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

# Re: st: Regressions with dependent continuous variable with bounded range

 From Suryadipta Roy To statalist@hsphsun2.harvard.edu Subject Re: st: Regressions with dependent continuous variable with bounded range Date Sun, 18 Dec 2011 08:29:00 -0500

```Dear Brendan and Nick,
Thank you very much for the suggestions! It seems that I was thinking
in these lines as well, since I see that my do file has the folllwing
code: twoway (qfitci y x, stdf) (scatter y x), by(year) , given that I
have a panel of observations (N = 183, T = 6). I believe that my worry
was really due to the (upper) bounded nature of the dependent
variable.

Sincerely,

On Sat, Dec 17, 2011 at 2:30 PM, Nick Cox <njcoxstata@gmail.com> wrote:
> In this case
>
> twoway qfit mpg weight
>
> is an alternative.
>
> Nick
>
>
> On 17 Dec 2011, at 19:17, brendan.halpin@ul.ie (Brendan Halpin) wrote:
>
>> On Sat, Dec 17 2011, Suryadipta Roy wrote:
>>
>>> I would very much appreciate some help as to how to interpret the sign
>>> of the coefficient associated with the squared term of an explanatory
>>> variable in such a model.
>>
>>
>> The best way of interpreting it is graphically. For example:
>>
>> sysuse auto
>> reg mpg c.weight##c.weight
>> twoway function x*_b[weight] + x^2*_b[c.weight#c.weight], range(weight)
>>
>>
> *
> *   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/
```