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

# Re: st: Test for significance of the difference between knee locations?

 From Austin Nichols To statalist@hsphsun2.harvard.edu Subject Re: st: Test for significance of the difference between knee locations? Date Fri, 19 Oct 2012 09:48:30 -0400

```Jordan Silberman <silberman.stata@gmail.com>:
You need to be a lot more clear about what you want. Below, you seem
to be describing estimation of a kink point, using two linear curves
that meet at one value of x, without discontinuities in the curves but
with a discontinuity in derivatives. Then in the original post you
describe comparing two quadratic curves (presumably with no kinks).
Can you show what you want using a publicly available dataset e.g.
wage as a function of age for union and nonunion:
sysuse nlsw88, clear
tw qfit wage age if union==0||qfit wage age if union==1

On Thu, Oct 18, 2012 at 5:26 PM, Jordan Silberman
<silberman.stata@gmail.com> wrote:
> According to one approach (from Dmitri Kaplan, described here:
> http://www.mathworks.com/matlabcentral/fileexchange/35094-knee-point/content/knee_pt.m),
> the knee is the x value that minimizes the sum of the squared errors
> of two linear regressions--one modeling all points left of the knee
> and one modeling all points right of the knee. It's an estimate of the
> point at which the slope "turns."
>
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/
```