Log in
Register6x3 Williams' design instead of 3x3 Latin Square! [Power / Sample Size]
Dear Matz!
Not going into details of the CV-calculation, why have you used a 3×3 design? There were already lengthy discussions (see this post, for example) and explanations why such a design is inappropriate - a 6×3 Williams' design should be used...
Obviously 3×3 studies don't die out - but why?
❝ We have done a pilot 3 way crossover study(3trt 3prd 3seq) compairing two
❝ test and one reference using a MIXED model with SUB(SEQ) as random.
Not going into details of the CV-calculation, why have you used a 3×3 design? There were already lengthy discussions (see this post, for example) and explanations why such a design is inappropriate - a 6×3 Williams' design should be used...
Obviously 3×3 studies don't die out - but why?
—
Regards, Jaime
Regards, Jaime
Complete thread:
- CV inter and intra for a 3 way crossover design mathews 2008-03-17 12:55 [Power / Sample Size]
![[Mix]](https://static.bebac.at/img/mix.png "open in Mix view")
- 6x3 Williams' design instead of 3x3 Latin Square!Jaime_R 2008-03-17 16:30
- 6x3 Williams' design instead of 3x3 Latin Square! mathews 2008-03-18 10:32
- CVintra / Cvinter Jaime_R 2008-03-18 15:22
- 6x3 Williams' design instead of 3x3 Latin Square! mathews 2008-03-18 10:32
- 6x3 Williams' design instead of 3x3 Latin Square!Jaime_R 2008-03-17 16:30
Ing. Helmut Schütz