The project proposal

n0: A promising trial

A small team tested a new booking form with twelve volunteers. Ten completed it without help. The project lead wants a recommendation for the next meeting. Jo must decide how to frame these encouraging but limited results.

• Draft an evidence summary first → n1
• Ask what decision the meeting must make → n2

n1: What the figures support

Jo records the trial size, the task and the number of independent completions. The volunteers already knew the service. Their experience may differ from that of first-time users.

• State that limitation beside the result → n3
• Seek feedback from the two who needed help → n4

n2: The decision boundary

The lead says the meeting will decide whether to fund a larger trial, not whether to replace every form immediately. This changes the level of certainty Jo needs to claim.

• Recommend a larger bounded trial → n5
• Compare the costs of two trial designs → n6

n3: A useful qualification

Jo writes, "Ten of twelve familiar users completed the task independently; a broader trial is needed." A colleague worries that this sounds less impressive than "Most people find it easy."

• Keep the precise statement → n7
• Explain why the wider claim is unsupported → n8

n4: The unresolved steps

Both participants who needed help struggled to find the final confirmation. Their comments suggest a possible design issue, but two reports alone cannot establish its frequency.

• Propose a focused confirmation-step test → n8
• Separate that observation from the success count → n9

n5: A bounded recommendation

Jo proposes a four-week trial with first-time users and a defined booking task. The proposal includes a transcript of instructions and a way to record where help is needed.

• Add a decision point after the trial → n9
• Invite the team to challenge the assumptions → n10

n6: Comparing designs

One trial design recruits a broader group; another observes fewer people in more detail. Each answers a different question. Jo lists the trade-off instead of calling one universally best.

• Choose according to the meeting’s question → n10
• Present both with the uncertainty stated → n7

n7: A decision with limits

The team approves a larger trial and a review date. Jo’s summary records the observed result, the unfamiliar-user limitation and what the next trial will test. No universal success claim is made.

Why this ending? The recommendation matched the evidence and the actual decision. Qualifying a claim made the proposal more usable, not less useful.

n8: An assumption made visible

A colleague points out that mobile users were missing from the first trial. Jo adds that gap to the plan and revises recruitment before the larger trial starts.

Why this ending? Inviting criticism exposed a testable assumption. The discussion improved the plan without pretending the original data answered every question.

n9: A focused investigation

The team tests the confirmation step with unfamiliar users and records when support is needed. It keeps the original twelve-person result separate from the new observations.

Why this ending? Jo turned a possible explanation into a question for testing. Combining unlike observations without context would have hidden what each result meant.

n10: An explicit trade-off

The team chooses a smaller observational study first because its immediate question is where people become stuck. The broader completion-rate trial remains a later decision.

Why this ending? The chosen method serves the immediate question. This ending is a different reasoned path, not a failure for choosing a smaller study.