Log in
Register|
Ken Peh ★ Malaysia, 2013-07-01 20:17 (4734 d 05:02 ago) Posting: # 10904 Views: 6,126 |
|
|
Dear All, We found a discrepancy in the calculation of F value in ANOVA table when we use SAS. The F value of Carryover effect in SAS is calculated by dividing the MS of Carryover over the MS of intrasubject residual. On the other hand, in Chou and Liu example in page 78 (Design and Analysis of Bioavailability and Bioequivalence studies, 2000), the F value of Carryover effect is calculated by dividing the MS of Carryover over the MS of intersubject residual. Wonder which is the correct one  . Appreciate if there is feedback from the group members.Thank you. Regards, Ken |
|
Ohlbe ★★★ France, 2013-07-01 20:40 (4734 d 04:39 ago) @ Ken Peh Posting: # 10905 Views: 4,949 |
|
|
— Regards Ohlbe |
|
Ken Peh ★ Malaysia, 2013-07-02 11:47 (4733 d 13:32 ago) @ Ohlbe Posting: # 10910 Views: 4,933 |
|
|
Dear Ohlbe, TQVM. Yes, I find the answer and solution to my problem. Some how I miss the thread although I have been searching the whole night ? Probably not using the right words during searching. Regards, Ken |
|
ElMaestro ★★★ Denmark, 2013-07-01 22:24 (4734 d 02:55 ago) @ Ken Peh Posting: # 10906 Views: 4,916 |
|
|
Hi Ken, ❝ Wonder which is the correct one There is a bit of religion in this but there seems to be widespread acceptance that the inter-subject variation should be applicable to the sequence or carry-over effect since it is an inter-subject term. Since SAS did not apply the between-MS for you I can imagine that you have omitted the random statement in your SAS code?— Pass or fail! ElMaestro |
|
Ken Peh ★ Malaysia, 2013-07-02 11:49 (4733 d 13:30 ago) @ ElMaestro Posting: # 10911 Views: 4,933 |
|
|
Dear ElMaestro, Thank you for your feedback. I have got excellent explanation from the thread given by Ohlbe. Regards, Ken |
Ing. Helmut Schütz