The Hidden Landscape of Missed Effects in Human Functional Neuroimaging

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Brain Research

Background:

• Functional neuroimaging studies aim to reveal brain processes but frequently suffer from underpowering.
• Existing literature lacks guidance for planning adequately powered neuroimaging studies.
• A significant barrier is the inflation of commonly reported effect sizes, which biases study planning.

Purpose of the Study:

• To introduce a correction for effect size inflation bias in functional neuroimaging.
• To provide guidance for planning more accurate and better-powered neuroimaging studies.
• To re-evaluate brain function understanding based on corrected effect sizes.

Main Methods:

• Conducted a mega-analysis of 63 functional neuroimaging studies across seven large datasets (52,979 participants).
• Introduced a novel correction method for effect size inflation bias.
• Utilized corrected effect size benchmarks for planning future studies.

Main Results:

• Common study planning methods using uncorrected effects yield approximately 50% of expected detections at typical sample sizes.
• Missed effects, when accounted for, explain significant additional variance in outcomes.
• Whole-brain multivariate approaches allow detection with smaller sample sizes (n < 50) compared to univariate effects (n > 1,000).

Conclusions:

• Corrected effect size benchmarks are crucial for informed neuroimaging study planning.
• Addressing effect size inflation can recover missed findings and improve understanding of widespread brain effects.
• These findings have implications for improving study design and reproducibility across neuroscience and biomedicine.