How to As You Study, Create a Variety of Questions—multiple-Choice, Short Answer, Essay—about the Content (Skill Sprint)

• The act of generating questions is retrieval practice in disguise. We interrogate the material twice: once to ask, then again to confirm.
• Varying the form (MCQ, short answer, essay) forces deeper processing and reduces the “format shock” when the real assessment appears in an unexpected style.
• Costs exist: writing clean multiple‑choice distractors takes time; we can fall into perfectionism; we risk writing questions that echo the text too closely. We will design constraints to keep throughput high and quality adequate.

A small scene to anchor it: we sit at the kitchen table at 7:15 p.m., the page open to “Cellular Respiration.” We set a 15‑minute timer. We plan to create 8 questions: 4 MCQ, 3 short answer, 1 essay outline. We accept that each MCQ will take 60–90 seconds. We promise ourselves to move on when the bell chimes. Imperfect questions today are better than slick questions never.

The core practice: choose a ratio, then build fast We use a basic ratio to start: 6:3:1 across a 25‑minute block—six multiple‑choice, three short‑answer, one essay prompt with a 3‑bullet outline. If we only have 15 minutes, we cut to 4:2:1. The aim is throughput: 10 questions in 25 minutes, 7 questions in 15 minutes. If we can sustain that for three days, we adjust based on friction.

We assumed long essay drafts per block would build depth → we observed fatigue and fewer total prompts created (two blocks yielded only two essays and zero breadth) → we changed to essay outlines: 3 bullets, 1 sentence each, max 3 minutes. Depth came later when we used those outlines for a timed 8‑minute write.

• Choose a live chapter or lecture: e.g., “Enzymes,” “Keynesian fiscal policy,” “Pointers in C,” “The French Revolution 1789–1794.”
• Decide the ratio and total question count for the block: e.g., 10 questions in 25 minutes, 6:3:1 split.
• Open Brali LifeOS → Question Bank Builder. Create a topic tag (e.g., “BIO101 Respiration”). Create a template with three fields: Type, Stem/Prompt, Answer/Key. Turn on the timestamp option.
• Place a visible constraint: a sticky note that reads “1 minute per MCQ, 90 seconds per SA, 3 minutes per essay outline.”

Mini‑App Nudge: In Brali LifeOS, enable the “Quick Add” tile for your question bank; pin a 1‑minute timer micro‑module on the same screen so you can tap, type, and submit without switching.

Multiple-choice: build stems that teach by contrast We think about MCQs as a way to clarify categories and trap common misconceptions. We use contrasts—what looks right but isn’t.

Our quick rules (and we keep them simple to keep moving):

• One idea per question. Avoid double‑barreled stems.
• Four options (A–D). One best answer. Two plausible distractors drawn from common confusions; one clearly wrong but content‑related.
• Aim to test meaning or application, not trivia when possible.
• Spend 45–60 seconds drafting the stem; 15–30 seconds on options.

Example, topic “Cellular Respiration”:

• Stem: Which statement best distinguishes oxidative phosphorylation from substrate‑level phosphorylation?
  • A. Both occur in the cytosol during glycolysis.
  • B. Oxidative phosphorylation uses a proton gradient to drive ATP synthesis; substrate‑level phosphorylation transfers a phosphate group directly to ADP. [Correct]
  • C. Substrate‑level phosphorylation occurs only in chloroplasts.
  • D. Oxidative phosphorylation requires oxygen only for the Krebs cycle.

We notice the distractors: A is a partial truth for substrate‑level in glycolysis but wrong for oxidative phosphorylation; C is entirely wrong but anchored in cell organelles; D confuses processes and conditions. We write, we move. If we get stuck on perfection, we accept one less distractor quality and log a note: “Revise D later.”

Short answer: tighten precision Short answers expose whether we can name and define with crisp boundaries.

We constrain ourselves to formats:

• Define: “Define X in ≤20 words.”
• Enumerate: “List 3 steps of Y in order.”
• Explain cause/effect: “Why does A lead to B? Answer in 2 sentences.”
• Compute: “Calculate Z given data P” (for STEM).

Example, topic “Statistics—Confidence Intervals”:

• Prompt: Define a 95% confidence interval in ≤20 words.
• Key: An interval from sample data that would contain the true parameter in 95% of repeated samples.

We try to make the expected answer shorter than a tweet. The restraint forces clarity. We push ourselves to include the number when warranted. “List 3 glycolysis investment phase steps” produces a numeric target. If the field demands equations, we include exact forms: “State the formula for standard error of the mean (symbols only).”

Essay prompts: force structure without drafting an opus Our essay prompts build the skill of organizing deep responses. We stay at outline level in the study block; we can do a timed write later.

We use three kinds:

• Explain a mechanism: “Explain how X leads to Y; include at least one counterexample.”
• Compare/contrast: “Compare A and B on 3 dimensions: D1, D2, D3.”
• Argue a position: “Should policy P be implemented? Weigh 2 benefits vs 2 costs and recommend.”

We create a 3‑bullet outline:

• Thesis sentence with stance.
• Three supporting bullets with 1 key piece of evidence each.
• Optional counterargument bullet.

Example, topic “Macroeconomics—Fiscal Policy”

• Prompt: Should expansionary fiscal policy be used in a liquidity trap? Weigh 2 benefits vs 2 costs and recommend.
• Outline:
  • Thesis: In a liquidity trap, expansionary fiscal policy is warranted due to impaired monetary transmission.
  • Benefit 1: Direct demand injection raises output; evidence: 2009 ARRA estimates show 0.9–1.5 fiscal multipliers in short term.
  • Benefit 2: Signals commitment to support employment; mitigates hysteresis.
  • Cost 1: Increased debt burden; long‑run interest costs if r > g.
  • Cost 2: Implementation lags; risk of misallocation.
  • Recommendation: Use time‑limited, investment‑heavy measures; pair with medium‑term consolidation.

We map a doable cycle:

• Block length: 25 minutes (Pomodoro) or 15 minutes if squeezed.
• Target throughput: 10 questions per 25 minutes: 6 MCQ, 3 SA, 1 essay outline. For 15 minutes: 4 MCQ, 2 SA, 1 outline.
• Stop and tag: After each block, tag items with difficulty (1–3) in Brali. We’ll use this for spaced review.
• Immediate test: Spend the last 3–5 minutes answering 2 of your own MCQs and 1 short answer without looking. Note errors.

• Block 1 (20 minutes creation + 5 minutes self‑test):
  • 6 MCQs on “Photosynthesis” (6 items)
  • 3 short answers on “Light reactions detail” (3 items)
  • 1 essay outline: “Compare C3 and C4 pathways” (1 item)
• Break (5 minutes)
• Block 2 (20 minutes creation + 5 minutes self‑test):
  • 5 MCQs on “Calvin cycle” (5 items)
  • 2 short answers on “Rubisco; photorespiration” (2 items)
  • 1 essay outline: “Climate change and crop yield: mechanisms” (1 item) Totals:
• Items created: 18 (11 MCQ + 5 SA + 2 outlines)
• Minutes spent: 45
• Immediate retrieval trials: 6 (answered 4 MCQ + 2 SA during self‑tests)

Notice the arithmetic: 18 items in 45 minutes is sustainable over a week (≈90 items across 5 days). We can measure that.

We face a classic trade‑off: should we spend 3 minutes to craft a devious distractor, or move on? Our rule: set a quality floor (“Would this catch my yesterday‑self?”) and a time ceiling (60–90 seconds per MCQ). If we regularly observe “too easy,” we schedule a 20‑minute revision block every third day to upgrade 10 old questions. That separates generation from polishing.

We also debate breadth vs depth. When a chapter is sprawling, we cut breadth into micro‑themes and assign ratios:

• First pass on a new chapter: 8:2:0 (heavier MCQ for coverage).
• Second pass after a lecture: 4:4:2 (more SA and essays for depth).
• Pre‑exam consolidation: 3:5:2 (precision and argumentation).

• Illusion: “If I can recognize it, I can recall it.” Reality: recognition ≠ recall. Short answers expose recall gaps within 30 seconds.
• Illusion: “I’ll remember this outline later.” Reality: outlines evaporate. We write the 3 bullets now; we’ll flesh them when we rehearse.
• Illusion: “Question quality must be high to help.” Reality: even rough questions drive retrieval benefits; polished questions help later.

A small pivot in the wild: we assumed we needed to write full essay answers to feel prepared → we observed that doing so consumed 20 minutes per prompt and left other topics untouched → we changed to outlines today, with a separate 8‑minute timed write tomorrow morning on one outline. We reduced time per prompt from 20 minutes to 3–4 minutes and increased topic coverage by ~3x.

We adapt the forms.

Math:

• MCQ: test classification and common algebraic errors. Example: “Which integral technique applies?” or “Which of the following is the correct derivative of f(x)?”
• SA: compute a numeric value or show a key step (“Differentiate y = x^x at x = 1; give value only.”).
• Essay outline: explain a proof strategy in 3 bullets (“Prove the contrapositive; key lemma is monotonicity of f.”).

Programming:

• MCQ: what does this snippet output? which complexity class applies? which API is appropriate?
• SA: write a one‑line function signature; state a regular expression; write a boundary test case.
• Essay outline: outline a design decision (e.g., “Trade‑offs between thread pools and async IO; 3 bullets.”).

Languages:

• MCQ: choose correct verb tense; pick correct gender; identify the odd preposition.
• SA: translate a short phrase; conjugate a verb; fill a cloze with exact spelling.
• Essay outline: 3‑bullet plan for a 100‑word paragraph (intro, detail, concluding sentence). We can later expand in 5 minutes.

• Overfitting to your own phrasing: We all slip into writing questions that echo how we think, not how the test asks. Mitigation: once daily, rewrite one MCQ in the style of a past exam or a different textbook. Copy its cadence.
• Time sink on MCQ distractors: If we routinely blow past 90 seconds per MCQ, use a distractor template: one “near miss” (a common misconception), one “scope error,” one “unit error.” Fill blanks; move on.
• Answer leakage: In a rush, we sometimes embed the answer in the grammar (e.g., only one option uses a precise term). Mitigation: after the block, skim for length and structure balance across options; adjust one.
• False sense of progress from raw counts: A pile of questions can mask poor retention. Solution: daily micro‑test. Randomly select 5 items from yesterday; answer cold; note accuracy.
• Cognitive fatigue: Creating questions is heavier than highlighting. If we feel strain, reduce total to 6 items in a 15‑minute sprint; bank the win; come back tomorrow.

We use Brali as the container for the practice. No confetti, just the right levers.

• Create a “Question Bank Builder” deck per course/topic.
• Turn on “Auto‑Tag by Type” (MCQ, SA, Essay).
• Enable the “Spaced Revisit” interval: 1 day, 3 days, 7 days. Each saved question gets a revisit date.
• Add a “1‑minute Quick Add” button to your home view. When a good distractor occurs during a lecture, tap and add it without switching context.
• Schedule a daily 15‑minute block named “Q‑Var Sprint.” The task lives with a checkbox; the journal note captures today’s friction.

We learned that friction in switching apps kills throughput. Keeping the prompt field and the timer in the same screen saved ~10–20 seconds per item—enough to matter over 10 items.

A walk‑through: one block on a real topic Topic: Microbiology—Antibiotic Mechanisms

• Ratio: 6 MCQ, 3 SA, 1 outline. Timer: 25 minutes.

MCQ (example stems):

1. Which antibiotic class inhibits peptidoglycan cross‑linking by binding PBPs?

• A. Macrolides
• B. Beta‑lactams [Correct]
• C. Fluoroquinolones
• D. Aminoglycosides

2. The primary target of fluoroquinolones is:

• A. DNA gyrase/topoisomerase II [Correct]
• B. 30S ribosomal subunit
• C. Folate synthesis
• D. Cell membrane ergosterol

3. Which adverse effect is most associated with aminoglycosides?

• A. Red‑man syndrome
• B. Nephrotoxicity and ototoxicity [Correct]
• C. Hemolytic anemia
• D. Photosensitivity

4. Which mechanism explains MRSA resistance to methicillin?

• A. Beta‑lactamase production
• B. Altered PBP (PBP2a) encoded by mecA [Correct]
• C. Efflux pumps
• D. Decreased porin expression

5. Which antibiotics are bacteriostatic at usual doses?

• A. Aminoglycosides
• B. Beta‑lactams
• C. Macrolides [Correct]
• D. Quinolones

6. Which combination risks serotonin syndrome?

• A. Linezolid + SSRI [Correct]
• B. Penicillin + NSAID
• C. Vancomycin + ACE inhibitor
• D. Daptomycin + Statin

Short answer:

1. Name two antibiotics that inhibit the 30S ribosomal subunit.

• Key: Aminoglycosides, Tetracyclines.

2. State one reason daptomycin is ineffective in pneumonia.

• Key: Inactivated by pulmonary surfactant.

3. In ≤20 words, define the mechanism of action of vancomycin.

• Key: Binds D‑Ala‑D‑Ala terminus, inhibits peptidoglycan polymerization.

Essay outline prompt:

• “Compare beta‑lactams and glycopeptides on mechanism, spectrum, and resistance.”
• Outline:
  • Mechanism: Beta‑lactams bind PBPs; glycopeptides bind D‑Ala‑D‑Ala.
  • Spectrum: Beta‑lactams broader; vancomycin mostly Gram‑positive.
  • Resistance: Beta‑lactamase/altered PBPs vs altered D‑Ala‑D‑Lac.

We log 10–12 minutes on MCQ, 5–6 minutes on SA, 3 minutes on outline, 3–4 minutes on self‑test. We resist the urge to expand the essay. We move on.

A question created today becomes a retrieval seed for tomorrow. In Brali, each item receives a revisit date automatically. On Day 2, we answer three of yesterday’s SA prompts cold; on Day 3, we pick two essay outlines and write one paragraph each in 8 minutes. If we miss an item, we tighten the interval (1–2–4 days instead of 1–3–7). This respects the spacing effect (memory decays roughly exponentially; retrieval at the edge of forgetting strengthens it).

• “Only professors can write good MCQs.” Not true. While expert MCQs are better calibrated, learner‑generated MCQs still deliver retrieval benefits and increase exam transfer, especially when followed by feedback (Agarwal, Bain, and Chamberlain, 2012 report ~15–20% gains in some classroom settings).
• “Question writing steals time from learning.” It is learning. Compared to rereading equal minutes, retrieval practice typically yields higher delayed retention (10–30%).
• “Essays are a different skill.” Yes and no. Outlining builds argument structure and recall of relationships; we can rehearse paragraphs later. A 3‑bullet outline today lowers the activation energy for a 6–8 minute write tomorrow.

• Roediger & Karpicke (2006): Repeated testing improved retention at 1‑week delay by ~14–31% compared with repeated study.
• Karpicke & Blunt (2011): Retrieval practice produced better concept mapping performance than study, concept mapping, or combined methods.
• McDaniel et al. (2007): Classroom quizzes improved unit‑exam performance by ~10 points, with larger effects for short answer than MCQ in some contexts.

We keep the numbers in mind but do not worship them. The real metric is whether we can answer our own questions after a delay—and whether we can do it consistently without burning out.

Design heuristics—the small moves that matter

• Use verbs in stems that match cognitive level: define, list, explain, apply, compare.
• Make at least one MCQ per block that tests application, not fact: “In scenario X, which step fails?”
• Tie each SA to a tiny number: “2 reasons,” “3 steps,” “≤20 words.”
• Keep essay outlines skeletal: 3 bullets, 1 sentence each; add evidence only if it takes ≤30 seconds to recall.
• Tag failures. When we miss an item during self‑test, tag it “red” in Brali; revisit in 1 day. Tag “yellow” if slow or unsure; revisit in 3 days.

Friction log—what we felt and changed

• We felt silly writing “easy” questions. So we added a self‑score (1 easy, 2 medium, 3 hard) and aimed for two 3s per block. This nudged difficulty without perfectionism.
• We felt rushed on essay outlines. We moved essays to the end of the block, so we could scale them (1 or 0 based on time). Result: fewer overruns.
• We noticed we answered our own style too fluently. We took 5 minutes every other day to steal styles: rewrite one of our MCQs in a clinical/board style or past‑paper cadence.

What about collaboration? If we study in a group, we can assign roles:

• Person A drafts 4 MCQs; Person B tries to answer them blind; Person A observes where B hesitates; they edit 1 distractor. Swap.
• Person C drafts 2 SA; Person D sets “model keys” with word limits; they compare.
• One shared Brali deck; tags include author initials. We can filter by peer to get fresh exposure.

This adds accountability and a sense of play. It also uncovers the gap between what we meant and what others read.

Two metrics matter:

• Count of items created per day or per block (throughput).
• Delayed accuracy on yesterday’s items (retention).

We define a threshold we can live with:

• Throughput: 8–12 items per 25‑minute block.
• Delayed accuracy target: 70–85% on prior‑day items. If we’re >90%, we might be writing too easy; if <60%, we might be overreaching or need more feedback.

We log both in Brali. A simple “Count” field auto‑increments per save; an “Accuracy” field logs 0–100% when we answer a 5‑item sample.

Busy‑day alternative (≤5 minutes)

• Pick one subtopic. Create 3 MCQs and 1 short answer in 4 minutes; spend the last minute answering one of them cold. Total: 5 minutes, 4 items. If we have a commute or a queue, we can dictate items into Brali’s voice input.

A short, lived moment like this makes the habit sticky. On days when everything slides, this tiny checkpoint keeps the identity intact: “I am someone who turns content into questions.”

Monday: Two blocks (20–25 minutes each). Focus on breadth; ratio 8:2:0 on a new chapter. Total ≈ 16–20 items. Tuesday: One block creation, one block review. Creation ratio 6:3:1; review random 10 items from Monday. Wednesday: One short block (15 minutes) plus 8‑minute essay write from an outline. Creation ratio 4:2:1. Thursday: Two blocks focusing on weak areas (filter by “red” tags). Creation ratio 4:4:2. Friday: One block mixed; one group swap session (if possible). Saturday or Sunday: Optional 20‑minute revision of MCQs (improve 10 distractors); quick 10‑item cold test from the week.

We review what the numbers say at the end of the week: counts, accuracy, and subjective ease. We adjust ratios.

If the questions start to feel repetitive, we deliberately change one dimension:

• Context shift: write application MCQs tied to a case vignette or scenario.
• Constraint shift: limit SA answers to 12 words. This forces sharper definitions.
• Perspective shift: for one prompt, write the wrong answer and explain why it’s wrong (metacognitive inversion).
• Time shift: do an 8‑minute “essay sprint” on an outline from three days ago; see what sticks.

• Risk: conflating question‑writing skill with content mastery. We can get good at writing plausible fluff. Counter: tie each item to a source page or lecture timestamp; include a reference line in the Brali item (“Source: Ch. 4 p. 87”).
• Risk: burnout from daily creation. Counter: cap at 25 minutes; take a “review‑only” day mid‑week.
• Limit: some domains reward problem solving over recall (e.g., abstract math). We shift weight to SA computations and essay‑style proofs, reduce MCQ.
• Limit: if we have test‑access accommodations or unique constraints (e.g., dyslexia), we might need audio prompts. Use Brali’s voice entry and playback; keep MCQs short and SA longer but spoken.

Our questions are only as useful as our corrections. We check keys. We compare to class notes, textbooks, or a trusted online source. If we can, we run one MCQ per day past a friend or tutor. In Brali, we add a “Feedback” field; a single sentence (“Option C ambiguous: ‘only’ is too strong”) guides the next edit.

We also set a weekly 15‑minute “Key Audit” task. We pick five items with the most “red” tags and verify answers. It’s dull and vital.

We are tired. It’s 9:40 p.m. We promised ourselves ten minutes. We open Brali, choose “European History—Revolutions.” We type without ceremony:

• MCQ: Which factor most precipitated the 1848 revolutions? A) Food prices [Correct], B) Colonial wars, C) Church reforms, D) Gold standard changes.
• MCQ: Which state avoided major revolt in 1848? A) Prussia, B) Britain [Correct], C) Austria, D) France.
• SA: In ≤20 words, define “nationalism” in the 19th‑century European context. Key: Identity and sovereignty claims of a people; desire for nation‑state formation.
• Essay outline: Compare outcomes of 1789 and 1848 in France (3 bullets). We jot: regime change depth, social reforms durability, international ripple effects. We hit save. It took 7 minutes. We answer yesterday’s two SA prompts in 2 minutes. 9 minutes total. We feel a small relief. The act of writing the questions made us see what we still blur. Tomorrow, we’ll test them.

If we already use flashcards, we add variety:

• Divide the deck: 60% SA, 30% MCQ, 10% “outline prompts.” We mark outline prompts as “type Essay” with three empty bullets to fill when reviewing.
• If we do Cornell notes, convert each cue to a SA or MCQ at the bottom of the page.
• If we do problem sets, append a SA “why” after each solution: “Which step would fail if assumption A changed?” The 12‑word limit forces focus.

• Time box: set the 1‑minute micro‑timer visible. When it beeps, we must pick the best available option and move.
• Number box: commit to a minimum count (e.g., 8 items). If we hit it early, we end early. If we lag, we stop at the time ceiling. Count wins; perfection does not.
• Difficulty box: set a target of two “3‑hard” items per block. We do not need all items to be hard; we need a touch of challenge.

• Pick one subtopic you studied in the last 24 hours.
• Open Brali LifeOS → https://metalhatscats.com/life-os/ai-question-bank-builder
• Create 4 MCQs, 2 SA, and 1 essay outline. Set a 10‑minute timer. Do not stop to polish.
• Spend 2 extra minutes answering 2 items blind. Log accuracy.

Tomorrow, we will review those items and add 3 new ones. The momentum comes from the doing.

We end where we began: at the table, in a real hour. We choose to ask better questions because the world will ask them of us—sometimes gently, often without warning. Each small question we craft is a thread in a net. When we need it, the net holds.

Check‑in Block

• Daily
  1. How many questions did we create today? (count)
  2. Did we include all three types (MCQ, SA, essay outline)? (yes/no)
  3. How did retrieval feel during the quick self‑test? (easy / effortful / stuck)
• Weekly
  1. On how many days did we complete at least one Q‑Var sprint? (0–7)
  2. What was our average delayed accuracy on prior‑day items? (percentage)
  3. Which type felt weakest this week? (MCQ / SA / essay) and why?
• Metrics
  • Items created (count per day)
  • Delayed accuracy (% correct on a 5‑item sample)
  • Optional: Minutes spent in Q‑Var sprints

• If the subject involves sensitive content (e.g., clinical scenarios), anonymize cases and respect confidentiality. In Brali, avoid storing identifiable details.
• If we experience anxiety spikes during self‑testing, scale down to 3 items per session and use supportive language in the journal. Over time, widen the window.

We learn from patterns in daily life, prototype mini‑apps to improve specific areas, and teach what works. The skill here is not merely question‑writing—it is disciplined curiosity. We make a habit of asking, “What would a good question be right now?” and answering it quickly. This is how knowledge hardens.