Loading page…
Why it works
Paper Tiger turns studying into an evidence loop: explain it first from memory, compare the attempt with a real answer key, repair the specific gap it exposes, then come back after a delay.
Recall · Compare · Repair · Review
Paper Tiger is built around evidence you can see: the learner produces an answer, the system checks it, and the next task depends on what actually happened.
Before — a short primer gives you enough to attempt the question, not to copy the answer.
Recall
Explain the topic from memory. Producing an answer in your own words makes the specific gaps visible.
Compare
Paper Tiger uses the task's answer key when available. If it is incomplete, the result is labelled as relying more heavily on AI judgment.
Repair
Feedback names what you covered and what's missing. When a gap matters, it is corrected or retested on the spot.
Review
Paper Tiger uses your result and review history to set a return date. Answers needing repair come back sooner.
After — once the building blocks hold, you link them into a fuller answer and take a stricter full-topic check. A topic only counts as mastered after repeated strong checks on different days.
Before marking, Paper Tiger pulls up the topic, the current task and any task-specific key points. If the key is incomplete, grading can continue with a clearly labelled AI assessment.
Answer-key grading is more grounded, but missing authoring should not erase the learner's attempt or progress. A fallback result therefore states that it relied more heavily on AI judgment and may be less reliable.
key first; AI fallback when needed
Blurting is a practical form of retrieval practice. The point is not to make studying harsher — it is to check confidence against output, not against a page that is still open.
Rereading can make knowledge feel available because the material is still supplying the cues. Paper Tiger removes that support, then uses feedback to identify the most important gap. When a repair is scheduled, that gap becomes the target.1 3
The result an answer earns — Strong, Nearly there, Developing or Needs repair — decides what the review queue does with it next. The last two rows are deeper checks, not scores.
Needs repair
Blank, too short, or centrally wrong.
Becomes a 'Repair now' task: fix the gap before it enters delayed review.
Developing
Useful content, but still has important gaps.
Marked 'Needs practice': try the relevant task again before moving on.
Nearly there
Good substance with a minor omission or wording issue.
Enters delayed recall; continue, and save the polish for later.
Strong
Clean enough to leave the page.
'Scheduled recall': back after a delay, then lengthening cautiously after repeated success.
Synthesis practice
Building blocks are ready to be linked.
A next task, not a score: a coached linking prompt. It does not mark the topic mastered.
Mastery Check
The learner is ready for a strict full-topic attempt.
A next task, not a score. A strong result can contribute evidence, but one result cannot make the topic Mastered.
Each citation on this page is tied to a design choice: answer first, compare to an answer key, return later, and raise the bar gradually.
A credible learning page should show the ceiling of the evidence, not only the promise.
Roediger & Karpicke, 2006
Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17 (3), 249-255.Used here for: Retrieval practice and delayed-retention evidence.
Karpicke & Blunt, 2011
Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331 (6015), 772-775.Used here for: Meaningful learning, synthesis and retrieval beyond isolated facts.
Rowland, 2014
The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin, 140 (6), 1432-1463.Used here for: Meta-analytic evidence on testing versus restudy.
Dunlosky et al., 2013
Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14 (1), 4-58.Used here for: High-utility learning techniques: practice testing and distributed practice.
Cepeda et al., 2006
Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132 (3), 354-380.Used here for: Distributed practice and the spacing effect.
Latimier et al., 2021
A meta-analytic review of the benefit of spacing out retrieval practice episodes on retention. Educational Psychology Review, 33 959-987.Used here for: Spaced retrieval, and caution about expanding versus uniform schedules.
Yang et al., 2021
Testing (quizzing) boosts classroom learning: A systematic and meta-analytic review. Psychological Bulletin, 147 (4), 399-435.Used here for: Applied and classroom evidence for retrieval practice.
Agarwal et al., 2021
Retrieval practice consistently benefits student learning: A systematic review of applied research in schools and classrooms. Educational Psychology Review, 33 1409-1453.Used here for: Classroom implementation and retrieval practice beyond laboratory tasks.
Endres et al., 2020
It matters how to recall: Task differences in retrieval practice. Instructional Science, 48 699-728.Used here for: Why different retrieval tasks should not be scheduled as identical misses.
These references support the learning principles used by Paper Tiger. They do not imply that any single interval ladder is universally optimal, or that retrieval practice replaces reading, teaching, worked examples or applied exam practice.
The invitation
Run the public demo, or create an account to study with saved progress, a review queue and flashcards.
