Were the subjects appropriately selected?

• Demographics? (age, sex, etc)
• Degree of illness?
• Treated and untreated?
• How well do the subjects reflect your patient population?