Dear Rajan,

For the question you are trying answer, you have modelled your data in the wrong way. You should include both conditions and model the difference between conditions as a modulation of the F or FB connections. You can find an example of this in the SPM manual. 

Best wishes
Martin

On 03 Sep 2015, at 10:10, Rajan Kashyap <[log in to unmask]> wrote:

Dear Martin

Thanks a lot for your suggestion. I would like to explain a bit more and take you suggestion in more detailed way.

 Through BMS i can see that for Condition A : Model F is winning over FB with Model exceedance probablity =.6
For condition B: FB wins over F with same model exceedance probablity.

Now i want to compare F model of condition A with F model of condition B. (This is because i find some interesting correlation of the non winning model with Reaction time).
I have some questions as i am new to SPM

(1)Can i compare a wiining model of one condition with the non winning model of another condition.

(2) Will BMA be better than BPA for this comparison. As i am performing T test on the model parameters for every connection in my model to see the effect.

Your reply will help me a lot.

Regards
Rajan
Phd Student
HKBU


On Thu, Sep 3, 2015 at 3:37 PM, Martin Dietz <[log in to unmask]> wrote:
Dear Rajan,

The idea behind model selection is to identify the model (hypothesis) with the highest model evidence among a set of alternative models. If there isn't one particular model that best explains your data, you should use Bayesian model averaging (BMA). This allows you to do inference on the connection strengths using an estimate from the entire model space.

Best wishes
Martin

> On 02 Sep 2015, at 10:44, Rajan Kashyap <[log in to unmask]> wrote:
>
> Dear SPM people
>
> I have ERP  data of two conditions (Suppose A and B). Based on the data i created two models F and FB.
> I fitted the data on individual subjects.
> I found on the group level :(1) F model was preferred over FB model in  condition A
>                                        (2) FB model over F for condition B.
>
> However, then i found that few participants who show a correlation to behavioural data (in my case faster reaction time)in condition A preferred FB model though F model was winning over there at group level.
>
> This motivated me to take the connection parameters (posterior means) of the non winning models too. Then i compared for significant connections between the F models  connection posterior means for 2 conditions and FB models posterior means for the two conditions, to see what variations in connections can cause such outcomes
>
> In general i always found in DCM papers that only winning models are preferred.
> In this regards i would like take the view of DCM experts, if this could be done.
>
> Regards
> Rajan
> PhD student