Types of Bias Flashcards
A to H (3)
Allocation
Attrition
Hawthorne
M to P (4)
Maturation
Measurement
Natural progression
Placebo
R to V (4)
Regression to the mean
Rosenthal
Selection/sampling
Volunteer
Allocation bias?
Different baseline between groups (RANDOMISE)
Attrition bias?
Patients withdraw from study
- MINIMISE LOSS TO FOLLOW UP
- ITT
Hawthorne effect?
Change from attention received (CONTROL)
Maturation bias?
Pt matures over time (esp children)
- CONTROL
- RANDOMISE
Measurement bias?
Errors
- expectations of pts/therapists/researchers
- BLINDING
Natural progression?
Disease/condition natural ly improves over time
- RANDOMISE
Placebo effect?
Pts expect to get better
- SHAM
- BLINDING
Regression to the mean
Outliers are less likely to be outliers when reassessed
- RANDOMISE
- CONTROL
Rosenthal effect?
Pts perform better because they are expected to
- SHAM/BLINDING
- RANDOMISE
Selection/sampling bias?
Those sampled are different from those not selected (can’t generalise to population)
- GOOD SAMPLING
Volunteer bias?
Systematically different people who volunteer
- RANDOMISE
- CONTROL