Large Sample Size Alone May Not Save The Day - Reliability Matters and Cohort Studies Care

Abstract: 

Neuroscientists are working to rapidly amass the large-scale sample sizes needed to meaningfully study individual differences and support reproducible biomarker discovery (e.g., big cohorts). Although promising, the measurement reliability of individual samples is often suboptimal, thereby requiring unnecessarily large sample sizes to achieve the same goals. Functional neuroimaging provides an example, as recent works suggest that data acquisitions may need to be made longer to achieve desired levels of reliability. The implications of the present commentary extend to the broader life sciences.

Sponsored by the NYU-ECNU Institute of Brain and Cognitive Science at NYU Shanghai