No, I want to compute the interaction between factor 1 and factor 2 , gender and age as covariates. thanks
Bo Xiang 在 2012-11-22 04:22:06，"Douglas N Greve" <gr...@nmr.mgh.harvard.edu> 写道： >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 _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
