A bias is a phenomenon that can interfere with the reliability of study results and therefore lead to poor analysis of the indicators measured. There are all sorts of them. A bias may be inherent in the way the study is designed. These are then methodological biases:
- Poor formulation of questions such as the use of double negation, the use of too vague terms, a strongly oriented question...
- Poor questionnaire structure such as the order of the questions
- Non-representativeness of the population interviewed
Other types of biases, which are more difficult to avoid, may be phenomena external to the study such as cognitive biases (lack of attention on the part of the respondent, degree of knowledge of the subject...), social desirable biases (the respondent seeks to present himself in a favorable light through his answers), social and cultural biases...
