Bias refers to interference in a research study which can lead to misleading results.

One kind of bias is response bias: for example, patients who know they are being given a drug may often show improvement -- whether the drug itself is effective or not.

Another kind of bias is observer bias, an unconscious tendency for experimenters to find the results they expect. For example, suppose an experiment is conducted in which a number of patients are divided into two groups and given either a promising new drug or an inactive placebo (e.g. a sugar pill). If the experimenter knows which patients are in each group, he may quite unconsciously notice more signs of improvement in the drug group than in the placebo group. This will have the result of making the drug seem more effective than it really is.

A third kind of bias is selection bias: if the patients who are assigned to the drug group are less ill than the patients assigned to the placebo group, then the drug group will appear to reduce symptoms -- simply because those patients were less ill to begin with.

A third kind of bias is selection bias: if the patients who are assigned to the drug group are less ill than the patients assigned to the placebo group, then the drug group will appear to reduce symptoms -- simply because those patients were less ill to begin with.

Further Reading : "Putting Ginkgo to the Test"