Selection bias, information bias, and confounding variable concepts and how to identify them in a study design.
30 cards · basic cards · AI-written, checked twice. Edit anything.
- What is selection bias?
- A systematic error arising from how study participants are selected or retained, leading to a non-representative sample
- What is Berkson's bias?
- A form of selection bias where hospital-based case-control studies produce a distorted association because hospitalized controls differ systematically from the general population
- What is the healthy worker effect?
- A form of selection bias where employed populations appear healthier than the general population because sicker people are less likely to be employed
- What is loss to follow-up bias?
- Bias introduced when participants who drop out of a study differ systematically from those who remain
- What is information bias?
- A systematic error in how exposure or outcome data is measured or classified
- What is recall bias?
- A type of information bias where participants with a disease recall past exposures more accurately or differently than those without the disease
- What is interviewer bias?
- A type of information bias where an interviewer's knowledge of a subject's status influences how questions are asked or recorded
- What is observer bias?
- Systematic error introduced when an evaluator's expectations influence how they measure or interpret outcomes
- What is nondifferential misclassification's typical effect on results?
- It occurs equally across comparison groups and typically biases results toward the null
- What is differential misclassification?
- Misclassification of exposure or outcome that differs between comparison groups, which can bias results in either direction
- What is confounding?
- A distortion of the true association between exposure and outcome caused by a third variable related to both
- What three criteria must a variable meet to be considered a confounder?
- It must be associated with the exposure, associated with the outcome independent of the exposure, and not be an intermediate step in the causal pathway
- What is confounding by indication?
- A form of confounding in observational studies where the reason a treatment was given is also related to the outcome
- What study design methods can control for confounding at the design stage through participant selection?
- Restriction or matching
- What analytic methods can control for confounding after data collection?
- Stratification or multivariable statistical adjustment
FlashKeepers