1)Use a comparison group of some kind

Accounting Oct 23, 2020

1)Use a comparison group of some kind. Although it may be difficult to randomly assign patients to groups, the use of a comparison group does strengthen validity, even if it comes from a convenience sample.

2)If using a non-random comparison group, match the groups as closely as possible on potential extraneous variables (e.g., age, severity of illness, and number of comorbid conditions).