Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: RE: RE: RE: effect size in nonlinear regression


From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   st: RE: RE: RE: RE: effect size in nonlinear regression
Date   Tue, 16 Nov 2010 18:29:59 +0000

It looks as if your curves are asymptotic to lower and upper limits. If so, plotting curves on a logit scale, i.e. logit((value - lower) / (upper - lower)), will probably straighten curves, regardless of the metric in which you measure or fit. See also

SJ-8-1  gr0032  . . . . . . .  Stata tip 59: Plotting on any transformed scale
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q1/08   SJ 8(1):142--145                                 (no commands)
        tip on how to graph data on a transformed scale

for more on logit scale in graphs. 

Nick 
[email protected] 


Airey, David C

I agree about the graphs. Good point on the awkward choice of parameter symbols! :)

I'll just choose reasonable constant spread and differing parameter values and look at resulting power.

> Very small but possibly not minor detail: it depends on your audience, but e as a parameter suggests to many exp(1) ~ 2.71828, so that some other symbol may be advisable. 
> 

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index