CoChat

The advantage of mixing examples in inductive learning: a comparison of three hypotheses

Open access
Guzman-Munoz, Francisco Javier · 2017 · Educational Psychology

The DOI matches its Crossref record; 24 of the 24 references it cites were found in a registry or an index; none of the 23 sources its claims cite is checked for support yet.

Abstract

Mixing examples of different categories (interleaving) has been shown to promote inductive learning as compared with presenting examples of the same category together (massing). In three studies, we tested whether the advantage of interleaving is exclusively due to the mixing of examples from different categories or to the temporal gap introduced between presentations. In addition, we also tested the role of working memory capacity (WMC). Results showed that the mixing of examples might be the key component that determines improved induction. WMC might also be involved in the interleaving effect: participants with high spans seemed to profit more than participants with low spans from interleaved presentations. Our findings have relevant implications for education. Practice schedules should be individually customised so society as a whole can profit from differences between learners.

Overview

CoChat Summary

TL;DR

Three experiments tested whether the advantage of mixing examples in inductive learning is due to interleaving, temporal spacing, attention attenuation, or working memory capacity. The results consistently favored spaced/interleaved over massed presentation, found little support for temporal gap or attention-attenuation accounts, and showed preliminary evidence linking WMC to the interleaving effect.

Key findings

  • Spaced/interleaved presentation consistently outperformed massed presentation: Experiment 1 new-item classification was .62 vs. .39, \(t[23]=4.1, p<.001\), and studied-item accuracy was .63 vs. .44, \(F[1,22]=18.75, p<.001\); Experiment 2 new-item accuracy was .68 vs. .47, \(F[1,80]=47.61, p<.001\); Experiment 3 immediate test accuracy was .56 vs. .38, \(F[1,113]=71.51, p<.001\), and delayed test accuracy was .54 vs. .35, \(F[1,95]=40.25, p<.001\).
  • Temporal gap did not clearly drive the effect: in Experiment 2 the long-gap advantage was marginal for new items (.62 vs. .53, \(p=.051\)) and studied items (.64 vs. .56, \(p=.059\)); in Experiment 3, temporal gap was non-significant in the immediate test (\(p=.562\)) and only marginal in the delayed test (\(p=.056\)).
  • Attention-attenuation was not supported: tests of studied items showed no effect of order of presentation in Experiment 1 or Experiment 2, and participants often misjudged massing as better (61% in Experiment 1, 52% in Experiment 2, 63% in Experiment 3 thought massing was better or equal, while 74%, 70%, and 60% respectively did better under spaced conditions).
  • WMC showed preliminary links to spacing: Experiment 2 found high WMC more accurate than low WMC (.63 vs. .53, \(p=.039\)) and a marginal presentation-style × WMC interaction (\(p=.094\)); Experiment 3 immediate test had a marginal interaction (\(p=.083\)); combined across experiments (\(N=208\)), WMC correlated with spaced-item accuracy (\(r=.141, p=.042\)) but not massed-item accuracy (\(r=.107, p=.123\)).

Methods

  • Three experiments with University of Groningen undergraduates: Experiment 1 \(N=23\), Experiment 2 \(N=88\) with 4 excluded, and Experiment 3 \(N=118\) with 3 missing the delayed test and 15 excluded from WMC analyses.
  • Materials were 10 paintings by each of 12 relatively unknown artists from Kornell and Bjork (2008); 6 paintings per artist were studied, and new paintings were used for classification tests. WMC was measured with an automated symmetry span task.
  • Presentation style (massed vs. spaced/interleaved) was manipulated within subjects; Experiments 2 and 3 added a between-subjects temporal-gap manipulation, and Experiment 3 added a next-day delayed test. Analyses used mixed ANOVAs, median splits for WMC, and combined correlational analysis.

Limitations

  • The paper states that Experiment 1 had a small sample, that WMC was split by median and may not capture extreme differences, and that the WMC effects were only marginal or preliminary.
  • The paper notes that Experiment 2’s temporal-gap manipulation extended presentation time, adding extra study time as a confound, which Experiment 3 attempted to control.
  • Evident from the text: the WMC evidence rests on marginal interactions and a small combined correlation, and the attention-attenuation conclusion is based on failure to replicate order effects rather than direct support for an alternative attention mechanism.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1080/01443410.2015.1127331

Cites

24 of 24 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 0 Not valid · 16 Not checked · 62 Own results and interpretations

33 more claims

Create a free account to see all 78 claims and to verify the 23 references still open.

See all claims — free account
  • It may state two separate claims as one. 30
  • Probably not a substantial claim. 12
  • The quote only partly states this claim. 3

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

5 of 6 · the rest aren’t loaded

1 more finding

Create a free account to see every finding of this check.

See all findings — free account

Citations

Cites

Research works printed but not linked: 2

  1. Cohen J., 1988, Statistical power analysis for the behavioural sciences
  2. Schneider W., 2002, E-prime reference guide

Cited by

See all 22

Metadata

Authors
Guzman-Munoz, F. J.
Venue
Educational Psychology
Published
2016-01-11
Volume
37
Issue
4
Pages
421-437
ISSN
  • 0144-3410 (ISSN-L)
  • 1469-5820
OpenAlex
W2275976124
Cited by
22
References
24
Citations by year
2025: 2 · 2024: 4 · 2023: 1 · 2022: 2 · 2021: 7 · 2020: 1 · 2019: 3 · 2018: 1 · 2017: 1
FWCI
2.37
Type
journal article
Language
en
Open access
Locations
  • Educational Psychology (publishedVersion, cc-by-nc-nd)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (publishedVersion, cc-by-nc-nd)
  • Data Archiving and Networked Services (DANS) (submittedVersion, cc-by-nc-nd)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (submittedVersion, cc-by-nc-nd)
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Visual and Cognitive Learning Processes
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Categorization, perception, and language
Keywords
interleaving, massing, inductive learning, time gap, working memory capacity, individual differences, practice schedules, individualised learning

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-09-29, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-09-29 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.