Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Graph Quadratic Term


From   Steven Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Graph Quadratic Term
Date   Tue, 19 Apr 2011 12:48:21 -0400

Dennis,

Quadratic trends can look linear. If you print out the graph and lay a ruler over the extreme points, what do you see?


Steve
sjsamuels@gmail.com


On Apr 18, 2011, at 1:00 PM, Nick Cox wrote:

I don't know what you mean by "does not plot correctly" and in any
case you give very little code.

-predict- is a command, not a function.

This example may help

sysuse auto
gen headroomsq = headroom^2
logit foreign headroom headroomsq
predict predict, pr
scatter foreign headroom || mspline predict headroom

A quadratic on the logit scale naturally translates to a bell-shape on
the original scale.

Nick

On Mon, Apr 18, 2011 at 5:36 PM, Dennis Kramer <dkramerii@gmail.com> wrote:

> I am having trouble graphing the quadratic relationship between my
> outcome dependent and independent variable. I am currently modeling a
> -logit- function with my outcome being 0/1 for passing a standardized
> test. When I run my --logit-- with the sq term in the model I get a
> significant quadratic function; however, when I use the -- predict, pr
> -- function to plot for predictions the quadratic function does not
> plot correctly... could you provide some insight.
*
*   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/


*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index