CoChat

The DOI matches its Crossref record; 51 of the 54 references it cites were found in a registry or an index; none of the 20 sources its claims cite is checked for support yet.

Learning words without trying: Daily second language podcasts support word-form learning in adults

Open access
Alexander, Elise, Van Hedger, Stephen C., Batterink, Laura J. · 2023 · Psychonomic Bulletin & Review

Abstract

Spoken language contains overlapping patterns across different levels, from syllables to words to phrases. The discovery of these structures may be partially supported by statistical learning (SL), the unguided, automatic extraction of regularities from the environment through passive exposure. SL supports word learning in artificial language experiments, but few studies have examined whether it scales up to support natural language learning in adult second language learners. Here, adult English speakers (n = 70) listened to daily podcasts in either Italian or English for 2 weeks while going about their normal routines. To measure word knowledge, participants provided familiarity ratings of Italian words and nonwords both before and after the listening period. Critically, compared with English controls, Italian listeners significantly improved in their ability to discriminate Italian words and nonwords. These results suggest that unguided exposure to natural, foreign language speech supports the extraction of relevant word features and the development of nascent word forms. At a theoretical level, these findings indicate that SL may effectively scale up to support real-world language acquisition. These results also have important practical implications, suggesting that adult learners may be able to acquire relevant speech patterns and initial word forms simply by listening to the language. This form of learning can occur without explicit effort, formal instruction or focused study.

Overview

CoChat Summary

TL;DR

Adult native English speakers with no Italian listened to daily Italian or English podcasts for 2 weeks; Italian listeners improved in discriminating real Italian words from nonword foils relative to controls. This suggests passive, unguided exposure to natural second-language speech can support initial word-form learning.

Key findings

  • In familiarity ratings, the L2 exposure group showed greater pre-to-post improvement discriminating nontrained Italian words from nonword foils than controls: Group × Session × Word Category, t(3556) = 2.20, p = .028; no group difference at pretest (p = .63), but a significant group difference at posttest (p = .009). Nontrained-word ratings increased in the L2 group (p = .007), while foil ratings did not significantly change (p = .12).
  • Overall, trained words were rated most familiar (M = 2.75), followed by nontrained words (M = 2.25), then foils (M = 2.00), F(2, 57) = 17.1, p < .001. However, there was no significant group difference for trained versus nontrained words over sessions, t(3556) = 1.30, p = .19.
  • The nontrained-versus-foil learning effect replicated after excluding any item that appeared even once in the podcasts: Group × Session × Word Category, t(3067) = 2.08, p = .038, suggesting generalization rather than word-specific knowledge.
  • Words with higher transitional probabilities were rated as more familiar overall, F(1, 2389) = 28.6, p < .001; a group × session × TP interaction was found, F(1, 2368) = 4.11, p = .043. L2 ratings correlated with TP at Session 2 (p < .001) but not Session 1 (p = .098), though the between-session change was only trend-level (p = .12). Syllable frequency was unrelated to ratings (p = .57). The word-detection task showed no significant effects of the manipulation.

Methods

  • Seventy-one young adults completed the protocol; one was excluded for listening at 1.5 speed, and seven more were excluded for poor word-detection performance, leaving 31 Italian listeners and 32 English controls. Participants were native English speakers, aged 18–35 (M = 21.42, SD = 3.14), with no prior Italian or significant Romance-language experience.
  • Participants were randomly assigned to listen to 14 daily 1-hour Italian or English podcasts over 2 weeks, in the background while doing daily activities; compliance was tracked with embedded “secret” English words and daily questionnaires.
  • Pre- and posttest Zoom sessions included an exposure + word-detection task with 300 Italian sentences (250 training, 50 target) and a familiarity-rating task (1–4) for 10 trained Italian words, 10 nontrained Italian words, and 10 nonword foils. Analyses used linear mixed-effects models and generalized logistic regression, with key contrasts of nontrained words versus foils and trained versus nontrained words.

Limitations

  • Stated in the paper: unguided learning has key limitations, and most aspects of language cannot be acquired through passive, context-free exposure. The TP-rating increase from Session 1 to Session 2 was only trend-level and not significant. The explicit behavioral measure may trail more sensitive neural indicators. The experiment was not preregistered, and the sensitivity power analysis indicated the study could reliably detect only effects of Cohen’s d = 0.60 or larger.
  • Evident from the text: the final analyzed sample was 63 after exclusions, not the initially reported 70. Testing was online/virtual and listening occurred in the background, so attention and compliance may have varied, although secret-word questionnaires and detection-based exclusions were used. The word-detection task itself showed no significant effects of L2 exposure, so the main learning evidence comes from familiarity ratings.

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

Integrity signals

Registry

Crossref · 10.3758/s13423-022-02190-1

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

18 tests recomputed · 0 with the reported and recomputed p on opposite sides of .05

Verification

Claims the paper makes

0 Valid · 1 Not valid · 15 Not checked · 77 Own results and interpretations

38 more claims

Create a free account to see all 93 claims and to verify the 17 references still open.

See all claims — free account
  • It may state two separate claims as one. 26
  • Probably not a substantial claim. 32

Create a free account to see these claims.

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

1 of 1

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.

  • Doesn’t add upt(2377) = 3.52reported p = .004 · recomputed p < .001

    “t(2377) = 3.52, p = .004”

  • ConsistentF(2, 57) = 17.1reported p < .001 · recomputed p < .001

    “F(2, 57) =17.1, p < .001”

  • ConsistentF(1, 61) = 8.56reported p = .005 · recomputed p = .005

    “F(1, 61) = 8.56, p = .005”

  • ConsistentF(2, 3557) = 2.44reported p = .087 · recomputed p = .087

    “F(2, 3557) = 2.44, p = .087”

  • ConsistentF(1, 61) = 0.48reported p = .49 · recomputed p = .491

    “F(1, 61) = 0.48, p = .49”

5 of 18 · the rest aren’t loaded

13 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.

1 of 1

Citations

Cites

Show all 54 references

Cited by

See all 12

Metadata

Authors
Alexander, E., Van Hedger, S. C., & Batterink, L. J.
Affiliations
Western University (CA)
Corresponding author
Batterink, Laura J.
Venue
Psychonomic Bulletin & Review
Published
2022-09-29
Volume
30
Issue
2
Pages
751-762
ISSN
  • 1069-9384 (ISSN-L)
  • 1531-5320
OpenAlex
W4297983851
Cited by
12
References
54
Citations by year
2026: 2 · 2025: 8 · 2024: 2
FWCI
1.59
Type
journal article
Language
en
Open access
Licence
Locations
  • Psychonomic Bulletin & Review (publishedVersion, other-oa)
  • PubMed (publishedVersion)
  • Scholarship@Western (Western University) (publishedVersion)
  • Scholarship@Western (Western University) (submittedVersion)
Funders
Social Sciences and Humanities Research Council of Canada
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Phonetics and Phonology Research
  • Social Sciences › Psychology › Developmental and Educational Psychology › Language Development and Disorders
  • Social Sciences › Psychology › Developmental and Educational Psychology › Second Language Acquisition and Learning
Keywords
statistical learning, word learning, second language acquisition, adult language learners, podcasts, passive exposure, word recognition, Italian as a second language
MeSH terms
Adult, Humans, Language, Language Development, Learning, Phonetics, Speech Perception, Verbal 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 2026-11-19, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-19 (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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Learning words without trying: Daily second language podcasts support word-form learning in adults. https://app.cochat.ai/papers/10.3758/s13423-022-02190-1
The paper
Alexander, E., Van Hedger, S. C., & Batterink, L. J. (2023). Learning words without trying: Daily second language podcasts support word-form learning in adults. Psychonomic Bulletin & Review. https://doi.org/10.3758/s13423-022-02190-1

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.