Engert, J.

study model method instrument university estimate phenotype bayesian association outcome group regression meta-analysi fi brinogen distribution mendelian analysis parameter randomization value mendelian randomization studies ratio 2 sls number thompson assumption brinogen genotype journal allele heart disease university college london mrc biostatistics unit example confounder ficient level fi tted values approach burges wiley collaboration genotypic groups figure genetic marker health bayesian methods london cambridge copyright london school health study variance analyses section group-based 1298--1311 mendelian randomization davey effect epidemiology statistic dnorm interval individual-based genotypic 1.0 error population fi ve subjects study m table 2 sls method medicine increase subject bayesian method school