Formula for linear mixed effect model

Hi Gang,

In my study, I used 3dLME to run linear mixed-effect models on functional connectivity data. In the model, the factor condition is the fixed effect, and the factor react is the random effect. Is a repeated measures design.

I was wondering if

[b]B[sub]ij[/sub] = a[sub]0j[/sub] + a[sub]1j[/sub]*X[sub]ij[/sub] + g[sub]0i[/sub] + g[sub]1i[/sub]*X[sub]ij[/sub] + e[sub]ij[/sub][/b]

is the correct expression to cite.

Here’s my 3dLME code:

3dLME \
-prefix LME_$seed\_glt6Up \
-jobs 4 \
-model "condition*react" \
-mask ../../groupmask/NoiseMaskcol_TN+tlrc \
-qVars "react" \
-qVarCenters '0' \
-ranEff '~1+react' \
-SS_type 3 \
-num_glt 7 \
-dataTable \
Subj	react		condition	InputFile \
s1	55.4636		neutral		../../subjF1/Decon_PPI_bis/bucket_T_ppi_$seed\+tlrc'[132]' \
s17	104.3677	negative	../../subjF20/Decon_PPI_bis/bucket_T_ppi_$seed\+tlrc'[136]'

Thanks, Simone

Simone, how many levels does the factor “condition” have, 2 or 3?

2 levels


The model looks good. You may want to mention that indices i and j code for subject and condition, respectively.

Of course. Thanks.