# PharmPK Discussion - Statistics software applications

PharmPK Discussion List Archive Index page
• On 28 Apr 2000 at 12:23:56, "Meyer Katzper 301-827-2514 FAX 301-827-2531" (KATZPER.at.cder.fda.gov) sent the message
`The hypothetical data with equal means presented by Reeve,Treat 1    Treat 250         7070         8090         90110       100130       110can be subject to an entirely different interpretation. The means areidentical but the standard deviations are very different. (28.3 vs 14.1)Imagine that there is an optimal value for some physiological variableand it is (in this hypothetical case) 90. If we look upon the data aspaired outcomes from 2 treatments then treatment 2 is clearly betterthan treatment 1 in getting all subjects closer to the optimal value.However, any means test will show the means not significantly differentfrom each other. An F-Test of Two-Samples for Variances (as implementedin Excel) will also not show significant difference for the variance asthe sample size is too small. Doubling each observation will yield asignificant result. The point is that your knowledge and understandingof the data should be primary and statistical tests secondary (unlessyou only know statistics).=========================================Meyer Katzper, Ph.D.FDA/CDER/ODEV/DAAODPkatzper.aaa.cder.fda.govTel (301) 827-2563`
Back to the Top

• On 28 Apr 2000 at 22:22:24, jw5886.aaa.cnsvax.albany.edu sent the message
`Knowledge from 5 data points convince you that trt 2 is better than trt1? Why is it better than an appropriate statistical test?  I thinkstatistics should test if what we know about data is correct, the less weassume the less bias will be introduced.  Having said, clinical judgementof the investigators is equally important.  So it will be nice if we finda balance.Of course looking at the data provided, you can argue due to the size ofvariability, there is not enough power to detect a difference. But thatdoesn't necessarily mean a second sample of 5 in each  treatment groupwill replicate the first experiment exactly.  After all, before a trial,you don't know what is going to happen, maybe trt 2 will be even worse.Peace.`
Back to the Top

• On 1 May 2000 at 20:02:01, "Nils Ove Hoem" (n.o.hoem.-a-.farmasi.uio.no) sent the message
`Regarding statistical testing;To me it seems that people tend to forget what statistics is all about. Whensomeone states that "Statistics should test if what we know is correct" Ithink that the matter has gotten way ot of line.Statistics deals with stocastic variables and has in its essence nothing todo with "truth" or proof of hypothesis.This is exactly why our 0. hypothesis usually states that two populationsare equal. What statistics can tell us somthing about is the chances thatsuch would be the case because of stocastics. So in my mind statistics is abase-level filter.  Its purpose is to rule out(at a certain level ofconfidence) that de differences we see can be ascribed to random variation.Now, with that possibility ruled out we can go ahead and ask ourselves whatmight then have caused the differences observed. What have to be kept inmind is that any non-random error (eg. non-stocastic) can of course causesuch observed differendes between populations. So again, statistics do nottell us any truth, only that what we see probably is not caused by stocasticvariation. And of course if we can not rule out that possibility, then anyfurther discussion is a vaste of time.N.O Hoem`
Back to the Top

• On 2 May 2000 at 21:38:54, "R.A. Fisher" (ra_fisher.-a-.hotmail.com) sent the message
`Hi All,I have some comments on a recent email posted under the above topic.Reeve proposed the hypothetical dataset:tmt1  tmt250    7070    8090    90110   100130   110Meyer Katzper commented:i) The standard deviations are 28.3 and 14.1.ii) Regarding the F ratio test, "doubling each observation will yielda significant result".iii) Treatment 2 is "clearly better" regarding hitting a target of 90 ?Unfortunately,i) is incorrect.  This is surely a 'sample' drawn from all possiblesubjects.  Hence the 'sample' standard deviations should becalculated (31.6 and 15.8 respectively).ii) is incorrect.  From my days at school, the F ratio = (sample var1/ sample var2).  If I double each observation, var1 becomes 4*var,and var2 becomes 4*var2.  Hence the F ratio, if used, is (4*var1) /(4*var2).  The 4's cancel.  This makes sense.iii) What a confident statement!  Aren't pharmaceutical companiessilly spending lots of money when N=5 could be sufficient for a NDAto demonstrate 'clearly better'.  Faced with this question:Researcher 1 (bright) says, "I can see tmt2 is more variable", andperhaps additionally calculates a correct SD for each sample.Researcher 2 (brighter) does the same as Researcher 1, but considersand uses appropriate statistic methods, because he/she knows thatthese are only samples (which are subject to random error), andwishes to generalise the result to the population as a whole.Researcher 3 (brightest) does the same as Researcher 2, but asksthemselves the question afterwards "How did I end up with data whichis unable to clearly answer my question?", and vows not to do thesame again.Researcher 4 (new role model for Researcher 3) always takes the easyoption.  She/he prefers not to guess.  She/hei) thinks through the question they wish to answer.ii) Prospectively plans what data, how much data, AND what analysisthey will carry out.iii) Does ii).iv) Gets results which are clear.  Perhaps not positive, but clear.I like Researcher 4, although perhaps I 'only know statistics'.RA Fisher`
Back to the Top

• On 25 May 2000 at 20:51:23, prashant bodhe (prashnvb.aaa.dr.com) sent the message
`Dear FiscerI have used book on staistical tables extensively.Which is the latest edition now available. In case you have the table in thesoft form would you be able to send those to meDr. prashant`
Back to the Top

Want to post a follow-up message on this topic? If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Statistics software applications" as the subject

Copyright 1995-2010 David W. A. Bourne (david@boomer.org)