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]
Re: RE: st: Checking to see if the association between two variables is linear or otherwise
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: RE: st: Checking to see if the association between two variables is linear or otherwise
Date
Sat, 13 Oct 2012 16:02:08 +0100
I am sympathetic to all parties here, but the difference seems a bit
bigger than this implies. Using dummies doesn't even commit the
analyst to the idea that level of education is ordinal scale, although
my teaching career has been futile if that's not true. Using a
regression spline does seem to treat level of education, or some
transform of it, as at least interval scale.
Nick
On Sat, Oct 13, 2012 at 2:51 PM, JVerkuilen (Gmail)
<[email protected]> wrote:
> On Sat, Oct 13, 2012 at 3:20 AM, Justina Fischer <[email protected]> wrote:
>> see my post to David's comment. His critique applies only if educational levels are employed as continuous variable, which I never suggested.
>>
>
> Dummies for education level bins is essentially a zero-order spline
> and so a regression spline gives a way to compromise between the
> "wiggly" regression implied by dummies and the overly smooth one
> represented by a linear term, so the choice isn't really an either/or.
*
* 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/