The following sources of bias apply to any type of study.
There are three main types: selection, information, and confounding bias.
- Sampling bias
- Ascertainment bias
- A form of sampling bias
- Volunteer bias
- Recall bias
- Recording bias
- Response bias
- Beliefs of observer (assessment bias) or subject - including blinding
- Publication bias
- Length-time bias
- Less aggressive, slow-growing cancers are present for longer and are more likely to be discovered by screening. Cancers detected by screening may therefore appear to do better because aggressive, fast-growing tumours are less represented.
- Lead-time bias
- An artefactual improvement in survival by simply diagnosing the condition earlier. For instance, an incidental CT may detect an asymptomatic, but incurable pancreatic cancer. Well defined in Clinical Epidemiology.
- Instrument validity
- especially interviews and questionnaires.