Log in
Register|
jag009 ★★★ NJ, 2012-02-22 23:19 (4830 d 15:33 ago) Posting: # 8158 Views: 4,486 |
|
|
Hello, Would it be possible if someone could tell me whether it's possible to compute the 95% upper confidence bound in Excel for a 3-way partial replicate study? If so, how? In a nutshell, my laptop is currently down (in the shop) and I don't have my SAS routine to do the computation. I have the following data from a three way replicate: Mean of ln transformed Test Cmax = 7.396 Mean of ln transformed Ref Cmax = 7.247 SWR (Within Subject Variation) = .403 Sample size is 58 Thank you. Jay |
|
d_labes ★★★ Berlin, Germany, 2012-02-23 10:12 (4830 d 04:39 ago) @ jag009 Posting: # 8163 Views: 3,623 |
|
|
Dear Jay, You need additionally to your givings the SD (standard deviation) of the difference µT - µR, calculated stratified over the sequences. This is from the SAS step of calculating an ANOVA (Proc GLM) with sequence as factor and the differences ilat as left hand variable in the model as described in the Progesterone guidance, current Feb 2011 version. But for missing code lines see here.Then it should not so hard to implement the formulas for the 95% CI of the scaled ABE criterion given f.i. here in M$ Excel. But I wouldn't recommend that remembering Martins "Do not": "Never never never never use Excel. Not even for calculation of arithmetic means."  BTW: How will you use Excel with your laptop not available?  — Regards, Detlew |
|
jag009 ★★★ NJ, 2012-02-23 16:11 (4829 d 22:41 ago) @ d_labes Posting: # 8166 Views: 3,478 |
|
|
Thank you. Well my laptop is the office laptop which has the softwares. In the mean time I am using my home computer which only has Excel  Maybe I will go and download "R" and try it out... Jay |
Ing. Helmut Schütz