Log in
RegisterThe secret of WinNonlin's BE's post-hoc Power analysis [Software]
go on validating WNL's and my formula using unbalanced data
Chow and Liu's famous data, delete Subject # 23 and 24's data both in period 1 and period 2.
For original data-----------------
My result is
For ln-transformed data----------
WNL's result is
My result is
also same at level of 0.0001. Seems my formula works as WNL's inside hiden formula.
The level of the precision is interestingly same to that of BE's 90% CI.
go on validating WNL's and my formula using full data
All the results are satisfactory. Seems I have touched into WNL's black box.
Edit: Merged two posts. [Helmut]
Chow and Liu's famous data, delete Subject # 23 and 24's data both in period 1 and period 2.
For original data-----------------
WNL's result is
Power of ANOVA for Confidence Level 90.00
Power at 20% = 0.99266427My result is
Power Analysis
-----------------------------------------------------------------------------
Parameter T stat 1 T stat 2 P value 1 P value 2 Power
-----------------------------------------------------------------------------
AUC -2.671468902 -6.120905388 0.007332086 2.78018E-06 0.992670694
-----------------------------------------------------------------------------For ln-transformed data----------
WNL's result is
Power of ANOVA for Confidence Level 90.00
Power at 20% = 0.97882724My result is
Power Analysis
-----------------------------------------------------------------------------
Parameter T stat 1 T stat 2 P value 1 P value 2 Power
-----------------------------------------------------------------------------
ln(AUC) -2.168810573 -5.618247059 0.02116397 8.44382E-06 0.978844474
-----------------------------------------------------------------------------also same at level of 0.0001. Seems my formula works as WNL's inside hiden formula.
The level of the precision is interestingly same to that of BE's 90% CI.
go on validating WNL's and my formula using full data
Original data, use 93% CI, 20% diff to detect
WNL's power = 0.99020906
My power = 0.990218894
Original data, use 93% CI, 13% diff to detect
WNL's power = 0.82872712
My power = 0.828839088
Ln-transed data, use 93% CI, 20% diff to detect
WNL's power = 0.97637515
My power = 0.976397234
Ln-transed data, use 93% CI, 13% diff to detect
WNL's power = 0.72133308
My power = 0.721483927All the results are satisfactory. Seems I have touched into WNL's black box.
Edit: Merged two posts. [Helmut]
Complete thread:
- The secret of WinNonlin's BE's post-hoc Power analysis yicaoting 2011-11-08 14:40
![[Mix]](https://static.bebac.at/img/mix.png "open in Mix view")
- The secret of WinNonlin's BE's post-hoc Power analysisyicaoting 2011-11-08 14:55
- The secret of WinNonlin's BE's post-hoc Power analysis Helmut 2011-11-08 15:06
- Power of superiority d_labes 2011-11-10 09:49
- Power of superiority Helmut 2011-11-10 14:39
- Power of equality d_labes 2011-11-10 16:42
- Power of equality Helmut 2011-11-10 17:00
- Power of equality test - 2nd try for the logs d_labes 2011-11-11 09:55
- Crtl-C/Ctrl-V Helmut 2011-11-11 14:39
- Power of equality test - 2nd try for the logs d_labes 2011-11-11 09:55
- Power of equality yicaoting 2011-11-12 07:05
- Power of equality Helmut 2011-11-10 17:00
- Power of equality d_labes 2011-11-10 16:42
- Power of superiority Helmut 2011-11-10 14:39
Ing. Helmut Schütz