Bias-free research is in the unseen
A biased research is a study that capitalises on certain factors to skew its outcome. These factors may include the sample selection, measurement or method, or publication process.
Bias is introduced when the data collection technique or outcome measured is intentionally selected solely based on its ability to produce significant results. This can compromise the study’s accuracy and integrity.
Again, selectively reporting only the variables that show some significance also introduces strong bias into a study.
Without relying on the outcome of the results, a study must be truly blind. This means the sample size, method selected and variables explored must all be chosen and pre-registered before the results are revealed. The results must also be presented without any selection or change in hypothesis to fit the results achieved.
Other practices that reduce bias are:
- Randomization
- Replication
The image below demonstrates the risk of introducing bias during the research process. Click on each stage of the research study flow diagram to view their risk of bias.