1. away from the null (false positive results) Potential sources of bias away from the
2. diminishing positive associations reported in null are less obvious in past studies of elec
3. past literature if they are actually a result of tric and magnetic fields and cancer. In the
4. errors in study design or execution; and c) to community-based studies selection bias in
5. enhance precision ofstudy results. the constitution of the control groups is an The principal source of potential bias important consideration. The constitution