Revista Colombiana de Psiquiatría
versão impressa ISSN 0034-7450
Investigation results are of major interest for both the investigator and the reader, this is the reason why their validity is important. While evaluating if a result is valid, we should keep in mind that this can be affected by the random error and by the systematic error. The bias is a systematic deviation (error) of the truth and can appear in any moment of the investigation. Many types of bias exist, some particular to each design and others common to many designs. It is of utmost importance to be able to recognize and then prevent them during the design phase or control them during the analysis in the case of the investigator or in the case of an interested reader to recognize them in order to estimate if the results are valid or not. We review the more common biases in analytical studies.
Palavras-chave : Bias, case-control studies; clinical trials; cohort studies.