S. Stanley Young, director of bioinformatics at the National Institute of Statistical Sciences, critiques statistical analysis by some epidemiologists