Dear Noelia,
In general, the contrast estimate is simply given by c' *b, where c is the weight vector or matrix and b is the beta estimates. This means the rows of your contrast define the number bars in your plot.
I hope this answers your question!
Best wishes
Martin
> On 19 Jan 2016, at 18:24, Noelia Martinez Molina <[log in to unmask]> wrote:
>
> Dear SPM experts,
>
> In order to plot the contrast estimates of my 2nd Level, I open a particular contrast in SPM and go to a local peak maxima. Before so and in addition to this contrast, I have defined a contrast named "Effect of interest" (thereafter eoi) using an F contrast and writing the vectors' weights in the columns to reduced design. Since my design matrix has 6 columns, I specified the eoi contrast as 7:maximum. This is the contrast I will use to plot the contrast estimates.
>
> Next I choose Plot in the SPM window and Contrast estimates and 90% C.I. Then I select the eoi contrast and get the contrast estimates from that coordinate. So far so good.
>
> My question is: what mathematical operation applies SPM to the beta coefficients of this contrast to compute the contrast estimates displayed in this plot?
>
> I'd appreciate any help to being able to calculate the contrast estimates that SPM provides in the structure contrast of the matlab workspace using the beta values.
>
> Many thanks in advance for your assistance.
>
> Kind regards,
> Noelia Martínez-Molina
|