Abstract
Errors that threaten the validity of the results should be considered in every study. Bias involves a systematic error in a study, leading to results that deviate from the real values. Therefore, investigators should make every effort to avoid bias. By using results from the past, retrospective studies are particularly sensitive to bias. As a consequence, retrospective studies, and patient-control studies in particular, are often taken less seriously. However, when a prospective study cannot be performed for practical or ethical reasons, the retrospective design provides the only good alternative. To arrive at a valid retrospective study, various forms of bias must be avoided as much as possible. This means that all possible forms of bias should be considered before starting the study.
Translated title of the contribution | Which forms of bias should be anticipated in retrospective studies? |
---|---|
Original language | Dutch |
Pages (from-to) | 65-70 |
Number of pages | 6 |
Journal | Tijdschrift voor fertiliteitsonderzoek |
Volume | 13 |
Issue number | 3 |
Publication status | Published - 1999 |