In a small pilot dataset with sparse repeated measurements and no correction for multiple testing, isolated nominally significant p -values following a non-significant global interaction test are highly vulnerable to random variation and carry no inferential weight.
← all excerpts
Feasibility of Repeated Patient-Reported Outcome Collection and Trial Design Implications for Structured Transition Care in Adolescents with Congenital Heart Disease: A Single-Center Pilot Randomized Controlled Study
1
—
—