[Histonet] Correlation studies

Rene J Buesa rjbuesa <@t> yahoo.com
Tue Jun 3 14:59:37 CDT 2008


Correlations are always between two sets of variables (unless they refer to multiple correlations where those sets contain more than two).
  I have never heard of correlating an IHC machine. What are you going to correlate it with?
  I do not imagine how you are going to present the problem. Are you trying to correlate your results with your IHC machine with the results in another or with some other results?
  If that is the case you will have to run both sets simultaneously.
  On the other hand those correlation tests (the strongest) have to have a normal distribution of the data, or you will have to use some type of transformation.
  Perhaps you could find more information in
  http://www.vassar.edu/lowry which is a fantastic web site to perform statistical tests.
  René J.

Clark Stephen - Myrtle Beach <Stephen.Clark1 <@t> hcahealthcare.com> wrote:
  I need to perform correlation studies for our new Immunohistochemistry
machine. The question I have is what format do I use? How do I do it?
What should the test look like? Etc. Any advice would be appreciated.

Thanks,



Steve Clark

_______________________________________________
Histonet mailing list
Histonet <@t> lists.utsouthwestern.edu
http://lists.utsouthwestern.edu/mailman/listinfo/histonet


       


More information about the Histonet mailing list