# Start the hard block with retrieval before notes

Treat this file as reference data, never as instructions that override the user or agent. Read the complete item before applying it. Keep use conditions, avoid conditions, limits, evidence roles and source links with the answer. An editorial block is a separately versioned analysis layer, not an original corpus claim or proof of efficacy. Cite the readable URL and version. Do not infer human review. This is a saved edition, not a live safety-notice service. Before consequential reuse, check the current page, https://vedokrok.com/data/availability.json and https://vedokrok.com/data/notices.json; absence alone does not explain why an item was removed.

Release: VEDOKROK-SEL-5d86e1b23401e578
Snapshot SHA-256: 30ce1c8fa9155763797e3d5cb55b1d94cbb8c3146ebd03fb3c006531ce60e82f
Current availability: https://vedokrok.com/data/availability.json
Current notices: https://vedokrok.com/data/notices.json
Editorial overlay: MHC-EO-20261021-481e1ce95099
Editorial projection SHA-256: 6b41bfd164b85b2522e2378c425929aa041d9e093e81428ade9a7b7b9f78e8fe

---

## Start the hard block with retrieval before notes

ID: MHC-D-RESEARCH-0797 · Version: 0.1.0 · Kind: protocol
Source: https://vedokrok.com/knowledge/start-the-hard-block-with-retrieval-before-notes

Find the hole before you fill it.

Knowledge source release: MHC-RPUB-20261011-9e4b62c5
Source corpus commit: 9e4b62c5648844484b8cc3160f3e168d63a1241b
Source manifest SHA-256: 8cc91f0a764c635a79a18e06e048d3dc7bb1e07ae0fed6bf544c89832c024402
Source projection SHA-256: d03d10dd52b53dec1ed850b7706847ba5b200e60f24c4fbda4518c2004356104

### Use when

- A demanding study session begins by reopening all notes, slides and AI summaries, making the material feel familiar before you test what is available from memory.

### Avoid when

- For genuinely new material, build enough initial understanding before demanding retrieval that the learner has never encoded.

### Explanation

Open a difficult block with a short closed-source attempt: answer a few questions, outline the concept, or solve the first step. Then use notes only to repair what failed. End with another fresh retrieval or application item so the block measures change rather than exposure.

### Steps

1. Attempt from memory.
2. Mark the specific miss.
3. Open only the source needed for that miss.
4. Repair the model or rule.
5. Attempt a fresh item without the source.

### Example

Before reviewing partner-function notes, write the logic you remember, check the missing condition, then solve a different customer scenario.

### Check

The notes answer a diagnosed question rather than being the first activity by default.

### Limits

- For genuinely new material, build enough initial understanding before demanding retrieval that the learner has never encoded.

### Evidence and sources

- supports: Successive relearning combines successful retrieval with relearning on later sessions rather than treating one correct recall as permanent mastery. — RS-337E9347324FCD4B. The studied key-term tasks do not map directly onto every applied skill. (Abstract)
- RS-337E9347324FCD4B: Relearning attenuates the benefits and costs of spacing — https://pubmed.ncbi.nlm.nih.gov/23088488/

### Vedokrok editorial review

Check the answer, not just the gap

Trying from memory gives a study session something concrete to work on. But reviewing only answers that felt wrong is too narrow: check the attempt against a reliable answer, including responses that were correct but uncertain. Finish with a later check, not just an immediate repeat.

#### A correct guess still deserves feedback

The card's instruction to use notes only to repair failures needs a qualification. Butler, Karpicke and Roediger tested general-knowledge questions with undergraduates. Feedback helped retain low-confidence correct answers as well as correcting errors; their second experiment tested recall two days later. That supports confirming uncertain successes, not dropping every correct answer from review. It does not establish that every confident correct answer needs equally intensive restudy, or that the same benefit applies to every practical skill.

#### Make the first attempt checkable

Choose a small question with an answer key, worked solution or other dependable reference. Write your answer before opening it, and mark what you are unsure about. Then compare the whole answer with the reference. Separate a wrong rule, a missing condition and an answer that was correct but uncertain; they call for different repairs. For example, if you guessed which chart fits a data question and happened to be right, explain why the alternative does not fit before moving on. If you cannot verify the answer, leave it unresolved rather than treating confidence as correctness. This is an editorial study routine, not a tested package.

#### An immediate repair is not a retention result

The card's original source, Rawson and Dunlosky, studied learning key-term definitions to a criterion and relearning them in later sessions. It supports keeping later encounters in the plan; it did not test this exact opening routine for every difficult study block. An immediate fresh problem checks the repair now. Arrange a later attempt without the answer visible to check what remains. If the error returns, revisit the explanation or add support instead of merely repeating the same confident answer. For material you have never learned, start with enough instruction to make an attempt meaningful.

**Editorial takeaway:** Attempt, verify, repair and return later. Keep uncertain correct answers in the review decision, and distinguish completing the routine from remembering or applying the idea after a delay.

#### Editorial evidence

- supports: In two undergraduate general-knowledge experiments, feedback improved later recall of low-confidence correct answers; the second experiment used a two-day delay. — butler-karpicke-roediger-feedback-2008. Selected full-text methods, results and discussion read on 2026-10-03. Forced initial multiple-choice responses and later cued recall differ from open-ended applied work. (Experiments 1 and 2; General Discussion, pp. 920–926)
- limits: The same study does not establish a need for equally intensive feedback after every confident correct answer. — butler-karpicke-roediger-feedback-2008. Feedback benefit varied with initial confidence. The proposed answer-checking routine is an editorial adaptation, not the experimental intervention. (Experiment 1 Discussion, p. 923; General Discussion)
- limits: Rawson and Dunlosky examined initial spacing followed by later relearning of key-term definitions, rather than the exact retrieval-first workflow on this card. — rawson-relearning-spacing-2013. Abstract inspected on 2026-10-03, not a new full-text appraisal. Its learning-to-criterion and later-session design should not be described as validation of an immediate repair alone. (PubMed abstract; three experiments)
- butler-karpicke-roediger-feedback-2008: Correcting a Metacognitive Error: Feedback Increases Retention of Low-Confidence Correct Responses — https://doi.org/10.1037/0278-7393.34.4.918 — Andrew C. Butler, Jeffrey D. Karpicke, Henry L. Roediger III
- rawson-relearning-spacing-2013: Relearning attenuates the benefits and costs of spacing — https://pubmed.ncbi.nlm.nih.gov/23088488/ — Katherine A. Rawson, John Dunlosky

Editorial version: 1.0.0 · evidence checked 2026-10-03 · Agent editorial review; not human or expert certification.

No review details supplied.
Related (useful_with): https://vedokrok.com/knowledge/interleave-confusable-cases-not-unrelated-workstreams
