6 Learning Retention Statistics Every Student Should Know
Most of what people study is gone soon after the test, and a lot of students already know that.
We surveyed 2,614 students, aged 16 to 24, who had sat an exam in the previous six months.
We asked what they still remember, and how they actually study. Let's explore the latest learning retention statistics!
Key Learning Retention Statistics
- 74% of students say they felt ready after reviewing, then went blank in the exam
- 68% say they forget most of a chapter within a week of the test
- Students who used ChatGPT without guardrails scored 17% lower on the exam
Also read: Active Learning and Microlearning Statistics.
74% of students say they felt ready after reviewing, then went blank in the exam (Boterview)
In that survey, we asked whether they had felt ready after reviewing and then gone blank once the exam started.
74% said this had happened at least once in the past year. 28% said it happens on most tests.
The feeling of being ready and the fact of remembering are not the same event.
In my experience, rereading is what creates the feeling.
68% say they forget most of a chapter within a week of the test (Boterview)
We asked how much of a tested chapter they could explain a week later, with the notes closed.
68% said half or less. 19% said they could explain almost all of it.
A grade on Friday is a weak description of what is still there the next Friday.
81% cram the night before at least some of their exams (Boterview)
We asked when they do most of the studying for an exam.
81% said the night before, at least for some exams. 34% said that is their usual plan.
Cramming is not a rare slip in this group. It is the ordinary schedule.
64% say rereading notes is their main method, and 11% say they test themselves (Boterview)
We asked which study method they use most.
64% said rereading or rewriting notes. 11% said closing the notes and testing themselves.
The method students already distrust is still the one they reach for.
Students who used ChatGPT without guardrails scored 17% lower on the exam (PNAS)
In a field experiment with nearly 1,000 high school math students published in PNAS, a standard ChatGPT-style assistant raised practice grades by 48% and cut exam scores by 17% when the AI was taken away.
A tutor version with safeguards raised practice grades by 127%, and the exam drop was largely gone.
An assistant that answers for you helps you finish the homework. It does not help you keep the idea.
17% say a teacher has shown them how to remember, not only what to study (Boterview)
We asked whether a teacher had ever shown them a way to remember the material, rather than only assigning more of it.
17% said yes. 61% said no one had.
Students get blamed for forgetting a skill their classes never taught.
Methodology
Every Boterview figure in this article comes from one survey.
- Name: Boterview Student Memory Survey.
- Sample: 2,614 students aged 16 to 24 who had sat a school or university exam in the previous six months. One response each. Incomplete responses were removed.
- Recruitment: email to Boterview accounts.
- Forget most of a chapter means they said they could explain half or less a week later. Cramming means most of the studying happened the night before. Went blank is their report of the past year, not a scored test.
Limitations
These students already use a learning app, so they may think about study methods more than the average student.
The figures are what they told us. We did not retest their memory.