Concept · psychology
Learning
15 studies1 discovery
Learning is a lasting change in behaviour produced by experience. The studies here are mostly about which change actually occurred: whether repetition built a habit or the last episode was simply retrieved, whether a cue was learned or only its context, and whether the person in front of you learns the contingency the same way the group average does.
The word covers mechanisms that make different predictions about what will transfer and what will persist. A frequency-built habit and a recently retrieved episode look identical on a reaction-time graph, and telling them apart changes what you would do to weaken the behaviour.
Does repetition show gradual learning or retrieval of the latest episode?
Start with the findings and their cited studies, then compare the boundaries below.
The gradual-strengthening interpretation did not survive controlling for the last trial.
What people choose to practise tracks how much they are still improving.
Before applying these findings
Inspect the most recent trial, the practice schedule and the outcome actually measured. Use the linked spacing, retrieval and interleaving pages for dedicated study comparisons; no single learning technique is established as best across all these tasks.
A tension to inspect
Whether external feedback helps depends on what it replaces. Curiosity-driven exploration was strongly guided by internally computed learning progress, while in motor sequence learning positive social comparative feedback produced no overall speed or self-efficacy advantage over no feedback, and plain reaction-time feedback without social context slowed learning.
Use the Claims tab to inspect each claim and its supporting, challenging, or qualifying studies.
Inspect source passages
Do we form habits from repetition or just our last action?
These passages occur verbatim in the stored paper text. They support the points listed below; they do not verify every statement in the explanation.
30 participants in the first experiment
Thirty native German-speaking FSU Jena students (18 female; M age = 23.03 years; range: 18–30 years) took part in the experiment.
4 word stimuli used in the experiments
Stimuli were the four neutral monosyllabic German adjectives “warm” (“warm”), “klein” (“small”), “ganz” (“whole”) and “fast” (“almost”).
Inspect source passages
How do learning progress and novelty drive human curiosity?
These passages occur verbatim in the stored paper text. They support the points listed below; they do not verify every statement in the explanation.
55 participants
The final sample consisted of 55 participants (M age = 20.44, SD = 2.85, F = 28, M = 27).
3 different cartoon characters
Participants were presented with three characters that were distinguishable by their color (green, blue, and red) and shape (A).
This is a selected set of studies, not an exhaustive literature review. A watch follows covered updates in PaperFren.
Studies
15
Findings
5
9 supporting · 0 challenging · 2 qualifying citations
Open tensions
1
Currently
What we know
- The gradual-strengthening interpretation did not survive controlling for the last trial.
- What people choose to practise tracks how much they are still improving.
- Well-connected knowledge acquires new links more readily — an effect of structure, not salience.
- Individual history predicted both the sharpness of learning and its persistence.
- Newly covered practice-schedule studies show better delayed vocabulary recall after distributed sessions, better transfer after retrieval with feedback than plain rereading, and benefits of mixed physics homework on surprise tests. Their comparisons and outcomes differ, so they do not justify a single universal learning gain.
Largest unresolved question
Whether external feedback helps depends on what it replaces. Curiosity-driven exploration was strongly guided by internally computed learning progress, while in motor sequence learning positive social comparative feedback produced no overall speed or self-efficacy advantage over no feedback, and plain reaction-time feedback without social context slowed learning.
Common misconceptions
Repeating an action builds a habit through accumulated frequency.
In these experiments the frequency advantage vanished once the most recent encounter was controlled, which is what an episodic retrieval account predicts and a strength-accumulation account does not.
Encouraging feedback improves learning.
Positive social comparative feedback did not produce faster motor sequence learning or higher self-efficacy than no feedback at all, and reaction-time feedback delivered without social context produced slower learning than no feedback.
Basic conditioning works essentially the same way in everyone.
Eighteen participants in one study failed to acquire basic conditioned inhibition at all, and higher neuroticism, anxiety and fear scores correlated with weaker inhibitory learning among those who did.
Related
Claim ledger
What the evidence shows
Drawn from 15 studies in this library. Mix labels say which citation roles are present; they are not a strength score. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope.
The gradual-strengthening interpretation did not survive controlling for the last trial.
An apparent frequency effect can be a single-episode retrieval effect. Participants responded faster to high-frequency colour-word pairs by 6 ms and 9 ms across two studies, but the effect disappeared entirely once what they did on the most recent encounter with that word was controlled.
What people choose to practise tracks how much they are still improving.
Exploration is guided by expected learning progress, not only by reward or novelty. Participants were more likely to keep playing with a character while learning progress was high (β = −0.13 for switching as progress fell), and when choosing something new were drawn to both expected learning progress (β = 0.24) and perceptual novelty.
Well-connected knowledge acquires new links more readily — an effect of structure, not salience.
Existing knowledge structure predicts what new material sticks. Cue words with higher degree centrality in semantic networks produced significantly better recall and recognition of their paired responses, and the advantage was not explained by concreteness or word frequency.
Individual history predicted both the sharpness of learning and its persistence.
Who learns a discrimination, and how rigidly, varies systematically between people. Higher tension-stress scores and an experimentally induced drop in self-efficacy each impaired discrimination between ambiguous safe and dangerous cues, and each additional point on a childhood trauma questionnaire raised the odds of persistent avoidance habits by 8.3% and 6.6% across two experiments.
Newly covered practice-schedule studies show better delayed vocabulary recall after distributed sessions, better transfer after retrieval with feedback than plain rereading, and benefits of mixed physics homework on surprise tests. Their comparisons and outcomes differ, so they do not justify a single universal learning gain.
- Can blocked practice outperform interleaving for novice ECG learners?— Novice ECG interpretation favoured blocked over interleaved practice.
- Are Anki flashcards associated with higher medical-school exam scores?— Flashcard use was observational; exam associations do not establish causation.
Study Role Design N Population Outcome Do spaced sessions and feedback improve vocabulary recall? Supports Human experimentMixed factorial experiment; spacing allocation partly constrained by scheduling N=79 · 87 recruited; 79 analysed Native or advanced German speakers, ages 16–77, without prior Finnish knowledge Delayed cued recall of Finnish–German translations Does retrieval practice beat rereading when key information is matched? Supports Human experimentTwo controlled repeated-measures experiments; retention interval varied in Experiment 1 Experiment 1: 56; Experiment 2: 54 analysed after one exclusion Psychology undergraduates at Erasmus University Rotterdam Correct answers to transfer questions applying passage information in a new domain Does mixed physics homework improve later problem-solving? Supports Human experimentCounterbalanced classroom intervention across two self-selected lecture sections N=290 · 350 enrolled; 290 analysed in Stage 1 and 286 in Stage 2 Undergraduate life-sciences students in introductory physics at UCLA Surprise-test formula recall and problem-solving; secondary midterm examination scores Can blocked practice outperform interleaving for novice ECG learners? QualifiesNovice ECG interpretation favoured blocked over interleaved practice. Human experiment2×2 factorial intervention; initially random allocation with four schedule switches N=51 · 80 enrolled; 70 completed all modules; 51 completed delayed test First-year medical students at McMaster University without prior ECG training Immediate module quizzes and delayed cumulative ECG interpretation quiz Are Anki flashcards associated with higher medical-school exam scores? QualifiesFlashcard use was observational; exam associations do not establish causation. CohortNon-randomised cohort comparison; MCAT-adjusted exam-score analyses N=130 · 78 ever-users and 52 never-users; exam-specific Anki samples 48, 39, 23 and 10 First-year medical students at Boonshoft School of Medicine, class of 2025 Three course-end examinations and an end-of-year basic-science assessment
Debates
Tensions and limits
Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes.
Whether external feedback helps depends on what it replaces. Curiosity-driven exploration was strongly guided by internally computed learning progress, while in motor sequence learning positive social comparative feedback produced no overall speed or self-efficacy advantage over no feedback, and plain reaction-time feedback without social context slowed learning.
Whether external feedback helps depends on what it replaces. Curiosity-driven exploration was strongly guided by internally computed learning progress, while in motor sequence learning positive social comparative feedback produced no overall speed or self-efficacy advantage over no feedback, and plain reaction-time feedback without social context slowed learning.
PaperFren reads this as a limit on how far one study travels — different assays, populations, or outcomes — not a forced fight between papers.
Timeline
How understanding moved
Study years are when the paper was published. Evidence edits are dated changes to this page's claims. Explanations are when PaperFren added a Discovery — not a claim that the science happened that day.
2026
- Working-memory training gains look like better strategies, not a bigger memory store
Concept page published
Learning
- Added a scope qualifier
Are Anki flashcards associated with higher medical-school exam scores?
- Added a scope qualifier
Can blocked practice outperform interleaving for novice ECG learners?
- Added a supporting study
Does mixed physics homework improve later problem-solving?
- Added a supporting study
Does retrieval practice beat rereading when key information is matched?
- Added a supporting study
Do spaced sessions and feedback improve vocabulary recall?
New claim
Newly covered practice-schedule studies show better delayed vocabulary recall after distributed sessions, better transfer after retrieval with feedback than plain rereading, and benefits of mixed physics homework on surprise tests. Their comparisons and outcomes differ, so they do not justify a single universal learning gain.
Change log
What changed
Dated edits to this page's evidence: studies added or removed from a claim, claims added or withdrawn, and new explanations tagged here. Coverage edits are distinct from reviewed research changes and corrections. Dates record PaperFren updates; study publication dates appear in the timeline.
Newly covered practice-schedule studies show better delayed vocabulary recall after distributed sessions, better transfer after retrieval with feedback than plain rereading, and benefits of mixed physics homework on surprise tests. Their comparisons and outcomes differ, so they do not justify a single universal learning gain.
- Coverage updated · New claim
- Coverage updated · Added a supporting study: Does mixed physics homework improve later problem-solving?
- Coverage updated · Added a supporting study: Does retrieval practice beat rereading when key information is matched?
- Coverage updated · Added a supporting study: Do spaced sessions and feedback improve vocabulary recall?
- Coverage updated · Added a scope qualifier: Are Anki flashcards associated with higher medical-school exam scores?
- Coverage updated · Added a scope qualifier: Can blocked practice outperform interleaving for novice ECG learners?
- Coverage updated · Concept page published
Papers
15 studies in this library bear on Learning, ordered by citations.
- Do we form habits from repetition or just our last action?
Habitual actions are driven by remembering our very last action in a situation, rather than the total number of times we have practiced it.
- How do learning progress and novelty drive human curiosity?
People decide to stop exploring a task when they stop learning from it, and they choose their next task based on both how much they expect to learn and how new it is.
- How does stress affect how we learn about safety and danger?
High levels of persistent tension and stress, rather than general anxiety, impair a person's ability to tell the difference between safe and dangerous situations when things are ambiguous.
- Does childhood stress make us more prone to rigid habits?
Adults with a history of early-life stress are more likely to fall into hard-to-break habits to avoid unpleasant outcomes.
- Do well-connected words help us learn new associations?
Words with many connections in our mental lexicon make it easier to form and remember new associations.
- Are Anki flashcards associated with higher medical-school exam scores?
Anki users had higher examination scores in one medical-school cohort after adjustment for MCAT scores, but self-selection prevents a causal conclusion.
- Does Believing You Can't Cope Impair Fear Learning?
People who were led to believe they cope poorly with stress struggled to distinguish between safe and dangerous cues during a subsequent fear-learning task.
- Does mixed physics homework improve later problem-solving?
Interleaved physics homework improved surprise-test performance despite feeling harder, but the study did not find significant gains on the subsequent high-stakes midterms.
Show 7 more studiesShow fewer studies
- Can blocked practice outperform interleaving for novice ECG learners?
Novice medical students did better after distributed instruction, while blocked practice outperformed interleaving on a delayed ECG interpretation test.
- Does personality affect how we learn to stop expecting rewards?
People with higher levels of anxiety and neuroticism are less successful at learning when a stimulus signals that an expected reward will be withheld.
- Is there one best interval between study sessions?
Vocabulary and visual-acuity practice showed different retention across spacing intervals, with the nominal 12-hour group doing well; the study does not establish a universal optimal schedule.
- Does Praising Performance Relative to Others Help Motor Learning?
Telling people they performed better than average does not improve how well they learn a hidden physical sequence compared to giving no feedback.
- Does retrieval practice beat rereading when key information is matched?
Retrieval practice with feedback improved transfer over rereading, but much of the advantage shrank when rereading also highlighted the information needed for the final test.
- Why can easier practice feel more effective than it is?
People rated blocked object-colour study as better even when interleaving improved the final test; judgments tracked immediate performance more closely than later retention.
- Do spaced sessions and feedback improve vocabulary recall?
German-speaking learners recalled more Finnish vocabulary after distributed sessions, while retrieval trials benefited from corrective feedback; spacing groups were not fully randomised.
Compare studies
Select 2–10 studies. Design and N are labels, not a ranking.
Nothing selected yet.
Questions
What is still open
Whether external feedback helps depends on what it replaces. Curiosity-driven exploration was strongly guided by internally computed learning progress, while in motor sequence learning positive social comparative feedback produced no overall speed or self-efficacy advantage over no feedback, and plain reaction-time feedback without social context slowed learning.
Study this conceptflashcards and short-answer questions
How would you distinguish a habit built by repetition from repeated retrieval of a recent episode?
By controlling for the most recent encounter. Both accounts predict faster responding to frequently repeated pairings, so the frequency effect alone does not separate them. When the law of recency study statistically removed what participants did on the last encounter with each word, the 6–9 ms advantage disappeared — which the strength-accumulation account cannot accommodate, since accumulated strength should survive removing one trial.
Why does a null result on positive feedback matter more than it first appears?
Because the comparison included a condition that was worse than nothing. Positive social comparative feedback matched no feedback on learning speed and self-efficacy, while reaction-time feedback without social framing produced slower learning than no feedback. That pattern argues feedback is not a monotonic good whose dose can be increased, but an intervention whose framing determines its sign.
What do the fear-learning studies imply about reporting group-average conditioning results?
That the average can describe nobody in particular. Tension-stress scores and induced low self-efficacy each impaired discrimination between safe and dangerous cues, early-life stress predicted more persistent avoidance habits, and in one study eighteen participants failed the basic conditioning manipulation outright. A mean discrimination score therefore mixes people who learned the contingency with people who did not.