In a statistical inquiry, bias is considered to be a characteristic of a question or an approach that causes the introduction of errors into the results or the interpretation of the results of the inquiry.
In a statistical inquiry, a question is said to be biased if the data collected does not reflect the characteristics of the population studied.
Examples
- In an inquiry aimed at finding out the tastes of Montrealers for French cinema, we surveyed customers at a local movie theatre.
- Here, the choice of the sample is biased, because it only contains people who already go to the movies.
- To find out the evolution of the salaries of people between the ages of 25 and 40, we asked the respondents if they believe that their quality of life declined over the course of the target period.
- This question does not really target the evolution of salaries, but rather the perception that the respondents have about their quality of life.