4.1.5. Errors caused by selection bias

Selection bias should also be considered as a source of error. Selecting an unrepresentative sample will lead to bias in the results. Actively selecting a random sample rather than a self-selecting sample who may well differ from those not included in the survey (volunteer bias), is the best way to avoid this. Using well-trained personnel will avoid any haphazard substitution of properly selected participants with more conveniently available participants who had not been selected, which is another possible source of selection bias.