Scatchard-type analysis and homoscedasticity

T. Mark Reboul MARK at CUCCFA.CCC.COLUMBIA.EDU
Thu Jul 29 07:42:45 EST 1993


Nick Holford makes a key point here.

Although we have now acquired Munson's Ligand for Mac to take care
of these analyses, originally I had wrongly proposed a quick & dirty
solution: using the constrained, non-linear regression function of 
some stats package (e.g., SPSS, or SAS) to treat 2-site data recorded
in (B,B/F) coordinates. That solution would have been wrong, the 
biggest reason being what Holford was talking about.

The software available from Munson/NIH does the right thing. If there
is any doubt about this, see the original papers by Munson and/or 
Rodbard, who are well aware of the issue. (And if the issue remains
unclear from the limited postings in this forum, then see those same

papers for a complete introduction!)

Mark Reboul
Columbia-Presyterian Cancer Center Computing Facility
mark at cuccfa.ccc.columbia.edu




More information about the Bio-soft mailing list