Verification record
Every claim in this article was checked against the abstracts of the papers it cites, by a separate pass that was given those abstracts and nothing else. A claim marked supported carries the sentence that supports it, quoted verbatim below.
- References
- 14
- Claims checked
- 6
- Fully supported
- 6
- Checked on
- Oct 1, 2026
Claim by claim
In a pre-registered multi-lab replication, looking times at consistent versus inconsistent test trials showed no significant effect.
SupportedLaboratory studyLaboratory study. Shown under laboratory conditions, not in the field.
“A mixed-model analysis of the looking times at consistent versus inconsistent test trials indicated no significant effect.”
The abstract confirms that in this multi-laboratory replication submitted as a registered report, looking times at consistent versus inconsistent test trials showed no significant effect.
Checked by google/gemini-3.8-flash on Oct 1, 2026
A Bayesian analysis of the data indicated strong evidence for a null effect.
SupportedLaboratory studyLaboratory study. Shown under laboratory conditions, not in the field.
“a Bayesian analysis of the data indicated strong evidence for a null effect”
The claim matches the exact text in the abstract.
Checked by google/gemini-3.8-flash on Oct 1, 2026
None of the pre-specified moderators, mono- versus multilingualism, age, or experimental paradigm, showed significant effects.
SupportedLaboratory studyLaboratory study. Shown under laboratory conditions, not in the field.
“moderating factors that were part of the design: mono- versus multilingualism, age, and experimental paradigm; none of these factors showed significant effects”
The abstract directly confirms that none of the moderating factors included in the design (mono- versus multilingualism, age, and experimental paradigm) showed significant effects.
Checked by google/gemini-3.8-flash on Oct 1, 2026
Robustness analyses across different preprocessing choices did not change the results, and 30 of 31 contributed samples showed no evidence of rule-learning.
SupportedLaboratory studyLaboratory study. Shown under laboratory conditions, not in the field.
“Robustness analyses indicated that the results were not different in the case of different choices in the (pre-)processing of the data. A follow-up analysis for single labs indicated that there was no evidence of rule-learning in 30 of the 31 contributed samples.”
The abstract directly confirms that robustness checks across different preprocessing choices did not alter the results and that 30 of the 31 samples showed no evidence of rule-learning.
Checked by google/gemini-3.8-flash on Oct 1, 2026
The original study being replicated reported that 7-month-old infants learned abstract rules over syllable sequences and generalised them to novel sequences.
SupportedLaboratory studyLaboratory study. Shown under laboratory conditions, not in the field.
“In a seminal study, Marcus et al. (1999) reported that 7-month-old infants learned abstract rules over syllable sequences and were able to generalize those rules to novel syllable sequences.”
The abstract explicitly states that the original study being replicated (Marcus et al., 1999) reported that 7-month-old infants learned abstract rules over syllable sequences and generalized them to novel sequences.
Checked by google/gemini-3.8-flash on Oct 1, 2026
A meta-meta-analysis of 25 phenomena in early linguistic and cognitive development found that effect sizes for most phenomena did not change meaningfully across age.
SupportedMeta-analysisMeta-analysis
“Surprisingly, for most meta-analyses, the effect size for the phenomenon did not change meaningfully across age.”
The abstract directly states that examining 25 phenomena via meta-meta-analysis revealed that effect sizes for most phenomena did not change meaningfully across age.
Checked by google/gemini-3.8-flash on Oct 1, 2026
What this does and does not prove
It proves that every DOI cited resolves to a real registered work whose returned DOI matches the one requested, that none of them is retracted, and that each claim has a named sentence in a cited abstract behind it.
It does not prove that the paper as a whole supports the claim. An abstract can overstate its own findings, and a sentence can be read out of context. That is why the quote is printed here rather than summarised: if the check is wrong, you can see that it is wrong.
Evidence grading is automatic and conservative. Anything it cannot classify is marked unclassified rather than flattered upward.