Brandon and Mariska,
I'm afraid I cannot give a good ref. It is just my personnal
experience that the results of RFX analyses with "small" groups can
really change with a few participants more/less (especially if you're
looking for subtle effects), and thus I infer - perhaps erroneously -
that this may affect comparisons across unequal groups. Sorry I
cannot be more precise. Do the real SPM experts have an idea on this
issue?
NB. Pragmatically speaking, I would suggest Mariska to first run the
analyses with equal groups. If she gets something significant and
meaningful, there is probably no reason to add more participants in
one group...
Mauro.
>Hello Mauro,
>
>This issue recently came up in a lab meeting and I've never read about
>inequality in any discussion of fMRI statistics. Perhaps I need to be looking
>at individual papers rather than book chapters. Can you recommend any good
>references (books or papers) for inequality?
>
>Thank You,
>Brandon Abbs
>
>--------------------------------
>Brandon Abbs, Ph.D.
>Postdoctoral Research Fellow
>Harvard Medical School
>Brigham and Women's Hospital
>1620 Tremont Street, BC-3-34 DWH
>Boston, MA 02120
>Phone: 617-525-8641
>Fax: 617-525-7900
>E-Mail: [log in to unmask]
>--------------------------------
>-----Original Message-----
>From: SPM (Statistical Parametric Mapping)
>[mailto:[log in to unmask]] On Behalf
>Of PESENTI Mauro
>Sent: Friday, October 17, 2008 8:54 AM
>To: [log in to unmask]
>Subject: Re: [SPM] compare between experiments
>
>LLN, le 17/10/08
>
>Dear Mariska,
>
>For behavioral results, the Anova is generally not much affected by
>the inequality of groups. For functional results, I'm not sure it's
>the case. Personally, I wouldn't trust too much a difference stemming
>from unequal groups, but I guess a small inequality should be OK if
>you have large groups (e.g., more than 15 participants). With small
>groups (< 15), adding/dropping 2-3 participants can significantly
>affect your RFX analyses.
>
>Hope this helps,
>
>Mauro.
>
>
>>Hi Mauro,
>>
>>
>>thanks for your reply!
>>
>>One more question:
>>Do I need to have an equal amount of males/females in both
>>experiments or doesn't it matter if the groups don't differ too much?
>>
>>
>>greetings,
>>Mariska
>>PS This is a statistically valid method?
>>
>>
>>
>>2008/10/15 PESENTI Mauro
>><<mailto:[log in to unmask]>[log in to unmask]>
>>
>>LLN, le 15/10/08
>>
>>Dear Mariska,
>>
>>Here is what I suggest:
>>
>>(1) Compute your contrast of interest at the first level for each
>>subject individually,
>>
>>(2) port the corresponding con_* files to 2-level analyses using:
>>
>>- 1-sample t-tests with a contrast set at [1] for each group separately;
>>
>>- 2-sample t-tests to compare your two groups. Select all your con_*
>>files for one groupe (say females), then all your con_* files for
>>the other group (say males); set your contrasts depending on what
>>you're looking at: ([1 -1] gives group 1 > group 2, [-1 1] gives
>>group 2 > group 1).
>>
>>Hope this helps,
>>
>>Mauro
>>
>>.
>>
>>Hi all,
>>
>>
>>Does anybody know how I can compare males vs females on one contrast
>>from two experiments with different subjects (containing both males
>>and females)?
>>
>>
>>Greetings,
>>Mariska
>>
>>
>>
>>--
>>_____________________________________
>>
>>!!! NOUVELLE ADRESSE MAIL -- NEW E-MAIL ADDRESS !!!
>> <mailto:[log in to unmask]>[log in to unmask]
>>_____________________________________
>>
>>Help fighting hunger: <http://www.hungersite.com>http://www.hungersite.com
>>
>>Just click your mouse and sponsors of The Hunger Site donate a
>>serving of food to a person in need - at no cost to you.
>>
>>______________________________________
>>
>>Mauro PESENTI
>>Research Associate, National Fund for Scientific Research (Belgium)
>>Unite de Neurosciences Cognitives
>>Departement de Psychologie
>>Universite Catholique de Louvain
>>Place Cardinal Mercier, 10
>>B-1348 Louvain-la-Neuve
>>
>>tel.: +32 (0)10 47 88 22
>>fax: +32 (0)10 47 37 74
>>E-mail: <mailto:[log in to unmask]>[log in to unmask]
> ><http://www.nesc.ucl.ac.be>http://www.nesc.ucl.ac.be
>><http://www.nesc.ucl.ac.be/mp/pesentiHomepage.htm>http://www.nesc.ucl.ac.be/mp/
>pesentiHomepage.htm
>
>
>--
>_____________________________________
>
>!!! NOUVELLE ADRESSE MAIL -- NEW E-MAIL ADDRESS !!!
>
> [log in to unmask]
>_____________________________________
>
>Help fighting hunger: http://www.hungersite.com
>
>Just click your mouse and sponsors of The Hunger Site donate a
>serving of food to a person in need - at no cost to you.
>
>______________________________________
>
>Mauro PESENTI
>Research Associate, National Fund for Scientific Research (Belgium)
>Unite de Neurosciences Cognitives
>Departement de Psychologie
>Universite Catholique de Louvain
>Place Cardinal Mercier, 10
>B-1348 Louvain-la-Neuve
>
>tel.: +32 (0)10 47 88 22
>fax: +32 (0)10 47 37 74
>E-mail: [log in to unmask]
>http://www.nesc.ucl.ac.be
>http://www.nesc.ucl.ac.be/mp/pesentiHomepage.htm
>
>The information transmitted in this electronic communication is intended only
>for the person or entity to whom it is addressed and may contain confidential
>and/or privileged material. Any review, retransmission, dissemination or other
>use of or taking of any action in reliance upon this information by persons or
>entities other than the intended recipient is prohibited. If you received this
>information in error, please contact the Compliance HelpLine at
>800-856-1983 and
>properly dispose of this information.
--
_____________________________________
!!! NOUVELLE ADRESSE MAIL -- NEW E-MAIL ADDRESS !!!
[log in to unmask]
_____________________________________
Help fighting hunger: http://www.hungersite.com
Just click your mouse and sponsors of The Hunger Site donate a
serving of food to a person in need - at no cost to you.
______________________________________
Mauro PESENTI
Research Associate, National Fund for Scientific Research (Belgium)
Unite de Neurosciences Cognitives
Departement de Psychologie
Universite Catholique de Louvain
Place Cardinal Mercier, 10
B-1348 Louvain-la-Neuve
tel.: +32 (0)10 47 88 22
fax: +32 (0)10 47 37 74
E-mail: [log in to unmask]
http://www.nesc.ucl.ac.be
http://www.nesc.ucl.ac.be/mp/pesentiHomepage.htm
|