February, 2003
© 2000, 2001, 2003 Scott S. Emerson, M.D., Ph.D.
Session 1- Fixed Sample Design: 27
Statistical Design Issues
Predictive value of statistically significant result depends on
Probability of beneficial drug
·fixed when treatment to test is chosen
Specificity
·fixed by level of significance (.05 by popular convention, so specificity is 95%)
Sensitivity
·statistical power made as high as possible by statistical design