Bias can be eliminated by all except?

Ans. is 'd' i.e., Multivariate analysis Matching eliminates confounding bias Blinding eliminates subject, investigator and analyzer bias, depending on the type of blinding. Randomization elminates selection (Investigator) bias and confounding bias.