From Practice to Progress: How Students Improved After Immediate Feedback

A student answering a question incorrectly is not the end of the learning process.

What happens next matters.

In many exam-preparation settings, students complete past papers, receive a mark and move on to another exercise.

The score shows that a student struggled, but it does not always help them understand the mistake or demonstrate whether they can correct it.

IGA AI’s Improvement Challenge was designed around a different approach.

Students first complete a short practice session using real exam-style questions. They then receive immediate feedback on their answers before attempting related questions again.

This creates a simple learning cycle:

1. Practise.
2. Identify mistakes.
3. Receive feedback.
4. Try again.
5. Measure improvement.

During the 2026 national examination preparation period, students completed 189 Improvement Challenges on IGA AI.

The results showed that students generally achieved higher practice scores after receiving feedback and making another attempt.

What is an Improvement Challenge?

An Improvement Challenge gives a student another opportunity to practice after completing an initial session.

Instead of showing only a final score, IGA AI provides feedback intended to help the student understand:

1. the concept being tested;
2. why an answer was correct or incorrect;
3. the reasoning needed to reach the correct answer;
4. common mistakes students make;
5. how to approach similar questions in the future.

The student then completes another set of related questions.

The purpose is not simply to repeat the test. It is to determine whether the student can use the feedback to improve their next attempt.

What the practice data showed

Across 189 completed Improvement Challenges, students recorded the following average changes in practice scores:

Applied Mathematics B showed the largest average change, followed by Chemistry and Physics.

These subjects often require students to apply formulas, understand processes and connect multiple concepts when solving a problem.

A student may know parts of a topic but still struggle to apply that knowledge correctly during an examination-style question.

The results suggest that giving students feedback and another opportunity to practise may help them correct some of these misunderstandings while the material is still fresh.

However, these figures describe changes in IGA AI practice scores. They should not be interpreted as equivalent changes in national examination performance.

Why trying again matters

A mark tells a student how they performed.

It does not always show whether they have learned from the mistake.

For example, a student may answer a Mathematics question incorrectly because they:

1. misunderstood the concept;
2. selected the wrong formula;
3. made an arithmetic error;
4. skipped an important step;
5. misread what the question required.

If the student only sees a score, those different problems can look the same.

Immediate feedback helps make the mistake more specific. A second attempt then provides another useful signal: whether the student can apply the correction.

This is important because improvement is not only about identifying weak students. It is also about identifying which students respond to feedback and which students may require additional support.

What this could mean for teachers

Teachers regularly use tests and exercises to assess their students. The challenge is that a single score may not provide enough information to guide the next intervention.

A student who scores 40% may have a temporary misunderstanding that can be corrected quickly. Another student with the same score may have deeper foundational gaps.

Allowing students to practise again after feedback can help teachers distinguish between these situations.

Schools could use this type of practice data to identify:

1. students who improve quickly after feedback;
2. students who continue struggling with the same concepts;
3. subjects where learners need more structured practice;
4. topics that require classroom revision;
5. students who may need individual teacher support.

This moves assessment beyond asking:

“What score did the student receive?”

It also allows schools to ask:

“Did the student improve after receiving help?”

What this could mean for students

Students sometimes treat mistakes as proof that they are not good at a subject.

A low score can reduce confidence, especially in subjects such as Mathematics, Physics and Chemistry.

Improvement Challenges present a different message.

A first score is not necessarily a final judgment. It can be a starting point for identifying what the student needs to understand next.

When students can see that their score increased after reviewing feedback, they receive visible evidence that improvement is possible.

This may help students focus less on avoiding mistakes and more on learning from them.

Practice should lead to information

Schools and students already have access to past papers, revision books and online questions.

The problem is not always a lack of questions.

The bigger challenge is turning student answers into useful information.

Teachers need to know:

1. which students are practising;
2. which subjects they are choosing;
3. where they are struggling;
4. whether feedback is helping;
5. who still needs support.

Students need to know:

1. why they got a question wrong;
2. what concept they should revise;
3. whether they can solve a related question;
4. whether their performance is improving.

IGA AI is being developed to help connect these two needs.

Students receive structured exam practice and immediate feedback, while schools can gain clearer visibility into participation, weak topics and improvement.

Important limitations

The Improvement Challenge results are encouraging, but they should be interpreted carefully.

The figures in this report:

1. come from students who completed Improvement Challenges on IGA AI;
2. measure changes in practice scores, not national examination results;
3. may be affected by the number of challenges completed in each subject;
4. do not yet measure whether improvement was retained over several weeks or months;
5. should not be treated as proof that IGA AI alone caused every improvement.

Further analysis will be needed to understand how improvement differs by subject, topic, class level and frequency of practice.

IGA AI also plans to study whether students retain what they learn and whether repeated practice leads to consistent long-term progress.

The main lesson

The most important finding is not that students completed more questions.

It is that students were given an opportunity to act on their mistakes.

Across 189 completed Improvement Challenges, students recorded higher average practice scores after receiving feedback and trying again.

This supports a simple principle:

Students do not only need to know that they were wrong. They need to understand the mistake and receive another opportunity to apply what they learned.

For schools, this creates an opportunity to move from simply measuring performance to actively tracking progress.

For students, it shows that an initial score does not have to be the final result.

Request a school demo

IGA AI helps schools see which students are practising, identify weak topics and track whether learners improve after receiving feedback.

Request a school demo to see how IGA AI can support your DOS, teachers and students.

Data source

The figures in this article come from internal IGA AI platform data collected during the 2026 national examination preparation period.