Bioequivalence and Bioavailability Forum

Dear SKR!

❝ In ANOVA procedure the GLM output gives P-values up to four digit number

❝ after decimal point (Ex: 0.6562).

Being not a SAS-user myself - but it should be possible even for this valuable piece of software to come up with more decimal places (keyword: ODS) . Maybe it's possible to configure it to come up with more significant digits (even better in scientific notation)...

❝ As we have to test at 5% LoS, i.e 0.05, is the output value has

❝ to be rounded off to two digit number after decimal before testing?

No rounding should be applied!

❝ For Ex. if ANOVA P-value is 0.0518, without rounding off, it is greater

❝ than 0.05. Hence Conclusion: Insignificant.

❝ With rounding off to two digit number (0.05), It is equal to 0.05, Hence

❝ Conclusion: Significant. Which one is right.

That's easy: 0.0518 > 0.05; the output should be modified in such a way that the conclusion (n.s.) is transparent to a reviewer of the printout.