Effects of influential points and sample size on the selection and replicability of multivariable fractional polynomial models.

Studies assessing the stability of variable selection procedures using bootstrap resampling show that the variables with stronger effects are selected in the vast majority of bootstrap replications, whereas those with weak or “borderline significant” effects may enter the model at random [ 20 , 21 ], and their inclusion can be heavily influenced by IPs.

also in 11,409 other papers