graph of est. marginal means ± 95% in ANCOVA?

[Ian Martin-2]

I've been doing some ANCOVA to adjust the dependent for an age
covariate.  I'm getting some significant effects of the independent
on the (adjusted) means of the groups, but would like to plot these
means and associated error bars.  GLM options gives the adjusted
means and 95% CI, but only plots the means.

Any suggestions would be most appreciated.

best regards,
Ian

Ian D. Martin, Ph.D.

Tsuji Laboratory
University of Waterloo
Dept. of Environment & Resource Studies

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Bruce Weaver]

Ian Martin-2 wrote

I've been doing some ANCOVA to adjust the dependent for an age
covariate.  I'm getting some significant effects of the independent
on the (adjusted) means of the groups, but would like to plot these
means and associated error bars.  GLM options gives the adjusted
means and 95% CI, but only plots the means.

Any suggestions would be most appreciated.

best regards,
Ian

You could use a Hi-Lo plot.  The syntax will be something like:

GRAPH HILO(SIMPLE)=VALUE( upper lower mean ) BY caselbl  .

This requires a data file (or data set) that has Group, Mean, Lower & Upper, with a row of data for each group.  If this is a one-off thing, you may wish to create that file manually.  If you want it to be more automated, you should be able to send the desired table of means and confidence intervals out to a data set via OMS.