Mahinda, As I understand it, you have 3 groups (Controls, P1, and P4) in your analysis and have done three contrasts: (1) P1vP4 --> no differences (2) P1vControls --> no differences (3) P4vControls --> significant differences
Your first question about the difference in the analysis --> three contrasts are all testing different relationships and thus will give you different results Your second question about whether these are valid --> all three contrasts are valid HOWEVER, from your wording, it seems that your interpretation is wrong. P1 and P4 are not different. The only test that you can do to test if P1 and P4 are different is to test them directly. Observing that P1vControls had no differences and P4vControls had a difference does provide any evidence that P1 and P4 are different. Here is a quick example. The threshold we will use is 3. P1vControls is 2.99. P4vControls is 3.01. There is virtually no difference in P1 and P4, even though only one group is different than controls. While this is an extreme example, it holds that the lack of a difference in one group, but not the other does not mean the groups are different. Here is a paper on the issue and its prevalence in Neuroscience: Nieuwenhuis et al. (2011). Erroneous analysis of interactions in neuroscience: a problem of significance. Nature Neuroscience. I suspect that with larger samples, that you should be able to see the difference, if the difference exists. You might consider the contrast [0 1.5 .5 -.5 -1.5] across all 5 groups showing a linear effect of disease severity. The first column being controls. Hope this helps. Best Regards, Donald McLaren ================= D.G. McLaren, Ph.D. Research Fellow, Department of Neurology, Massachusetts General Hospital and Harvard Medical School Postdoctoral Research Fellow, GRECC, Bedford VA Website: http://www.martinos.org/~mclaren Office: (773) 406-2464 ===================== This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is intended only for the use of the individual or entity named above. If the reader of the e-mail is not the intended recipient or the employee or agent responsible for delivering it to the intended recipient, you are hereby notified that you are in possession of confidential and privileged information. Any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited and may be unlawful. If you have received this e-mail unintentionally, please immediately notify the sender via telephone at (773) 406-2464 or email. On Mon, Nov 26, 2012 at 11:59 AM, Mahinda Yogarajah <y.mahi...@gmail.com> wrote: > Hi Doug and Donald, > > Again many thanks for the advice. With regard to the analysis described > previously - I have n=21 in my control group while P1=4 patients, P2=7 > patients, P3=3 patients and P4=5 patients. Given that i have small numbers I > elected to compare only 2 patient groups (p1 and p4) rather than all 4. > > My aim was to show that P4 patients (higher level of endogenous factor) are > more affected (eg smaller area and greater thickness-age decline compared to > controls) than P1. Given that as Donald points out that (c-p1) - (c-p4) is > like comparing p1 and p4 directly, I did the latter but there is nothing > significant. > > However when i compare p1 to controls and then p4 to controls separately i > see nothing signficant in the p1 analysis but significant results in the p4 > analysis - this makes sense as the p4 gp are more affected than the very mild > p1 gp. > > My questions are: > > 1) why the difference in results between the two analyses (? due to larger > overall group in second analysis) > 2) is the latter analysis valid ? > > Thanks for your help. > > Mahinda > > _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
