Allocation bias
Allocation bias results when treatment comparisons fail to ensure that, apart from the treatments being compared, ‘like will be compared with like’. See also N-of-1 crossover, Crossover tests, Cluster allocation and Factorial design.
JLL Essay
2.2 The need to compare like-with-like in treatment comparisons Topics
- Individual patient data
- Pre-clinical
- Fair tests of treatments
- The need to address treatment uncertainties
- Treatment comparisons are essential
- Treatment comparisons must be fair
- Biases
- Design bias
- Allocation bias
- Co-intervention bias
- Observer bias
- Analysis bias
- Biases in judging unanticipated possible effects
- Reporting bias
- Biases in systematic reviews
- Researcher/sponsor bias and fraud
- The play of chance
- Bringing it all together for the benefit of patients and the public