Five myths about variable selection

Multivariable regression models are often used in transplantation research to identify or to confirm baseline variables which have an independent association, causally or only evidenced by statistical correlation, with transplantation outcome. Although sound theory is lacking, variable selection is a popular statistical method which seemingly reduces the complexity of such models. However, in fact, variable selection often complicates analysis as it invalidates common tools of statistical inference such as P-values and confidence intervals. This is a particular problem in transplantation research where sample sizes are often only small to moderate. Furthermore, variable selection requires computer-intensive…