In a pre-registered direct replication (N = 96), the disfluency × WMC interaction did not significantly predict retention, failing to replicate Lehmann et al. (2016)¶

Claim · All claims
Evidence · 1 study · causal r2 · q3
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? The interaction of disfluency manipulation and standardized WMC was not a significant predictor of retention in the replication regression model (β = 4.19, p = 0.397). → Sophia C. Weissgerber 2024

Evidence¶

Sophia C. Weissgerber 2024¶

Sophia C. Weissgerber, Denia Indah Permatasari Terhorst, Ralf Rummer. (2024). Is the perceptual disfluency effect moderated by working memory capacity? Direct replication of Lehmann et al. (2016). Metacognition and Learning. https://doi.org/10.1007/s11409-023-09366-7

q3 · i? · causal · r2

Pre-registered supervised online replication experiment (final N = 96; fluent n = 47, disfluent n = 49) using the same materials, WMC measure, and analytical approach as Lehmann et al. (2016). The regression model was significant overall (F(9, 84) = 2.76, p = 0.007, R2 adj. = 0.15), but "the interaction between disfluency manipulation * WMC, (β = 4.19, SE = 4.92, t(84) = 0.85, p = 0.397), were not predictors of retention". Based on the significance-and-direction criterion, this does not replicate the primary study's interaction (β = 11.66, p = 0.02).

"Disfluency manipulation (β = 0.82, SE = 4.77, t(84) = 0.17, p = 0.864), WMC (both in fluent (β = -0.07, SE = 3.94, t(84) = -0.17, p = 0.865; disfluent β = 3.52, SE = 3.04, t(84) = 1.16, p = 0.25 condition), and the interaction between disfluency manipulation * WMC, (β = 4.19, SE = 4.92, t(84) = 0.85, p = 0.397), were not predictors of retention"

Discussion¶

Related Claims¶

• The null effect of the disfluency × WMC interaction on transfer replicated the primary study's null finding, with both studies showing negative coefficients
• Working memory capacity predicts comprehension performance only in the disfluency condition, with higher WMC yielding better comprehension — related
• Working memory capacity predicts retention performance only in the disfluency condition, with higher WMC yielding better retention — reports the opposite
• Disfluency and WMC × disfluency did not affect extraneous cognitive load, contrary to hypotheses — related
• Across the four existing experiments on WMC moderation of disfluency, results are inconsistent: one finds the effect restricted to high-span, one to low-span, and two find null moderation — a broader claim this one bears on
• Working memory capacity moderated disfluency effects in one study but not another — a broader claim this one bears on
• The primary study (Lehmann et al., 2016) had only about 45% power to detect its observed interaction effect; roughly 150 participants would be needed for 80% power — related
• Disfluency and WMC × disfluency did not affect intrinsic or germane cognitive load — related
• The learning benefits of perceptual disfluency reported by Diemand-Yauman et al. (2011) do not replicate in any general way — related
• All 13 special-issue studies failed to show overall better performance due to disfluency — related
• Disfluency benefits retention and comprehension only for learners with high working memory capacity — reports the opposite