chi-squared problems in Systat

#1
Hi all, I went through college and grad school only using SAS and now at my new job, all we have is a old version of Systat 8. I put a post yesterday on the stats homework help about what test to use and I thought chi square and it was agreed. So now I am trying to run it and am having trouble. The variables are:
-Year (2001 and 2006)
-Transect (37 transects)
-Plot (50 points along transect where the nearest shrub measured, this was done for 3 shrub types so there is 3 species) Between the two years, they may not be the same plants nor the same species observed at each plot.
-Form class (the amount or severity that the shrub was eaten by deer and elk) This is in classes- 1-lightly browsed, 2-moderately browsed, and 3-severely browsed.

I want to test the significance of the increase or decreases of browsing. Lightly browsed shrubs decreased greatly and severely browsed increased greatly.

I am having trouble setting up a Chi-squared in systat and need some help.

I am assumming that the chi-squared in systat I want would be the 2 sample KS-test?

I have been bringing the data in, choosing data by groups, and using species as my grouping variable since I want to analyze each shrub species seperate.

I then go to 2 sample KS test,
Putting form class as my variable, year as my grouping variable.

But that seems to just analyze the difference of the numbers between the 2 years, not the classes themselves. I need to know whether there was a significant difference in the number of class 1's, class 2's, and class 3's.
Am I grouping something wrong??


Thank you so much for the help.
Russ
 

JohnM

TS Contributor
#2
Try coding the variables like this - for each case, it will be either 1-lightly browsed, 2-moderately browsed, or 3-severely browsed, so for example, if it's lightly browsed:

lightly browsed-----mod browsed-----severely browsed
-----Yes-----------------No----------------No-------

if it's moderately browsed:

lightly browsed-----mod browsed-----severely browsed
-----No-----------------Yes----------------No-------

if it's severely browsed:

lightly browsed-----mod browsed-----severely browsed
-----No-----------------No----------------Yes-------