BIAS is defined as any process at any stage of inference that tends to produce results or conclusions that differ systematically from the truth.1 Inference is the mental process whereby one uses specifics (eg, the results of a study) to support a generalization. By tradition, one confines the formal discussion of biases to those that originate in the inference performed by the investigator. I will call this form of faulty inference investigator, bias. Thirty-five investigator biases have been defined and catalogued.2 But bias, as defined previously, may also occur during the inference conducted by the reader; I'll call this form of bias reader bias.
Thus, the defining feature of these two classes of bias is their source—investigator bias arising from the investigator (or his investigation) and reader bias from the reader. A second important difference lies in how these two types of bias are avoided or remedied. Usually,