Start with the key idea
Agree a specific learning goal, keep a short starting task and compare it with a later fresh task under similar conditions. Record the help used, the reasoning and whether the idea transfers to a new example. Use this evidence alongside the child’s experience and school feedback. One score or one enjoyable lesson cannot establish the whole effect of tutoring.
Start with one observable goal
Replace “get better at Maths” with a task the child can demonstrate. For example: “Read a two-step word problem, choose the operations and explain the answer.” In English, the goal might be “write a paragraph with a clear main idea and supporting details”. In Science, it might be “explain a circuit calculation with the correct quantities and units”.
Agree what a successful response looks like with the tutor. Include method, explanation and level of help, not only a final mark. Keep the goal narrow enough to inspect in a short sample but meaningful enough to matter in schoolwork.
Ask why the goal was chosen. A useful answer connects it to the child’s actual work or a prerequisite difficulty. A chapter label can describe content covered; it does not identify the precise change you are looking for.
Set a review date together rather than demanding a universal number of sessions. Starting knowledge, attendance, topic difficulty and practice conditions vary. The review date is a decision point, not a guaranteed deadline for mastery.
Keep a small starting sample
Before a teaching sequence, keep one or two suitable tasks and record what support was given. A photo or saved page can be enough for your private record. Do not post identifiable student work publicly as part of this process.
Record whether the instructions were read aloud, a formula was supplied, a diagram was started or a step was prompted. These are different forms of help. They may be appropriate support, but they change what an independent performance claim means.
Keep the learner’s first response as well as the correction. If only the polished answer remains, you cannot see which step needed teaching. Avoid creating a high-pressure test just to make a baseline; use ordinary appropriate work agreed with the tutor.
The starting sample is evidence about those tasks under those conditions. It is not a diagnosis, an intelligence score or a fixed label for the child.
Compare like with like, then test a new application
At review time, use a fresh task with the same underlying skill and a similar level of difficulty. Repeating the identical question may show memory for that answer. A substantially harder question may make genuine progress difficult to see.
First compare under similar support conditions. Then, where appropriate, see whether less support is needed. Record both results rather than combining them into an unexplained percentage. A learner who now draws their own diagram may have improved even if the final calculation still needs work.
Ask for the reasoning. In a word problem, correct arithmetic applied to the wrong operation does not meet the modelling goal. In reading, a plausible answer without evidence may not meet an inference goal.
Finally change the context or representation. The child might move from counters to a short story, from a worked sentence to their own paragraph, or from a numerical formula to a graph. This checks transfer; it should stay within what has reasonably been taught.
What useful evidence looks like in three subjects
Maths: a learner previously divided every number they saw. In the later task, they explain that 27 children in cars with four seats need seven cars, even though 27 ÷ 4 is 6 remainder 3. The evidence is the interpretation of the remainder, not only the division.
English: a learner previously listed three unrelated sentences. In the later paragraph, each supporting sentence develops the same main idea. Keep both drafts and ask the child to point to one edit and explain its purpose. Fewer spelling errors alone would not establish progress on paragraph organisation.
Science: a learner previously multiplied voltage and current when asked for resistance. Later they identify R = V/I, substitute the quantities and write ohms. A fresh example should check formula selection again rather than merely changing the last digit.
These are illustrative examples, not BananaLearn outcome claims or promises about a particular child. Use a goal and task relevant to your own learner.
Separate learning evidence from useful supporting signals
Enjoyment, willingness to ask a question and attendance matter to the learning experience. They can support a constructive lesson, but they are not substitutes for evidence that the child can use the idea. Record them separately.
School marks are also useful but have context: paper difficulty, topics, marking, time limits and support may change. A rise in a school mark does not by itself prove that tutoring caused the rise. Discuss it alongside actual work samples and teacher feedback.
Look for patterns across a few opportunities rather than treating one difficult day as a verdict. At the same time, repeated vague reports such as “doing well” without examples are a reason to ask for clearer evidence.
If the child feels uncomfortable or cannot follow the explanations, raise that directly. A lesson should allow questions and correction without humiliation. Do not wait for a formal score review to discuss a poor learning experience.
Use a focused review conversation
Bring the agreed goal, the starting sample and the later work. Ask the tutor: “Which part now works independently?”, “What support is still needed?”, “What is the next small target?” and “How will we check it?”
Ask the child which explanation helped and which step still feels unclear. They may describe a difficulty the work sample does not reveal. Keep the conversation specific and avoid asking the child to defend every mistake in front of adults.
Choose a response based on the evidence. If the skill is improving, continue and widen the application. If a prerequisite is missing, adjust the goal and teaching. If progress evidence remains unclear, agree a better check and a short review interval. If explanations or fit remain poor, consider a different approach or tutor.
Do not add more lessons automatically without identifying what the extra time would address. Frequency is a planning choice; this guide does not set a recommended package or promise a return on a purchase.
A progress log you can actually maintain
Use one row per review: date; agreed skill; task; support used; what worked; first difficulty; next step; next check. A short concrete entry is more useful than a long list of completed chapters.
Example: “Remainders in transport problems. Needed help drawing groups before; now draws groups independently. Chose seven cars for 27 children with four seats. Still forgets to state the unit. Next: two fresh contexts and label the answer.”
Keep records private and proportionate. The purpose is to make teaching decisions, not compare the child publicly with classmates or create an extra reporting burden. Ask the tutor for an example from ordinary lesson work where possible.
The practice tasks below help you interpret evidence and draft a review note. Their model answers are examples of reasoning, not a scoring scale for tutors.
4 worked examples with reasoning
1. The report says “Fractions chapter completed”. What follow-up makes it useful?
- Ask which specific skill was taught and request a fresh example of the child using it.
- Ask what support was needed and what the child can explain.
- Record the evidence separately from the chapter-completion statement.
2. A learner gets 9/10 on repeated questions and 5/10 on new questions of the same type.
- The repeated score may include memory for those answers.
- Inspect the first wrong step on the new tasks to see whether selection, understanding or execution is difficult.
- Use fresh matched tasks for the next check; do not label the child from these two totals.
3. A child now explains the operation but still makes one multiplication error.
- Separate the modelling goal from the calculation.
- Record that operation choice is now explained independently, while the fact or calculation needs checking.
- Retain the progress and set a precise next target instead of writing only “wrong answer”.
4. The tutor and parent disagree about progress.
- Compare the goal and conditions used by each person.
- Review the same work sample and note hints, difficulty and expected reasoning.
- Agree a fresh appropriate check rather than debating an undefined idea of “better”.
Save the guide or print the practice
Download to save or share
Complete guide with answers (PDF) · Worksheet without answers (PDF) · Worksheet and answer key (PDF)
Try each question before opening its explanation. Show the calculation, evidence or reasoning that supports your answer.
10 practice questions with explanations
Try each question before opening its explanation. Show the calculation, evidence or reasoning that supports your answer.
Write an answer and your reason before opening the explanation. Check the reasoning as well as the final answer.
- Goal
Rewrite “Improve English” as an observable goal.
Answer and reasoning
Example: write a paragraph with one main idea and two relevant supporting details.
The goal names a task whose features can be inspected.
- Evidence
A correct answer follows three prompts. How should it be recorded?
Answer and reasoning
Correct with three prompts; name the prompts if useful.
It is supported success, not yet evidence of the same task being independent.
- Comparison
Why should a much harder later task not be compared only by percentage?
Answer and reasoning
The task difficulty changed.
Inspect matched skills and conditions before interpreting the difference.
- Transfer
Suggest a fresh task after a child learns to share 24 counters equally among four groups.
Answer and reasoning
Example: share 35 stickers equally among five children and explain the result.
The context and numbers change while equal sharing remains the underlying idea.
- Maths
27 children travel in cars with four child seats each. How many cars are needed?
Answer and reasoning
Seven cars.
Six cars seat 24, leaving three children; a seventh car is needed.
- English
A paragraph has accurate spelling but unrelated sentences. Has paragraph organisation been demonstrated?
Answer and reasoning
Not necessarily.
Spelling and coherence are different skills; inspect the agreed goal.
- Science
What is missing from “R = 8/0.4 = 20”?
Answer and reasoning
The resistance unit, ohms, in the final result.
Also check that 8 V and 0.4 A refer to the same component.
- Review
Write one question to ask the tutor about next steps.
Answer and reasoning
“Which step still needs support, and what fresh task will check it?”
The question connects the plan to evidence.
- Decision
Should more sessions be the automatic response to unclear progress?
Answer and reasoning
No. First clarify the difficulty, teaching approach and evidence.
Extra time is a planning option, not a substitute for a clear goal.
- Record
Write a short review entry using a real work sample.
Answer and reasoning
State skill, support, what worked and the next check.
Use actual observations; do not invent an improvement to fill the log.
Use the ideas without looking at the notes
Review record: Goal ___. Starting task/date ___. Later task/date ___. Support used each time ___. What the child can now explain ___. Remaining difficulty ___. Agreed next step ___. Next check ___.
Bring the record to a tutor conversation. Keep outcome claims limited to the work and conditions you actually observed.
Common questions
How quickly should tutoring show results?
There is no single timetable for every learner or goal. Agree a review date and observable target, then inspect the evidence rather than relying on a promised deadline.
Is enjoyment enough?
Enjoyment is valuable feedback about the experience. Check independent learning separately with a suitable task.
Should parents mark every lesson?
No. A small sample and a focused conversation can be enough for review. Avoid turning home into a second assessment session every day.
Can a score improve for reasons other than tutoring?
Yes. Task difficulty, school teaching, practice and support can also change. A score alone does not establish cause.
When should we change the plan?
When work samples show a different prerequisite need, the current explanation is not helping, or the agreed goal no longer fits. Discuss the evidence with the tutor and child.
Connect your learning
Ask about guided Learning support learning if your child needs help explaining these ideas.
About this guide
Original explanations, examples, diagrams and practice by BananaLearn. The grade label indicates the intended learning level; this is not an official textbook extract or exam paper.