Give the second reader a better brief

A request to ‘have a quick look’ at a decision paper leaves the most important part of the review undefined. One colleague may tidy the wording, while another may question the whole proposal. Both can be helpful, but neither can tell whether they have done what the author needed. A better review begins with a specific question: what should a second reader be able to confirm or challenge before this paper is used?

Explain what the paper is for

Give the reviewer the decision in one sentence. Is the paper asking someone to choose between options, approve a trial or agree that more information is needed? Explain who will decide and what will happen afterwards. A colleague can then judge the draft against the job it needs to do.

Consider a hypothetical library deciding whether to extend its Saturday opening hours. A paper that asks for a short trial needs enough evidence to justify testing it. A paper asking for a permanent change must answer a different question about demand and staffing over time. The reviewer needs to know which decision is on the table before commenting on whether the evidence is sufficient.

Check the work and challenge its use

The UK government’s AQuA Book distinguishes verification, which checks whether analysis was carried out correctly and meets its specification, from validation, which asks whether it suits its intended purpose. A similar distinction can help when commissioning a review of an ordinary decision paper.

In the library example, one check is whether the draft correctly describes the period covered by the visitor figures. Another is whether those figures say anything useful about demand outside existing opening hours. A perfectly transcribed count can still leave the decision-maker with an unanswered question.

Ask the reviewer to identify the assumption that carries most weight. If demand is expected to extend beyond current opening times, what supports that expectation? Could a different explanation fit the same evidence? A good challenge gives the author something concrete to investigate.

Give the reviewer enough to work with

Send the current draft and the material needed to check its main claims. Point to the page, table or section that matters. If a source cannot be shared, say what the reviewer can reasonably confirm and arrange an appropriate check where necessary.

Match the effort to the consequences. The AQuA Book recommends making the level of assurance proportionate to the risks of using the analysis and the complexity of the work. For a straightforward paper, ask a colleague to trace the main recommendation back to its evidence. For a decision with serious consequences, allow time for more scrutiny and involve someone with the relevant expertise.

Make disagreements easy to act on

Ask reviewers to explain the consequence of a concern. ‘This needs more detail’ could mean that a sentence is unclear or that a decision rests on a gap in the evidence. A useful comment identifies the passage, explains what a reader might infer and suggests what would resolve the problem.

Keep a short record of material points and the response to each. A revision may settle a concern; another may need to be stated openly as a limitation. Where the author and reviewer disagree about something that could change the decision, make that disagreement visible to the decision-maker.

Say what was checked

When the paper goes forward, explain the scope of the review. A reader should be able to tell whether a colleague checked the figures, examined the assumptions or reviewed the wording. If an important question remains open, say who will address it and when it needs an answer.

For the next paper, try a three-sentence review request. State the decision, identify the claim you most need challenged, and ask the reviewer to flag anything that would make the recommendation unsafe to rely on. That gives the second reader a clear task and makes their work useful to the person who must decide.

Sources and further reading

• The AQuA Book: UK government guidance on analytical quality assurance