Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 66 of the 67 references it cites were found in a registry or an index; none of the 75 sources its claims cite is checked for support yet.

Perceptual fluency affects judgments of learning: The font size effect

Open access
Yang, Chunliang, Huang, Tina S.-T., Shanks, David R. · 2018 · Journal of Memory and Language

Overview

CoChat Summary

TL;DR

Across three experiments, the authors show that large-font words are identified faster than small-font words in a continuous identification (CID) task, and that this perceptual fluency partially mediates the font size effect on judgments of learning (JOLs) while recall remains unaffected. The CID task is more sensitive than lexical decision, and fluency is at least as important as beliefs about font size in producing the JOL effect.

Key findings

  • Experiment 1 (N = 28): JOLs were higher for large (M = 51.56, SD = 14.90) than small words (M = 47.50, SD = 14.63), difference = 4.05, 95% CI [1.98, 6.13], p < .001, d = 0.76, while recall did not differ (difference = 0.9%, 95% CI [−3.3%, 5.1%], p = .66, d = 0.08). CID identification RTs were faster for large (M = 1.19 s, SD = 0.34) than small words (M = 1.44 s, SD = 0.30), difference = −0.25 s, 95% CI [−0.33, −0.17], p < .001, d = −1.25; fluency mediated 21%, 95% CI [8%, 42%], of the font size effect on JOLs.
  • Experiment 2 (N = 12): CID RTs were faster for large (M = 1.04 s, SD = 0.38) than small words (M = 1.23 s, SD = 0.41), difference = −0.19 s, 95% CI [−0.28, −0.09], p = .001, d = −1.27, but lexical decision RTs did not differ between large (M = 0.63 s, SD = 0.11) and small words (M = 0.60 s, SD = 0.07), difference = 0.02 s, 95% CI [−0.02, 0.06], p = .24, d = 0.36. The task × font size interaction was significant, F(1,11) = 24.97, MSE = 0.13, p < .001, ηp² = .69.
  • Experiment 3 (N = 29): study-phase JOLs were higher for large (M = 50.63, SD = 11.38) than small words (M = 46.40, SD = 10.10), difference = 4.23, 95% CI [2.47, 5.99], p < .001, d = 0.91, with no recall difference (difference = 0.5%, 95% CI [−2.2%, 3.2%], p = .72, d = 0.07). RTs were faster for large (M = 1.17 s, SD = 0.38) than small words (M = 1.43 s, SD = 0.56), difference = −0.26 s, 95% CI [−0.36, −0.16], p < .001, d = −1.00; fluency mediated 15%, 95% CI [7%, 28%], of the font size effect on sJOLs, while beliefs (oJOLs) mediated 9%, 95% CI [−17%, 38%], and the difference between indirect effects via fluency and beliefs was −0.02, 95% CI [−0.90, 0.86].
  • Experiment 3 also found an inverse RT–sJOL relationship (fixed effect = −3.34, 95% CI [−4.49, −2.17]) but no RT–oJOL relationship (fixed effect = 1.41, 95% CI [−4.41, 7.22]); beliefs about fluency did not mediate the fluency effect on JOLs (indirect effect = −0.12, 95% CI [−0.94, 0.58]). The authors conclude the results support dual-basis theory and that the CID task is a more sensitive measure of perceptual fluency than lexical decision.

Methods

  • Experiment 1 used a within-subjects font size manipulation (large 48-point vs. small 18-point). Twenty-eight UCL participants identified 100 monosyllabic 5-letter English nouns via a CID task in which a word and mask alternated, with word presentation increasing in 17-ms steps and mask presentation decreasing in 17-ms steps over 14 cycles; after each correct identification, participants made a 0–100 JOL. A 2-min math distraction and free recall test followed. Analyses included multilevel regression and Bayesian multilevel mediation using the R bmlm package.
  • Experiment 2 used a 2 (font size: large/small) × 2 (task: CID/lexical decision) within-subjects design with 12 UCL participants. Eighty words from Experiment 1 and 40 nonwords were used; participants completed both tasks with the same participants and materials, but no item-by-item JOLs or final memory test were included. The main analysis compared RTs between tasks and font sizes.
  • Experiment 3 used 29 participants (30 recruited, one lost to computer failure). The study task was the same CID procedure as Experiment 1 with 100 words and item-by-item JOLs. This was followed by an observation task in which participants viewed their own identification trials replayed with the letter string “abcde” replacing words, at the same font size and duration, and made oJOLs purportedly for another participant. Free recall and effort ratings followed. Multilevel mediation analyses used bmlm and Mplus for multiple mediators.

Limitations

  • Stated: In Experiments 1 and 3, participants made a JOL immediately following each correct identification, which might draw attention to fluency and inflate its influence on JOLs; this might also contribute to the null difference between the indirect effect through fluency and that through beliefs in Experiment 3.
  • Stated: In Experiment 3, fluency (RTs) and sJOLs were collected in the study task, but beliefs (oJOLs) were measured in the observation task, which might contribute to the null difference in the indirect effects. The authors reiterate that they do not reject analytic processing theory and encourage future research to develop more elegant procedures to avoid drawing attention to fluency and to measure fluency and beliefs simultaneously.
  • Evident from the text: Experiment 2 omitted JOLs and the final memory test by design, so it only compares the sensitivity of CID and lexical decision to perceptual fluency and cannot directly test mediation of the font size effect on JOLs. Sample sizes are modest (12 in Experiment 2; 28 and 29 in Experiments 1 and 3), although the paper reports power analyses.

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

Integrity signals

Registry

Crossref · 10.1016/j.jml.2017.11.005

Cites

66 of 66 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

0 Valid · 1 Not valid · 50 Not checked · 100 Own results and interpretations

94 more claims

Create a free account to see all 151 claims and to verify the 61 references still open.

See all claims — free account
  • Probably not a substantial claim. 37
  • It may state two separate claims as one. 22

Create a free account to see these claims.

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentt(27) = 1.08reported p = .29 · recomputed p = .290

    “t(27) = 1.08, p = .29”

  • Consistentt(27) = 0.44reported p = .66 · recomputed p = .663

    “t(27) = 0.44, p = .66”

  • Consistentt(27) = -6.6reported p < .001 · recomputed p < .001

    “t(27) = -6.60,p < .001”

  • Consistentχ²(1) = 24.14reported p < .001 · recomputed p < .001

    “χ 2 (1) = 24.14, p < .001”

  • Consistentt(11) = 0.27reported p = .8 · recomputed p = .792

    “t(11) = 0.27, p = .80”

5 of 17 · the rest aren’t loaded

12 more recomputed tests

Create a free account to see every test the Stats check recomputed.

See all tests — free account

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.

3 of 3

Citations

Cites

Cited by

See all 75

Metadata

Authors
Yang, C., Huang, T. S.-T., & Shanks, D. R.
Affiliations
University College London (GB)
Corresponding author
Yang, Chunliang
Venue
Journal of Memory and Language
Publisher
Elsevier BV
Published
2018-04
Volume
99
Pages
99-110
ISSN
0749-596X (ISSN-L)
OpenAlex
W2772581473
Cited by
75
References
67
Citations by year
2026: 5 · 2025: 7 · 2024: 12 · 2023: 7 · 2022: 7 · 2021: 10 · 2020: 13 · 2019: 9 · 2018: 4
FWCI
3.44
Type
journal article
Language
en
Open access
Licence
www.elsevier.com/tdm/userlicense/1.0/ (tdm)
Locations
  • Journal of Memory and Language (publishedVersion)
  • UCL Discovery (University College London) (submittedVersion)
Funders
China Scholarship Council
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
  • Social Sciences › Psychology › Developmental and Educational Psychology › Reading and Literacy Development
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Cognitive Functions and Memory
Keywords
judgments of learning, perceptual fluency, metacognition, metacognitive confidence, lexical decision, stimulus size, memory confidence

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 2026-11-18, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-18 (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.