Practice questions: how the questions inside study material are written, placed and answered

Question: how are the questions inside study material written, placed and answered, so that answering them makes the learner learn more, and what does the evidence say about their kind, format, wording and feedback?

Steps frame review state of the art position against baseline protocol sources bibliography

One reviewer searched on 2026-10-29, by the shared method.

Files

File Role
frame the problem, the terms, and what research and practice call them
review existing reviews, then one strand per sub-question, practice and Q/A sites
state of the art best evidence against most adopted, with adoption counts
position what is known, what is not, and where practice and evidence part
against null results, controversies and limits
baseline the named baseline and how a shape would be compared with it
protocol every string, count and rule of the search
sources the flow and one row per source, with grey scores
bibliography every cited work, once
shapes and checks proposals for the question block and for checks, with grades
handoff works met here that belong to other topics
flow.json, openalex.tsv the flow as data; the formal strings in OpenAlex syntax

Claims

Claim Grade Where
Testing transfers to new questions, d = 0.40; to application and inference questions d = 0.32 measured (meta-analysis, full text) review.md 0
Higher-order quiz questions raise higher-order scores, and in two classroom studies lower-order scores too; fact questions do not raise higher-order scores measured, few studies (abstracts) review.md 1
Introductory course items sit mostly at the two lowest Bloom levels: 93% of 9,713 measured, descriptive review.md 1
In classrooms, recall and recognition quizzes teach about equally when feedback follows (0.541 against 0.607) measured (meta-analysis) review.md 2
In the lab, cued recall teaches more than recognition (0.61 against 0.29) measured (meta-analysis) review.md 2
Multiple-choice items with competitive options induce retrieval and teach the facts behind wrong options measured, lab (abstracts) review.md 2
Lures read on a multiple-choice test return as false answers; feedback reduces this measured (abstracts) review.md 2
Answering in one's head works (g = 0.23), and writing the answer beats it by a further g = 0.17 measured (meta-analysis, secondary) review.md 2
Three options are optimal for test quality measured for tests, not for learning (secondary) review.md 3
Item-writing flaws are frequent (about half of items, secondary) and some make items harder or slower measured, descriptive review.md 3
Item-writing guidelines rest on textbook consensus; nearly half of 43 rules had no study measured, descriptive (secondary) review.md 3
Feedback with more information has larger effects (high-information d = 0.99) measured (meta-analysis, full text) review.md 4
Explanation feedback beats the bare answer on new inference questions measured, one study (secondary) review.md 4
Feedback timing makes no average difference, g = 0.03 measured (meta-analysis, secondary) review.md 4
Peeking at the answer before responding reduces learning supported with caveat (old studies, abstracts) review.md 5
Where to place answers in a study text (folded, chapter end, back of book) not known; convention, untested review.md 5
Generated items match expert items after refinement and fall short without it (secondary); learning from them was not found tested measured for item quality, mixed review.md 6
Quiz tools offer deferred, immediate and multiple-try feedback, and hold billions of questions convention (the owners' pages) review.md 7

Named baseline

  • To beat: B1, a self-test of short-answer recall questions with the correct answer shown after each attempt.
  • Practice control: B0, end-of-chapter review questions as a textbook prints them (Spielman 2020).
  • Comparison: delayed test on practised, new inference and distractor-linked questions, with peeking logged (baseline).

Linked, not reviewed again