Log in
RegisterR1 vs R2 [RSABE / ABEL]
Dear ElMaestro,
Yessss Sir. Now we come to "des Pudels Kern". I have done it along these lines above, once with only the R data, once with all data in the usual ANOVA (all effects fixed as the EMA requests us to do).
The third evaluation was done via appropriate intra-subject contrasts (aka Senn's basic estimator approach). BTW: This is the way the FDA code in the progesterone guidance takes in evaluating the reference variability within the scaled ABE framework.
What bothers me is that these evaluations gave such distinct answers. And the most obvious evaluation in the spirit of the EMA crippled model, the first mentioned above, gave the worst answer if our goal is showing equivalence of reference versus itself within a partial replicate study.
To repeat my question: Which result do you trust?
What bothers me further is that a distinction of R into R1 and R2 is purely arbitrary. An ratio R vs. R different from 1 is horrible for me.
Moreover the distinction of R into R1 and R2 and evaluation of the R data alone via ANOVA with effects tmt2, period, sequence and subject within sequence is not universally valid. Try it with the EMA dataset I, a 4 period full replicate study. I got:
Trapped in the type III hell! No estimate of R1 vs R2 obtainable.
It only functions if you omit period from the model.
No idea from my side how to get an estimate of R vs. R within this framework
.
❝ The standard formula applies, doesn't it? The df's are easy from the model or anova, the critical t-value is obtainable at your chosen alpha and the derived df's, the PE (R1 - R2) is extracted via from the model effects, the sigma thorugh the residual, and you have the harmonic mean of no's of sequences under the square root sign.
Yessss Sir. Now we come to "des Pudels Kern". I have done it along these lines above, once with only the R data, once with all data in the usual ANOVA (all effects fixed as the EMA requests us to do).
The third evaluation was done via appropriate intra-subject contrasts (aka Senn's basic estimator approach). BTW: This is the way the FDA code in the progesterone guidance takes in evaluating the reference variability within the scaled ABE framework.
What bothers me is that these evaluations gave such distinct answers. And the most obvious evaluation in the spirit of the EMA crippled model, the first mentioned above, gave the worst answer if our goal is showing equivalence of reference versus itself within a partial replicate study.
To repeat my question: Which result do you trust?
What bothers me further is that a distinction of R into R1 and R2 is purely arbitrary. An ratio R vs. R different from 1 is horrible for me.
Moreover the distinction of R into R1 and R2 and evaluation of the R data alone via ANOVA with effects tmt2, period, sequence and subject within sequence is not universally valid. Try it with the EMA dataset I, a 4 period full replicate study. I got:
Source DF Type III SS Mean Square F Value Pr > F
tmt2 0 0.0000000 . . .
sequence 0 0.0000000 . . .
subject(sequence) 74 118.3549864 1.5993917 8.02 <.0001
period 1 0.3995422 0.3995422 2.00 0.1612Trapped in the type III hell! No estimate of R1 vs R2 obtainable.
It only functions if you omit period from the model.
❝ We could perhaps alternatively think along the lines of a mixed model with a covariance matrix having both a within- and between-sigma2 (the betweens would be off diagonal, the within on the diagonal) and optimise it by REML? I have no idea how to actually do this in R or any other package. Have I been sniffing too much glue? My understanding of covariance matrices is still a bit backward.
No idea from my side how to get an estimate of R vs. R within this framework
.—
Regards,
Detlew
Regards,
Detlew
Complete thread:
- Ref. vs. Ref. in partial replicate design d_labes 2012-04-25 15:52
![[Mix]](https://static.bebac.at/img/mix.png "open in Mix view")
- Ref. vs. Ref. in partial replicate design again d_labes 2012-10-10 10:56
- Ref. vs. Ref. in partial replicate design ElMaestro 2012-10-10 12:39
- Ref. vs. Ref. in partial replicate design d_labes 2012-10-10 16:26
- Ref. vs. Ref. in partial replicate design ElMaestro 2012-10-10 17:16
- R vs. R in partial replicate design d_labes 2012-10-11 09:43
- R1 vs R2 ElMaestro 2012-10-11 10:26
- R1 vs R2d_labes 2012-10-11 12:01
- R1 vs R2 ElMaestro 2012-10-11 12:09
- R1 vs R2 d_labes 2012-10-11 12:46
- treatment df ElMaestro 2012-10-11 13:38
- Missunderstanding d_labes 2012-10-11 16:34
- Missunderstanding, crippled ElMaestro 2012-10-11 18:12
- EMA crippled method - R peculiarities d_labes 2012-10-12 14:54
- All is good, just forget ANOVA - you don't need it! ElMaestro 2012-10-12 16:25
- Just forget ANOVA. What next? d_labes 2012-10-15 08:35
- Just forget ANOVA. What next? ElMaestro 2012-10-15 10:04
- Just forget ANOVA. What next? d_labes 2012-10-15 08:35
- All is good, just forget ANOVA - you don't need it! ElMaestro 2012-10-12 16:25
- EMA crippled method - R peculiarities d_labes 2012-10-12 14:54
- Missunderstanding, crippled ElMaestro 2012-10-11 18:12
- Missunderstanding d_labes 2012-10-11 16:34
- treatment df ElMaestro 2012-10-11 13:38
- R1 vs R2 d_labes 2012-10-11 12:46
- R1 vs R2 ElMaestro 2012-10-11 12:09
- R1 vs R2d_labes 2012-10-11 12:01
- R1 vs R2 ElMaestro 2012-10-11 10:26
- R vs. R in partial replicate design d_labes 2012-10-11 09:43
- Ref. vs. Ref. in partial replicate design ElMaestro 2012-10-10 17:16
- Ref. vs. Ref. in partial replicate design d_labes 2012-10-10 16:26
Ing. Helmut Schütz