Switch course ▾
Session 1. From a School Problem to a Term Study Plan
Objective. Turn a school problem into a narrow question, choose a fair design (before and after, two classes or rotation), and write a twelve-week plan that names its threats.
Most school improvement runs on feelings: 'it seemed to work'. A term study replaces the feeling with a plan that a sceptical colleague would accept. The work is mostly done before the first lesson: a narrow question, a fair comparison and a calendar that survives Navruz and exam weeks. A day spent on design saves a term of unusable data, and it makes the later report much shorter and calmer to write.
Key ideas
A study is more than one cycle
- Course 8.10 Action Research taught one short plan-act-observe-reflect cycle in your own class. A study runs longer, adds a comparison and uses instruments others can check.
- The reason is trust. A colleague or the methodological council will act on your result only if it is clear what you compared, how you measured and what else could explain it.
- Keep it small: one question, one grade, one change, one term. A small study done carefully is worth more than a big one done loosely.
The narrow question
- Write it in five parts: WHICH pupils, WHAT change (exactly), COMPARED with what, measured by WHAT outcome, over HOW long.
- Example: 'Does a 5-minute retrieval starter on Mondays and Thursdays for 8 weeks change Grade 6 scores on a 10-item fractions test, compared with the usual opening?'
- Test it: can I do the change the same way every time, and can I measure the outcome with a short instrument? If not, narrow it again.
Three fair comparisons
- BEFORE AND AFTER: the same pupils tested before and after. Easy, but pupils also grow and other lessons help, so you cannot say what caused the change.
- TWO CLASSES: one gets the change, a parallel class carries on as usual. Check that both start at a similar level, and never give the change to the more willing class.
- ROTATION: the classes swap halfway, so both get the change and each is the comparison once. It is the fairest to pupils and strengthens the result. Decide who goes first by a coin toss.
The term timeline and its threats
- Plan twelve weeks: weeks 1 to 2 baseline test and pilot; 3 to 8 the change; 9 final test; 10 interviews; 11 analysis; 12 report. Mark Navruz, exams and holidays first.
- Name the threats before you start: groups that differ, tests that differ, extra effort because the idea is new, pupils absent for weeks, weeks lost. For each threat write one fix.
- Put the plan on one page and ask a colleague to attack it: 'What else could explain the result?' A weak spot found in week 0 costs nothing.
From a school problem to a term plan
Why a comparison matters
If you test pupils before and after a new routine and scores rise, many things could have caused it: the pupils are older, other lessons helped, the second test was easier, you tried harder. A comparison group does not remove all these, but it shows how much of the change happens anyway. That is the question a council member will ask first.
The comparison need not be dramatic. A parallel class, last year's class on the same unit with the same test, or the same class in a block where the change is switched off, are all workable. What matters is that the comparison is chosen before you see the results, and by chance or by clear rule, not because one group looks more promising.
Rotation as the fair default
In a rotation, class A gets the change in block one while B carries on as usual, then they swap. Each class is both the changed group and the comparison, so no pupil is denied the idea for a whole term, and a difference between the classes cannot explain everything: if A gains more in block one and B gains more in block two, the pattern points to the routine, not to the class.
The cost is time: two blocks, two final tests and a stable routine. Keep the blocks short (four to five weeks), use two parallel forms of the test, and decide who goes first by tossing a coin. If only one class is available, a before-and-after design with a clear note of its weakness is still honest.
Examples from the classroom
The willing class
Scene. Aziz in Fergana wants to test a weekly error-analysis task in Grade 7 mathematics. He first thinks of giving it to 7A, the class that always does homework. A colleague asks what he would conclude if 7A gained more. He sees that he could not tell the task from the class, tosses a coin, and 7B goes first; 7A follows in block two.
Insight. Choose who goes first by chance, not by promise: this removes the biggest hidden difference between groups.
Counting real weeks
Scene. Malika plans an 8-week routine in Grade 5 reading in Tashkent. On the calendar she finds a holiday week, an exam week and two school events, so only 6 teaching weeks remain before the end of term. She shortens the change to 6 weeks, books the final test in week 8 and keeps one spare week.
Insight. A plan is only as good as its calendar: count teaching weeks, and keep a spare.
One change at a time
Scene. Dilnoza in Samarkand wants to improve Grade 6 science reports. She plans a report frame, group discussion and a new marking scheme at once. Her mentor asks which would explain a better result. She keeps only the frame for this term and puts the others in a list for later studies.
Insight. Change one thing at a time, otherwise a good result cannot be traced to a cause.
Common pitfalls
- Choosing the more willing or stronger class to try the change.
- Writing a question so broad ('improve mathematics') that no short test can answer it.
- Planning twelve weeks on paper without subtracting holidays, exams and school events.
- Changing the routine or the test halfway through because it seems better, which breaks the comparison.
Design the Term Study
- Open the Activity Kit: six SAMPLE school situations (A to F), the design menu and the one-page plan form. Form trios, ideally from different subjects or two parallel classes. (5 min)
- Trios (15 min): choose one situation and write its narrow question in the five parts. Test it: same change every time? short instrument? Cut until both answers are yes.
- Trios (15 min): pick a design from the menu and draw the twelve-week timeline with calendar losses. Name three threats and one fix for each.
- Two trios (15 min): swap plans. The reading trio asks the challenge questions from the Kit and writes the single biggest weakness they see. The owners revise.
- Second short activity, pairs (10 min): take your OWN real worry. Rate it against the five 'small enough' checks in the Kit. Cut it until all five are yes.
- Alone (5 min): write your school-based task from the Kit and name the colleague who will be your second pair of eyes.
Your facilitator will guide you through this activity in the session.
Questions to discuss
- Which change was hardest to describe so that you could do it the same way every time?
- Which threat did the other trio find that you had missed, and what is the fix?
- In your own school, which design is realistic: two classes, a rotation, or only before and after? Why?
Reflect
- Which worry in your teaching would be worth a whole term of careful study?
- Is there a parallel class or a colleague you could compare with, fairly?
- What would convince you that the change did not help?
Try it tomorrow
- Write your narrow question in the five parts and read it aloud to a colleague.
- Draw your twelve-week calendar and cross out every lost week before you place the tests.
- Toss a coin to decide which of two parallel classes goes first, and write down the result.
Further reading and links
One narrow question, one change, a fair comparison, a twelve-week plan and the threats named before you start.