Dear Aaron,
At 12:44 25/05/99 -0400, Aaron Field wrote:
| In the SPM99b contrast manager, when I specify an F-contrast of [1 0] to
| test a single effect of interest (a boxcar), the contrast is indicated
| above the design matrix with the color white indicating "one" (above the
| boxcar column) and gray indicating "zero" (above the constant term).
| However, when I specify the very same contrast in the t-contrasts
| window, the gray/white indicators are reversed. If this is not a bug, it
| is a very confusing feature! Please advise, and thank you.
It's a confusing feature!
Since F-contrasts can consist of multiple constraints (multiple linear
contrasts), they are depicted as images. This further differentiates one
dimensional F-contrasts from t-contrasts, which are displayed as
bar-charts.
The design matrix and F-contrast images map zero to mid-grey, with white
taking the maximum value (usually 1 in F-contrasts). With the old contrast
bar-plots, the bars are grey and the background white, so no bar looks like
a white square.
You can (just) differentiate between the two, since the bar-plots only have
X & Y axes, whilst the F-contrast images are surrounded by a box.
Additional contrast exploration and presentation facilities are planned for
the full release of SPM99. (For other planned features, see the SPM99 web
page.)
Hope this helps,
-andrew
+- Dr Andrew Holmes ------------------ mailto:[log in to unmask] -+
| ___ __ ___ Robertson Centre for Biostatistics |
| ( ,) / _)( ,) Boyd Orr Building, |
| ) \( (_ ) ,\ University Avenue, |
| (_)\_)\__)(___/ Glasgow. G12 8QQ Scotland, UK. |
+----------------------------------------- http://www.rcb.gla.ac.uk/ -+
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|