What is it and why is it a problem?

By Jelte M. Wicherts, professor Methodology at the Department of Methodology and Statistics of the Tilburg University.

Bias in research refers to systematic errors in the collection of data, (un)intentional selection of desirable or anticipated analytic or empirical results, unfair assessments of persons, or the drawing of preferred and often unjustified conclusions based on ignoring or downplaying of contrary evidence or arguments. 

For instance, biased research on a particular medical treatment might create the false impression that the treatment works well for most patients and shows no harmful side-effects, while unbiased research on the same treatment might show that it only works for a limited group of patients and shows adverse effects for most patients. 

Biased research could also lead other scientists hoping to build on earlier findings in the wrong direction, leading to wasted resources and less useful research for users of scientific knowledge (e.g., practitioners, governmental organizations, or businesses). And biased research might lower trust in science among the general public. Biases might create unfairness in many settings, which bears on integrity because of its negative consequences for science and society.