sdm-help-list > Meaning of contrast maps? (single population)
Showing 1-3 of 3 posts
May 31, 2016 04:05 AM | Nobody
Meaning of contrast maps? (single population)
Meaning of contrast maps? (in the context of single population
functional data meta-analysis)
For 2 groups of studies comparison, it seems intuitive:
1 = meta-analytic map of all studies with value 1 in column
0 = meta-analytic map of all studies with value 0 in column
1m0 = actual contrast between the two maps
For 3 groups, it is less intuitive for the encoding but here is my understanding:
Filter1: value 1 = meta-analytic map of all studies with value 1 in column "Filter1" indicating studies belonging to group/condition n°2. Value 0 = meta-analytic map of all studies with value 0 in column "Filter1" indicating studies NOT belonging to group/condition n°2 (i.e., group/condition n°1 & n°3).
Filter2: value 1 = meta-analytic map of all studies with value 1 in column "Filter2" indicating studies belonging to group/condition n°3. Value 0 = meta-analytic map of all studies with value 0 in column "Filter2" indicating studies NOT belonging to group/condition n°3 (i.e., group/condition n°1 & n°2).
Then, "10m01" means difference between Group/Condition n°2 versus n°3.
And "01m00" means difference between Group/Condition n°3 versus n°1.
And "10m00" means difference between Group/Condition n°2 versus n°1.
Is this correct ?
Thanks.
Henryk
For 2 groups of studies comparison, it seems intuitive:
1 = meta-analytic map of all studies with value 1 in column
0 = meta-analytic map of all studies with value 0 in column
1m0 = actual contrast between the two maps
For 3 groups, it is less intuitive for the encoding but here is my understanding:
Filter1: value 1 = meta-analytic map of all studies with value 1 in column "Filter1" indicating studies belonging to group/condition n°2. Value 0 = meta-analytic map of all studies with value 0 in column "Filter1" indicating studies NOT belonging to group/condition n°2 (i.e., group/condition n°1 & n°3).
Filter2: value 1 = meta-analytic map of all studies with value 1 in column "Filter2" indicating studies belonging to group/condition n°3. Value 0 = meta-analytic map of all studies with value 0 in column "Filter2" indicating studies NOT belonging to group/condition n°3 (i.e., group/condition n°1 & n°2).
Then, "10m01" means difference between Group/Condition n°2 versus n°3.
And "01m00" means difference between Group/Condition n°3 versus n°1.
And "10m00" means difference between Group/Condition n°2 versus n°1.
Is this correct ?
Thanks.
Henryk
Jul 6, 2016 12:07 PM | Joaquim Radua
RE: Meaning of contrast maps? (single population)
Dear Henryk
I think the filters and contrasts are correct.
To make it easier, note that you have coded the groups as follows:
00: group 1
10: group 2
01: group 3
and thus:
10m00: difference between group 2 and group 1
01m00: difference between group 3 and group 1
10m01: difference between group 2 and group 3
Best,
Joaquim
I think the filters and contrasts are correct.
To make it easier, note that you have coded the groups as follows:
00: group 1
10: group 2
01: group 3
and thus:
10m00: difference between group 2 and group 1
01m00: difference between group 3 and group 1
10m01: difference between group 2 and group 3
Best,
Joaquim
Jul 6, 2016 12:07 PM | Henryk Bukowski - University of Vienna
RE: Meaning of contrast maps? (single population)
Thanks for clarifying !
