Systematic analyses show that normative model performance strongly depends on sample size and covariate distributions, larger samples yield more stable fits, while misaligned covariates introduce ...

Testing for associations in big data faces the problem of multiple comparisons, wherein true signals are difficult to detect on the background of all associations queried. This difficulty is ...

This is a preview. Log in through your library . Abstract We consider two-armed clinical trials in which the response and/or the covariates are observed on either a binary, ordinal, or continuous ...

Missing data are a frequently encountered problem in epidemiologic and clinical research.1, 2 One approach is to include in the analysis only those participants without missing observations (complete ...

The plots used by Viale et al 2 are one way of exploring prognostic effects of continuous variables nonparametrically, but a more relevant question is that of an interaction between treatment and a ...

Targeting the Hepatocyte Growth Factor–cMET Axis in Cancer Therapy In most analyses appearing in the medical literature, the most common way of dealing with missing (covariate or response) data is to ...