or feasibility, and evidence needed next. If the changed design answers a different question, say so. Do not promise a numerical power improvement without simulation evidence.

Discuss for five minutes, revise for eight, exchange explanations for five, hear four 90-second reports and finish with six minutes of discussion: 30 minutes total.

## A — Calves: weighing more often or enrolling more calves?

Researchers can enrol only 24 calves per milk strategy. In each pen, they would randomly assign equal numbers of calves to the two strategies and use individual feeder settings. A colleague suggests doubling weighing visits to replace calves the facility cannot enrol.

Explain what more weights on the same calf might reveal, and why they do not add calves assigned to a strategy. Could better baseline information or a narrower question help? State which assumption about differences among calves or repeated weights you would investigate first.

## B — Broilers: more birds or more pens?

Researchers propose assigning diets to whole pens within groups of pens in similar locations. The facility offers either more birds in existing pens or additional pens that could receive their own diet assignment. Stocking and costs must remain acceptable. A colleague proposes counting every bird as a separate diet assignment.

Explain what each option changes. Keep feed conversion defined as total pen feed disappearance divided by total pen gain during the same 35 days. Explain how location groups and costs affect the choice, and what you need to know about differences among pens.

## C — Pigs: shared disease exposure and missing checks

Researchers propose assigning feed to whole pens, with eight pigs per pen, then recording whether each pig develops any diarrhoea during seven days. Pigs in one pen share disease exposure. Clinical care may prevent some planned observations.

Explain why adding pigs to existing pens differs from adding pens that can receive separate feed assignments. State how you would record the number observed, the number affected and reasons for missing observations. Clinical care must follow the animals' needs regardless of feed group. Name one planning change to investigate; do not propose a student-led infection challenge.

## D — Horses: from a weather relationship to a shelter trial

Your original data describe weather and shelter use in existing pasture groups. Researchers now propose randomly assigning an additional refuge feature to whole groups, while keeping adequate shelter for every group. Fewer groups are available than hoped; a colleague suggests collecting more camera images instead.

State the new treatment comparison and group-average outcome. Explain why more images or days do not add groups assigned to refuge treatments, while they may describe each group's behaviour more accurately. Explain why shelter use alone is not a complete welfare measure. Name evidence needed before recommending the changed study.

## Alternative — A question from another group

Choose a question that could change your proposed study. Write a realistic constraint and one revision. Explain what the revised study would actually answer.