Back to the on-screen lesson ·
Producing material from memory with the source out of sight, and why rereading feels better on the day than it holds up a week later.
Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.
You will tell a retrieval move from a review move, read what the rereading and recalling groups wrote back a week later, run one free recall trial, and tally a blank-page round honestly.
Nothing is assumed. If you have ever said a phone number over to yourself, or read a shopping list at home and then tried to recite it in the store, you have already met the two moves this lesson compares. This is the first lesson of the course, and every later lesson builds on the one habit it teaches.
| Term | What it means |
|---|---|
| Retrieval | Pulling something out of memory with the source out of sight. |
| Review | Going over material that is in front of you: rereading, highlighting, copying, rewatching. |
| Free recall | Retrieval with no hint at all: a blank page and whatever comes. |
| Idea unit | One separate point in a passage; researchers count them when they score what somebody wrote back. |
| Retrieval practice | Using retrieval on purpose as a way to study, and not only as a check at the end. |
Most people study by putting in: read the page, read it again, mark the important lines. It feels like progress, because each pass is smoother than the last.
The move with the strongest evidence behind it goes the other way. Cover the source, produce what you can, then check. That order, attempt first and look second, is the whole technique. If you look before you try, what follows is review, however hard you concentrate.
Free recall is the plainest version: a blank page and no hints. It feels thin and a little uncomfortable, and that feeling is information. The points that do not come are the ones you would otherwise have found missing at the worst moment: in the meeting, at the counter, in the exam room.
Another way: action
Try it now. Look away from the screen and say the five terms the vocabulary table defined. Then look back and check. That was one whole cycle, attempt then check, and it took about fifteen seconds.
Another way: diagram
Picture two lanes. In the review lane the page stays open and you pass over it again and again. In the retrieval lane the page is closed at one point, something is written from memory, and then the page opens for checking. The closed moment is the whole difference between the lanes.
In a well-known experiment, Henry Roediger and Jeffrey Karpicke (2006) had college students study short prose passages. One group read a passage four times. Another read it once and then spent the same time on three attempts to write it back from memory.
| Tested after | Read four times | Read once, recalled three times |
|---|---|---|
| 5 minutes | 83% | 71% |
| 1 week | 40% | 61% |
Read the table twice. Five minutes later, rereading looked better. A week later it had fallen furthest, and the recalling group was ahead by 21 percentage points. The students were also asked how well they expected to do, and the rereaders were the more confident.
This is a finding about two ways of using the same time, averaged over groups. It does not predict what any one person will remember, and it does not make reading useless: you must meet the material once before there is anything to retrieve.
Karpicke and Janell Blunt (2011) set retrieval against a respected alternative, drawing a concept map with the text open. A week later the students who had practiced recalling the text did better, including on questions that asked them to draw inferences rather than repeat facts. Most students had predicted the opposite.
Olusola Adesope and colleagues (2017) pooled 272 comparisons from 118 published studies. Practice tests beat restudying on average, with an effect size of about 0.5. An effect size measures a gap between groups in units of how spread out the scores are; 0.5 is a medium gap, not a miracle. Some studies found larger gains and some smaller, and the average says nothing certain about any single learner.
The check is honest only if nothing was peeked at during step 3. A point you read first and then wrote down is review. It is worth doing, but it is not counted as produced.
Under the 2025 version of the US naturalization civics test, set by US Citizenship and Immigration Services (USCIS), an officer asks up to 20 questions aloud from a published list of 128, and 12 correct answers pass. There is no page to look at in the interview: every answer has to come out of memory. That makes it a plain case for retrieval practice.
Work through a fictional case. Sam prints the list. Instead of reading it through each evening, he covers the answers with a card, reads a question aloud, says his answer, and only then slides the card down. He puts a tick beside each one he produced and a dot beside each one he missed or peeked at. The first evening he produces 9 of 20; the dots show him which 11 to study. Reading the list would have felt smoother and would have hidden all eleven.
The most common reason people give up retrieval is that it feels worse than rereading. That sensation is real, and it misleads. Rereading feels easy because the words are in front of you; the ease tells you about the page, not about what you could produce without it.
A second trap is treating a thin attempt as wasted time. An attempt that brings back three of eight points has shown you exactly which five to look at, which no amount of rereading would have done.
A third is saving retrieval for the end, as a final check once you feel ready. Retrieval is how you study, not only how you find out whether you have.
Read the leaflet through once, properly.
Home fire safety leaflet: 10 points, read once
There has to be something in memory before anything can come out.
Turn it face down and write every point you can.
Blank page: 6 points written, then nothing for a minute
This is the retrieval; it will feel thin, and that is expected.
Turn it back over and mark right, wrong and missing.
Right: 5. Wrong: 1 (the alarm test is monthly, not yearly). Missing: 4.
Checking comes after producing, so the attempt stays an attempt.
Write where tomorrow's round starts.
Tomorrow: the 1 wrong point and the 4 missing ones, then a fresh blank page
The next round begins where this one showed a gap.
Read the week-later row first.
1 week: recalled 61%, reread 40%
The later test is the one that shows what lasted.
Subtract to find the gap.
61 − 40 = 21 percentage points
A difference between two percentages is counted in percentage points.
Now read the five-minute row.
5 minutes: reread 83%, recalled 71%, so the order is reversed
That reversal is why rereading feels better on the day.
Find what one idea unit is worth.
100 ÷ 20 = 5 percent per unit
A percentage is out of 100, and the passage has 20 units.
Multiply by the units written back.
12 × 5 = 60
Each unit written back adds 5 percent.
State the result.
You want to learn fifteen new words for a Spanish class. You have ten minutes. Which of these moves is a retrieval attempt?
A fictional class tries both methods on a 20-unit passage. A week later the rereading group wrote back 7 idea units and the recalling group 12. Complete the worked comparison.
Turn the rereading group's count into a percentage.
7 ÷ 20 × 100 = p percent
Each unit of 20 is 5 percent.
Turn the recalling group's count into a percentage.
12 ÷ 20 × 100 = q percent
Same passage, same conversion, so the two can be compared.
Subtract to find the gap in percentage points.
q − p = g percentage points
A gap between two percentages is counted in percentage points.
A free recall trial, for practice. Study the six words while the timer runs. When they disappear, write back every one you can, in any order. No cue is offered, so whatever comes is unaided. This is practice, not a test of your memory.
Ask someone to read you the list, then cover it and write what you remember:
A fictional class repeats the classic comparison on a passage of 20 idea units. One group reread it; the other read it once and then practiced recalling it. A week later the rereading group wrote back 5 idea units on average and the recalling group wrote back 12. What percentage of the passage did the recalling group write back?
A week later, the recalling group had written back n percent of the passage.
A fictional blank-page round. Leo's safety leaflet has 12 points. With the leaflet face down he wrote 6 points; when he turned it over, 1 of those turned out to be wrong. Fill in his tally.
| Number of points | |
|---|---|
| Points right from memory | |
| Points to look at first | |
| Points never written |
The 2025 US naturalization civics test is oral: the officer asks up to 20 questions from a published list of 128, and 12 correct answers pass. Rosa, a fictional applicant, practices 20 questions with the answers covered. She answers 16 correctly before uncovering anything; for 3 of those she had peeked at the answer sheet first. Fill in her record.
| Answers produced before looking | Questions to study next | At least 12 produced? | |
|---|---|---|---|
| Rosa's practice round |
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
A fictional study hour before a biology quiz. For each move, say whether it is retrieval or review, and write how many of its minutes were spent retrieving (0 for a review move).
| Retrieval or review | Minutes of retrieval | |
|---|---|---|
| Reread the chapter notes (13 min) | ||
| Covered the notes and wrote the key terms (7 min) | ||
| Highlighted the handout (9 min) | ||
| Answered the chapter questions before looking back (10 min) |
Without looking back: what is the one test that separates retrieval from review, and which group was ahead after a week? Say both aloud before you check. Next: what forgetting looks like over days, and why that makes retrieval worth doing. A trial here is practice, not a test of your memory. What is kept is the list you wrote back on this device today, and it is never set beside anyone else's. You can stop at any point: closing the page is the whole of stopping, and the lesson reopens where you left it.
12. Your turn: a passage has 20 idea units, and someone writes back 12. What percentage is that?, step 3
60 percent of the passage came back
A percentage lets you compare passages of different lengths.