Switch course ▾
Session 3. From Evidence to a Policy Brief
Objective. Turn evidence and local data into the first half of a brief: a sharp problem, at least three fairly compared options and a recommendation with risks.
A brief is not a smaller report. It is a tool for someone who has minutes, not hours, and who must choose between real alternatives with real limits on money and time. If your document makes that person work to find the question, the evidence or the cost, the decision will probably wait or go to whoever spoke loudest. This lesson shows how to write so the choice is easy to see.
Key ideas
Write for the person who decides
- A policy brief is a short document that helps a named person choose: a head, a deputy, the district education officer, the school board. It is not a study report in miniature. IDRC advises about two pages or under 1,500 words.
- Ask first: who reads it, what decision can they take, how much time do they have, what do they already believe? Write the decision requested in one sentence before anything else.
- Put the answer first ('bottom line up front'). Evidence and method follow for those who want to check. Write a new brief for this purpose; do not cut down an old report.
The five parts
- PROBLEM: who is affected, how many, how big, with one local number you can source. OPTIONS: at least three, including 'keep current practice'. EVIDENCE: what is known about each option and how sure we are. RECOMMENDATION: one option and why. RISKS: what could go wrong and what you will do.
- Add cost, timeline and the next step. A recommendation without a price in money and teacher minutes, and without a date, is a wish.
- The problem needs your own school data (11.5 gave you the habit): a count from a term check, an attendance record, a short survey. Anonymise it, and say how it was collected.
Compare options fairly
- Use the same criteria for every option: expected benefit (with evidence strength), cost in money, cost in teacher minutes, feasibility here, equity (who gains, who may lose), and time before results show.
- A simple 1 to 3 rating or traffic light is enough, if the reasons sit next to the colours. Do not hide a weakness of your favourite or inflate the others; the reader will find it.
- Name trade-offs openly: 'Option B gives the larger expected gain but needs 40 more teacher hours a term.' Say who bears each cost. Equity matters: ask which pupils might gain least.
Plain, honest, short
- One message per paragraph, short sentences, no jargon. Give each number a source or say 'our school count'. Round sensibly. Use one table rather than three pages of text.
- Match your verbs to the claim ladder from session 1: 'probably improves', not 'proves'. Say what you do not know, and how you would find out cheaply (a pilot).
- Test the draft on a colleague who does not know the topic: can they say the problem, the ask and the cost back to you in 30 seconds? Remove names of pupils and colleagues.
The skeleton of a brief
The reader decides in three passes
Most decision-makers read a brief in three passes. First they look at the title and the first lines to learn what you want. Then they scan the headings, tables and numbers. Only if they are still interested do they read a paragraph in full. Design for that: the ask in the first three lines, bold headings that say something, one table for the options, and the sources in a short list at the end.
Write the executive summary last, once you know what the brief really says. Then read it as the reader would, with a cup of tea and five minutes. If you still need to explain it aloud, rewrite the summary, not the speech.
Your own data is your strongest paragraph
A policy brief that quotes only other countries can feel remote. One sourced number from your own school or cluster, such as 38 of 120 Grade 3 pupils below the expected level, makes the problem real. It also keeps you honest: if the number is small, the problem may be smaller than you thought.
Collect the number the way you learned in 11.5: a clear definition, a stated date, and no names. Say how it was collected, and what it cannot tell us. Then place research evidence beside it, so the reader sees the local problem and the wider knowledge about what might help.
Examples from the classroom
The history lesson
Scene. Malika in Tashkent starts her brief with two paragraphs on how mentoring began in other countries. The district officer reaches the bottom of page one and still does not know what she wants. A colleague underlines the one sentence that matters. Malika moves it to the top and the brief shrinks from three pages to two.
Insight. Put the ask first and history last, or leave history out.
The missing option
Scene. Aziz in Bukhara writes a brief with one option: buy a set of reading books for each class. The head asks, 'And if we reorganise the school library hours instead?' Aziz has no answer. In the revision he adds that option, a no-change option, and a cost for each in money and teacher minutes. The head now picks the library hours as a first step.
Insight. A fair set of options, including no change, lets the reader decide and trusts them to do so.
The unsourced number
Scene. Dilnoza in Samarkand writes 'about half of our pupils struggle with fractions'. A board member asks where that comes from. It is her impression. She replaces it with the count from the January fractions check of her own Grade 6 classes, with the date and 'school records' next to it, and the board reads on.
Insight. Every number needs a source or a plain label such as 'our school count'.
Common pitfalls
- Starting with background and history so the ask appears on page two.
- Writing a brief with one option and presenting it as a choice.
- Hiding weaknesses of the favoured option, which the reader will find anyway.
- Using jargon or acronyms the reader would have to look up.
Options Table and Brief Skeleton
- Open the Activity Kit: three INVENTED local problems with sample data (A, B, C), the options table and the brief skeleton. Trios (5 min): choose one problem and name the decision-maker who would read the brief.
- Trios (15 min): write the Problem paragraph (3 sentences with one number and its source) and the decision requested in one sentence.
- Trios (15 min): list three options, one of them 'keep current practice'. Fill the options table with the six criteria and short reasons. Choose one recommendation and write its two biggest risks with an action for each.
- Part B, two trios (10 min): swap skeletons. The reading trio asks 'so what?' after each paragraph and marks any number without a source and any verb too strong for the evidence.
- Whole group (10 min): one trio reads its recommendation and the biggest risk aloud. The group says the problem and the ask back in one sentence. Compare with the model in the kit.
- Alone (5 min): choose the real problem for your own brief and list the two local numbers you will collect this week.
Your facilitator will guide you through this activity in the session.
Questions to discuss
- Which option did your trio add that the decision-maker might not have thought of?
- Where was it hardest to write a short, honest risk without sounding negative?
- How much of your first draft was for you, and how much for the reader?
Reflect
- Who actually decides in your context, and what do they usually ask first?
- Which option do you tend to leave out of your own proposals because it seems too cautious?
- What local number could make your problem impossible to ignore, and can you source it?
Try it tomorrow
- Write the decision requested for your real brief in one sentence and a date; read it to a colleague who does not know the topic.
- List three options for it, one of them 'keep current practice', and put a cost in teacher minutes next to each.
- Find one sourced number from your own school that shows the size of the problem.
Further reading and links
A brief opens with the decision requested, uses one local number, compares at least three options on the same criteria and names its risks. Cost in teacher minutes counts as much as cost in money.