Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 1 down vote favorite

I have a dataset with a categorical treatment (1 if treated, 0 if control) and many columns (34) of response variables. Each column represents a species and its response (some measured abundance) to the presence or absence of the treatment. Each row is a separate site; exactly half of them were treated and the other half were controls. The experiment was replicated over two years.

I am interested in finding out which of these species respond to the treatment, so a LDA was suggested, but I am confused because my treatment is categorical and all of the examples I can find of LDA use continuous predictor variables.

So the data look like:

Treatment(0/1) Site(ID#) Year(1/2) Species 1 Species 2 ... Species 34

Can I use a LDA for this question? If not, what sort of ordination is capable of accounting for categorical predictor variables?

share|improve this question
Hi Laura, welcome to the site! I don't think LDA is the right approach here; LDA wants continuous predictors and a categorical outcome. If I understand correctly, you want to know which species differ in abundance when the treatment is applied - which means you have categorical predictors and continuous outcomes! To me, that suggests something more like t-tests, at the most basic, or multilevel models/generalized estimating equations, if you want to get fancier and have explicit estimates of things like site and year effects. – Matt Parker Jun 29 '12 at 18:14

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.