Review

<<>>

Review

TL;DR

Reviewing isn't rereading — it's testing yourself and spacing that testing out over time. You'll learn why retrieval practice and spaced repetition beat highlighting, how to build a review schedule that actually sticks, and how to spot when you're fooling yourself into false confidence.

1. The Mental Model

Your brain doesn't strengthen memories by looking at them again — it strengthens them by struggling to pull them out. Think of each fact as a path through a forest: walking it once leaves a faint trace, but walking it repeatedly, with gaps in between, cuts a clear trail. Review is forest-trail-cutting, not forest-sightseeing. If reviewing feels easy, you're probably not reviewing right.

2. The Core Material

Why rereading fails and retrieval works

When you reread notes, you recognize the material — "yeah, I know this" — but recognition is cheap. Your eyes glide over familiar words and your brain mistakes familiarity for mastery. This is called the fluency illusion. The fix is retrieval practice: closing the book and forcing your brain to reconstruct the answer from scratch.

Roediger and Karpicke's 2006 study is the classic evidence. Students who studied a passage once and then took three practice recall tests remembered 61% of it a week later. Students who reread the same passage four times remembered only 40% — even though the rereaders felt more confident during the study session. The group that felt worse about their learning actually learned more. This gap between felt confidence and real performance is the single biggest trap in exam prep.

Retrieval practice works because every time you successfully pull a memory out, you don't just access it — you rebuild and strengthen the retrieval pathway itself. Each attempt (even a failed one, followed by checking the answer) makes the next attempt easier. This is sometimes called the testing effect.

Spacing: when you review matters as much as how

Cramming creates a tall, narrow spike of memory that collapses within days. Spacing creates a lower, wider plateau that lasts. The mechanism is the forgetting curve, first described by Hermann Ebbinghaus in the 1880s: memory decays roughly exponentially after learning, fast at first, then slower. Each time you review right before you'd forget, you reset the decay and flatten the curve — and crucially, the next decay is slower than the last.

A rough but usable spacing schedule for exam material:

The gaps widen each time because a stronger memory decays more slowly, so you need fewer "top-ups." This is why cramming the night before an exam is the single worst use of review time per minute spent — you're reviewing material you already half-remember instead of material about to slip away.

flowchart LR
    A["Learn topic (Day 0)"] --> B["Review 1 (Day 1)"]
    B --> C["Review 2 (Day 4)"]
    C --> D["Review 3 (Day 11)"]
    D --> E["Review 4 (Day 25)"]
    E --> F["Exam day: quick final check"]
    B -.->|"Forgot? Shorten next gap"| B
    C -.->|"Forgot? Shorten next gap"| C

Building an honest review session

A good review session has three moves, in order:

1. Generate before you check. Cover your notes. Write or say everything you remember about the topic — definitions, examples, the logic connecting ideas — before looking anything up. This is the retrieval step, and it's where the learning happens, not the checking step.
2. Check and correct immediately. Compare what you generated against the real material. Don't just note "I got that wrong" — rewrite the correct version in your own words. Immediate feedback after a retrieval attempt is what makes the attempt count.
3. Interleave topics instead of blocking them. Studying AAAA-BBBB-CCCC (one topic fully, then the next) feels smoother, but mixing them as ABCABC forces your brain to actively decide which method or concept applies each time — which is closer to what an exam actually demands. Rohrer and Taylor (2007) found students who interleaved math problem types scored roughly double on a surprise test compared to those who blocked practice, even though the blocked group felt more fluent during practice.

A practical way to combine all three: use active recall questions instead of notes. Turn every page of notes into 3–5 questions you can't answer by skimming — "Why does X cause Y?" not "What is X?" Store these on flashcards (physical or an app like Anki) and let the app's spacing algorithm handle the scheduling for you.

# Minimal spaced-repetition scheduler (SM-2-style, simplified)
from datetime import date, timedelta

def schedule_next_review(ease, interval, quality):
    """quality: 0-5 self-rated recall score (5 = perfect, <3 = failed)"""
    if quality < 3:
        return 1, max(1.3, ease - 0.2)          # reset interval, soften ease
    if interval == 1:
        new_interval = 1
    elif interval == 2:
        new_interval = 6
    else:
        new_interval = round(interval * ease)
    new_ease = ease + (0.1 - (5 - quality) * (0.08 + (5 - quality) * 0.02))
    return new_interval, max(1.3, new_ease)

interval, ease = 1, 2.5
next_date = date.today() + timedelta(days=interval)
print(f"Next review: {next_date}")

You don't need to build this yourself — Anki does it — but seeing the logic matters: a correct, confident recall pushes the next review further away; a failed recall snaps it back to day one. That's spacing encoded as a rule, not a vibe.

3. Worked Example

Say you're reviewing the Krebs cycle for a biology exam, three weeks out.

Day 0 (first pass): You read through the cycle once, slowly, making sure you understand why each step happens (not memorizing yet). You write 8 recall questions: "What two carbons leave as CO2, and at which steps?" "What does NADH carry out of the cycle?"

Day 1: Cover your notes. Try to answer all 8 questions from memory, speaking out loud or writing them down. You get 5/8 right. For the 3 you missed, you don't just reread the answer — you write out the full correct explanation in your own words, then immediately re-ask yourself that question once more to confirm it stuck.

Day 4: Same 8 questions, now mixed in with questions from the electron transport chain (interleaving) so you're forced to distinguish between the two topics. You get 7/8.

Day 11: You get 8/8, so these questions go into a "mastered — check again in 3 weeks" pile. You add 2 harder follow-up questions that connect Krebs to ATP yield calculations, since simple recall is now too easy.

Day 25 (near exam): A 10-minute mixed recall session across all your biology topics, Krebs included, confirms retention. No new rereading, no last-minute cramming — just confirming the trail is still cut.

Notice what didn't happen: no highlighting, no rereading the textbook chapter five times, no "studying" that consisted of staring at notes. Every session involved producing an answer first and correcting second.

4. Key Takeaways

4.1 Most Important Concepts

• Retrieval practice beats rereading — forcing recall builds stronger memory traces than passive re-exposure.
• The fluency illusion is dangerous — feeling confident while rereading does not predict exam performance.
• Spacing beats massing — reviewing with gaps, timed near the point you'd forget, builds durable memory.
• The forgetting curve justifies the schedule — decay is fastest right after learning, so early reviews matter most, and gaps can widen as memory strengthens.
• Interleaving builds discrimination — mixing topics forces you to choose the right method, which mirrors real exam conditions.
• Immediate correction closes the loop — a retrieval attempt without feedback can reinforce the wrong answer.
• Desirable difficulty is the goal — if review feels effortless, it's too easy to be building memory.

4.2 Common Misconceptions

• "Highlighting and rereading are studying." They build familiarity, not retrieval ability — swap them for self-generated questions.
• "If I feel confident, I know it." Confidence measures fluency, not recall strength; the only real test is closing your notes and trying to produce the answer.
• "Cramming the night before works because I remember it in the exam." Short-term cramming can survive one test but decays within days — fine for a quiz, terrible for cumulative finals or long-term understanding.
• *"Studying one topic fully before moving to the next is more efficient." Blocked practice feels smoother but produces weaker discrimination between concepts than interleaving does.

4.3 Compare & Contrast

5. Now Try It

Pick one topic you have an exam on. Close your notes right now and write down, from memory, 5 questions you think could be asked about it — then answer all 5 without looking. Check your answers against your actual notes, and for anything you got wrong or vague on, rewrite the correct answer in your own words. Finally, schedule your next review of this same topic for 3 days from today (put it in your calendar or phone right now, not "mentally").

Success looks like: you have a written list of 5 self-generated Q&As, honestly marked right/wrong, corrected in your own words, and a real scheduled date for your next retrieval session on this topic.

<<>>