Hi Bo, I don't understand the contrast you are trying to make. Are you really trying to compute the interaction between four variables (factor 1, factor 2, gender, and age)? doug
On 11/21/2012 10:49 AM, xiangbo_2010 wrote: > Dear Donald McLaren > Thank you for your reply! I make the contrast according to your method > is following，but I want to make interaction between factor 1 (A,B)and > factor 2(C,D,E), gender (M,F) and one continuous variable (age) as > covariates, the following contrast: > 2 -2 0 -2 2 0 2 -2 0 -2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 > 0 2 -2 0 -2 2 0 2 -2 0 -2 2 0 0 0 0 0 0 0 0 0 0 0 0 > > is correct? Thanks! > Bo Xiang > > > > > > At 2012-11-21 04:07:37,"MCLAREN, Donald"<mclaren.don...@gmail.com > <mailto:mclaren.don...@gmail.com>> wrote: > >On Tue, Nov 20, 2012 at 12:56 PM, Douglas N Greve > ><gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu>> wrote: > >> Thanks Donald. Is this the standard way to do this? I had used 8 rows > >> instead of 4 with the difference being that 8 rows gives you an > >> opportunity > >> to look for an effect in males OR females. > > > >Yes. Having 8 rows would tell you if you have an interaction between > >factor 1 and 2 in either males or females. My 4 rows only tell you if > >the interaction exists. Technically speaking, one would run the > >three-way interaction first. If nothing existed then you do the > >two-way interaction as I suggested. If there is a three-way > >interaction, then you would use Doug's approach of the interaction in > >either males or females. > > > >If there is an effect in both > >> males and females but the effects go in opposite directions, then the 4 > >> row > >> implementation will resolve to 0 (no effect). Or am I misunderstanding > >> something (again:)? > > > >Nope. You are right. If the male and female effects are different, > >then they could cancel each other out. If you suspect this to be the > >case, then you should be able to demonstrate a three-way interaction. > > > >> thanks! > >> doug > >> > >> > >> On 11/20/2012 01:50 PM, MCLAREN, Donald wrote: > >>> > >>> Bo, > >>> > >>> Doug asked me to chime in on your issue. Here are some points that you > >>> (and others) will hopefully find useful. > >>> > >>> (1) Inferences are two-step process. First, you create and estimate > >>> the design matrix. Every column in the design matrix accounts can > >>> account for some of the variance in the data. Second, you have > >>> contrasts that allow you to infer specific effects. Because the model > >>> contains your covariates, you are always controlling for the > >>> covariates and by extension any factor/covariate not in the contrast. > >>> > >>> (2) Forming contrasts is often the most difficult thing to do. I > >>> assume that your three factors (1, 2, and gender) are all > >>> between-subject factors. If one of them is a within-subject factor > >>> please let me know and disregard the rest of the email. The final > >>> F-contrast will have 4 rows (factor 1 levels-1)*(factor 2 levels > >>> -1)=(3-1)*(3-1)=2*2=4 > >>> > >>> The following is an outline for creating contrasts: > >>> (a) Start simple - difference between levels of 1 factor > >>> (b) Define your null hypothesis: AO=AP=AQ > >>> (c) Make it equal to 0: AO-AP=0 AND AP-AQ=0 > >>> (d) Repeat for the other levels of the factor... > >>> BO-BP=0 AND BP-BQ=0 > >>> CO-CP=0 AND CP-CQ=0 > >>> > >>> (e) Now combine them AO-AP=BO-BP=CO-CP AND AP-AQ=BP-BQ=CP-CQ > >>> > >>> (f) Make them equal to 0: > >>> AO-AP-BO+BP=0 > >>> BO-BP-CO+CP=0 > >>> AP-AQ-BP+BQ=0 > >>> BP-BQ-CP+CQ=0 > >>> > >>> (g) Expand them to include gender, for example: > >>> AO-AP-BO+BP=0 becomes FAO-FAP-FBO+FBP+MAO-MAP-MBO+MBP=0 > >>> > >>> Since the contrast now has 2 columns per level, you should divide all > >>> values by 2. This will produce the correct amplitude and statistics. > >>> If you leave the values as 1 and -1, then you will have an incorrect > >>> amplitude, but the statistics will still be correct. > >>> > >>> (h) Fill in the respective columns of your design matrix. > >>> > >>> (3) The degrees of freedom are defined based on the rows of the > >>> F-matrix and the number of rows in the design matrix. The F-test has a > >>> numerator and denominator degrees of freedom. F(n,d). > >>> > >>> 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 Tue, Nov 20, 2012 at 8:17 AM, xiangbo_2010<xiangbo_2...@126.com > >>> <mailto:xiangbo_2...@126.com>> > >>> wrote: > >>>> > >>>> Dear doug > >>>> Thank you for your reply! > >>>> +AOM -BOM -APM -BPM 0 0 0 0 0 > >>>> +AOM -BOM 0 0 -AQM +BQM 0 0 0 > >>>> +AOM 0 -APM 0 0 0 -COM +CPM 0 > >>>> +AOM 0 0 0 -AQM -COM 0 +CQM > >>>> there should be use 1 -1 or 0.5 -0.5? whether the -BPM should be change > >>>> BPM? > >>>> Thanks > >>>> > >>>> Bo Xiang > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> At 2012-11-19 07:23:31,"Douglas Greve"<gr...@nmr.mgh.harvard.edu > >>>> <mailto:gr...@nmr.mgh.harvard.edu>> wrote: > >>>> > >>>> > >>>> Hi Bo, you can think of the Ftest as a logical 'OR' between the t-test > >>>> contrasts indicated in each row. Each row is a difference of > >>>> differences, > >>>> so > >>>> > >>>> 1. (A-B)om - (A-B)pm --> Does the difference between A and B differ > >>>> between > >>>> groups O and P for Males? > >>>> 2. (A-B)om - (A-B)qm > >>>> 3. (A-C)om - (A-C)pm > >>>> 4. (A-C)om - (A-C)qm > >>>> 5. (A-B)of - (A-B)pf --> Does the difference between A and B differ > >>>> between > >>>> groups O and P for Females? > >>>> 6. (A-B)of - (A-B)qf > >>>> 7. (A-C)of - (A-C)pf > >>>> 8. (A-C)of - (A-C)pf > >>>> > >>>> I've put together the first 9 columns of the first 4 rows. The last 9 > >>>> columns are all 0s. For the last for rows, the 0s and below matrix are > >>>> swapped to give you the same for the females > >>>> > >>>> doug > >>>> > >>>> AOM BOM APM BPM AQM BQM COM CPM CQM > >>>> > >>>> > >>>> -------------------------------------------------------------------------- > >>>> +AOM -BOM -APM -BPM 0 0 0 0 0 > >>>> +AOM -BOM 0 0 -AQM +BQM 0 0 0 > >>>> +AOM 0 -APM 0 0 0 -COM +CPM 0&! nbsp;&nb sp; > >>>> > >>>> +AOM 0 0 0 -AQM -COM 0 +CQM > >>>> > >>>> > >>>> > >>>> On 11/17/12 9:21 PM, xiangbo_2010 wrote: > >>>> > >>>> Hi Freesurfer experts, > >>>> > >>>> I'm very sorry to bother you, but I am very confused with the following > >>>> questions: > >>>> > >>>> My experimental design includes three discrete factors: factor 1 with > >>>> three > >>>> levels (A,B,C ); factor 2 with three levels (O,P,Q); gender (F, M), and > >>>> one > >>>> covariate. > >>>> > >>>> So I can get 18 classes: FAO, FAP,FAQ,FBO,FBP,FBQ,FCO,FCP,FCQ,MAO, > >>>> MAP,MAQ,MBO,MBP,MBQ,MCO,MCP,MCQ. I want to perform the interaction > >>>> between > >>>> factor 1 and factor 2 regressing out the effect of gender and one > >>>> covariate, > >>>> but I don't know the rules for setting the contrasts for the F-test. > >>>> The > >>>> contrast matrix I used is: > >>>> > >>>> 1 1 1 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 1 -1 0 1 -1 0 1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 1 1 1 0 0 0 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 1 0 -1 1 0 -1 1 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> 0 > >>>> > >>>> 0 0 0 0 0 0 0 0 0 1 1 1 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 0 0 0 0 0 0 0 0 0 1 -1 0 1 -1 0 1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> 0 0 0 0 0 0 0 0 0 1 0 -1 1 0 -1 1 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > >>>> 0 > >>>> > >>>> is it correct? > >>>> > >>>> > >>>> > >>>> Any help will be very appreciated. > >>>> > >>>> > >>>> > >>>> Best wishes, > >>>> > >>>> > >>>> > >>>> Bo Xiang > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> _______________________________________________ > >>>> Freesurfer mailing list > >>>> Freesurfer@nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu> > >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> _______________________________________________ > >>>> Freesurfer mailing list > >>>> Freesurfer@nmr.mgh.harvard.edu <mailto: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. > >>>> > >>> > >> > >> -- > >> Douglas N. Greve, Ph.D. > >> MGH-NMR Center > >> gr...@nmr.mgh.harvard.edu <mailto:gr...@nmr.mgh.harvard.edu> > >> Phone Number: 617-724-2358 > >> Fax: 617-726-7422 > >> > >> Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > >> FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html > >> Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ > >> > >_______________________________________________ > >Freesurfer mailing list > >Freesurfer@nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu> > >https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > > > > > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
