Although no formal multiple comparison computations have been done, the fact that: (1) many of the nominal p-values for associations of T1D with outcomes were at the P <.0001 level, (2) the relevant findings were all in the direction of higher risk for T1D group, and (3) the 3 composite variables were associated with the T1D with P <.0001, allow us to infer that the majority of the nominally significant findings are not due to type 1 errors.
← all excerpts
Maternal and Perinatal Outcomes in Pregnancies With Type 1 Diabetes and Controls During a 15-Year Period (2006-2020).
1
—
—