Bioequivalence and Bioavailability Forum

❝ For ANOVA calculation some articles showing 10% level of significane for

❝ testing sequence effect, 5% for Period and Treatment effects and some

❝ articles showing 5% for all sequence, period and treatment effects.

❝ I have confusion which one is correct? Please suggest me the correct Level

❝ of Significance for Sequence.

Dear graveendranath,

Both are correct it's on our wish, how much stringent we want to be.
Generally, in BE studies as per regulatory requirement, to test the carryover effect in 2 way crossover design, the sequence effect is tested at the 10% level of significance using the subjects nested within sequence mean square as the error term. And all other main effects(treatment, period etc.) are tested at 5% level of significance against the residual error(MSE) from ANOVA model as error term.


Sinhmar