Multiple Linear Regression using Dummy Variables

#1
Hi,

I need to do some analysis for my undergrad dissertation but I am completely stuck.

I have 7 forest research plots that have had various thinning treatments and I need to know if the treatment makes any difference to germination and growth of seedlings..

6 of the plots are paired plots of the same treatment, and the 7th is a single plot giving 4 treatments in all.

However, not all of the plots are the same. It is known that the density of mature trees is significant in how well seeds germinate and grow, but the density figure is significantly different for each plot (tested with ANOVA).

I need a way to compare treatments, but remove the effect of density. I have had a method describe to me whereby the treatments are replaced by dummy variables in the following manner:-

Take 3 columns, and if it is treatment 1, all columns have '0', if it is treatment 2 only column 1 has '1', if it is treatment 3 only column 2 has a '1' and if it is treatment 4 only column 4 has '1'. (see attached spreadsheet)

Still with me?

This then gives me 'number of seedlings' as the dependent variable, and column 1, column 2, column 3 and density as independent variable.

I am using SPSS (although I do have Minitab at my disposal) but I cannot work out what to put where, or what to look for in the output.

Any help would be much appreciated