Solve practice questions before checking the answer
Work through exam-style questions on material you just studied — and attempt every question before checking.
Key takeaways
- What it is: Work through exam-style questions on material you just studied — and attempt every question before checking.
- Why it works: Practice questions with real answers create an error-detection and correction loop: you commit to an answer, receive feedback, and update the model. The commitment step — having to decide — forces engagement with your actual current knowledge rather than a floating sense of familiarity. Errors made on practice tests are especially well retained because the discrepancy between your predicted answer and the correct one is informationally rich.
- Evidence: Backed by randomized trials / meta-analyses.
- Avoid: Abandoning a question and checking the answer the moment it feels hard — the productive struggle before and during the error is where much of the learning happens, not in reading the correct answer.
Why it works
Practice questions with real answers create an error-detection and correction loop: you commit to an answer, receive feedback, and update the model. The commitment step — having to decide — forces engagement with your actual current knowledge rather than a floating sense of familiarity. Errors made on practice tests are especially well retained because the discrepancy between your predicted answer and the correct one is informationally rich.
How to do it
- 1After studying a topic, find or create 5–10 practice questions at the level you need to perform.
- 2Complete all questions before checking any — grading mid-way destroys the commitment effect.
- 3When grading, note not just what was wrong but specifically why your reasoning failed.
- 4Re-test on wrong items in 24 hours without re-reading the full material first.
What the evidence says
RCT / meta-analysisErrors on practice tests are corrected and retained at high rates when feedback is provided, because the error-correction creates a distinctive encoding event. Kornell et al. (2009) showed that unsuccessful retrieval attempts followed by correct answer presentation produced better learning than simply reading the answer.
Honest caveat: The benefit of error-correction is most reliable when feedback is provided promptly after the error; delayed or absent feedback can consolidate the wrong answer.
- — Kornell, Hays & Bjork (2009), unsuccessful retrieval attempts enhance subsequent learning, Journal of Experimental Psychology: Learning, Memory, and Cognition
Common mistake
Abandoning a question and checking the answer the moment it feels hard — the productive struggle before and during the error is where much of the learning happens, not in reading the correct answer.
IX Coach frames exercises as questions you attempt before seeing elaboration — so you commit to your current understanding and receive correction where needed, rather than passively consuming correct answers.
Practice this with IX Coach →