Most working memory and speech-in-noise studies are underpowered to detect real effects.
Evidence
Correlation coefficients in the surveyed studies are broadly distributed, spanning almost half of the possible range of r values (i.e., from ‑0.29 to 0.58). Confidence intervals (CIs) are generally large and include the null hypothesis in 21/25 and 24/25 cases for CIs of 95 and 99 %, respectively, suggesting that these studies are not appropriately powered. — Dijk, Pim van, Physiology, Psychoacoustics and…, 1123
Confidence intervals (CIs) are generally large and include the null hypothesis in 21/25 and 24/25 cases for CIs of 95 and 99 %, respectively, suggesting that these studies are not appropriately powered. — Dijk, Pim van, Physiology, Psychoacoustics and…, 1129
Categorization
- QUESTION: A caution worth remembering when reading positive claims in cognitive hearing science — sample sizes are systematically too small.
- XREF: Part of the wider well-powered-samples critique in cognitive hearing science, echoing replication-crisis findings in psychology. — connects to: Published studies linking working memory to speech-in-noise are underpowered.; Working memory capacity explains under two percent of speech-in-noise variance in young listeners. — Dijk, Pim van, Physiology, Psychoacoustics and…, 1123
- corroborated: 3 independent passages support this claim — Dijk, Pim van, Physiology, Psychoacoustics and…, 1123; Dijk, Pim van, Physiology, Psychoacoustics and…, 1129
Broader Connections
- Working memory capacity explains under two percent of speech-in-noise variance in young listeners.
#complicates(0.85) — Dijk, Pim van, Physiology, Psychoacoustics and…, 1123