Data Analysis, Simple Experiment

How do you want the correlation to be determined?

If the correlation is less than .95 and it is known from experience to within .05 or .10, there should be no problem. Alternatively, if the correlation is less than .95 and the number of replicates per treatment is at least 6, the correlation can be successfully estimated from the data. If neither of these conditions holds, then a paired tight t or an analysis of covariance approach should be taken rather than a pooled tight t approach.

For further information, please contact Steve Verrill at or 608-231-9375.
Last modified on 11/23/95.