How Teachers Can Use Test Analytics to Teach Better

The short answer: start with the item-level results, find the questions and topics the class struggles with, work out whether the cause is a concept, speed or a flawed question, re-teach what the class missed and re-test. Analytics help only if they change what happens in the next class.

What data you usually have

Most online test platforms, including the institute dashboard on TestMint, give you results at three levels: class or batch (average score, distribution), student (score, time, attempts, section breakdown) and question (how many got it right, which options were chosen). Check which of these your dashboard shows, and use the question-level data first, since it tells you what to teach.

Step 1: Look at the distribution, not just the average

An average hides the shape. A class with an average of 60% may have everyone at 55 to 65 or half at 90 and half at 30. These need different responses. An illustration:

Both classes average about 60%, but class X is clustered, so a whole-class lesson will do; class Y is split, so you need two tracks, one for extension and one for repair.

Step 2: Find the questions most students missed

Sort questions by the share of students who got them right. Look at the lowest ten. For each, ask:

1. Is the question flawed? If strong students miss it or a distractor is popular among them, check the key and wording first.
2. Is it a concept the class has not understood? Many students choose the same wrong option for the same reason.
3. Is it a speed problem? Students got it right when given time, or left it unattempted.

Step 3: Read the wrong options

When students choose a wrong option, the choice shows how they think. An illustration for the markup-and-discount question (25% markup, 20% discount; correct answer 0%):

Over half the class made the same error. That is a teaching point, not a student problem: re-teach successive percentage change with multiplying factors, then re-test with similar questions.

Step 4: Separate concept from speed

Step 5: Group students for action

• Need-concept group: students with low accuracy on a topic. Give a short re-teaching session and 15 practice questions.
• Need-speed group: students with good accuracy but low attempts. Give timed sets with a ceiling.
• Need-extension group: students who got most right. Give harder questions or let them tutor peers.
• Need-attention group: students who did not attempt the test or whose scores dropped sharply. Talk to them.

Keep groups flexible: students move between them as topics change.

Step 6: Close the loop with a re-test

After re-teaching, assign a short test on the same topics a week later. Compare the question-level results with the first test. If the share of students getting the key question right rose from 24% to 70%, the re-teaching worked; if not, the explanation or the practice needs another approach. Keep the data in a simple table by topic and test.

Share analytics with students

Students benefit from their own time and topic breakdown. Show them their section accuracy and time, point them to their top two weak topics and ask them to keep an error log. See how to analyse mock test results and how to build an error log. Be careful with public rankings in a class: they can discourage the students you most want to help. See how to use leaderboards.

Privacy and fairness

• Share individual results with the student and, where appropriate, their guardians, not with the whole class.
• Use data to support students, not to label them.
• Remember that a single test is one data point.

Common mistakes

• Looking only at averages.
• Not checking the question when everyone misses it.
• Re-teaching without re-testing.
• Treating speed problems as concept problems.
• Publishing rank lists that discourage students.

For how teachers run assessments in the first place, see how teachers can conduct online assessments and online testing for teachers.