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, is already up and running.

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

AW: AW: st: prgen option quadratic term

From   "Meulemann Max" <>
To   "" <>
Subject   AW: AW: st: prgen option quadratic term
Date   Wed, 7 Nov 2012 11:29:34 +0000

I had hoped that was taken care of by the margins command 

ologit y capitagdp c.capitagdp#c.capitagdp i.socioecovars 
margins, at(c_capitagdp=(0.5(0.5)9))  atmeans predict(outcome(4))

which looked the same as 

ologit y c_logcapitagdp c_logcapitagdpsquared $socioecovars
oprobpr c_logcapitagdp, levels(c_logcapitagdpsquared=c_logcapitagdp^2) cat(4) newobs(100)

In the linear model that graph is downward slooping and in the quadratic model it is downward slooping and then turning up

Von: []&quot; im Auftrag von &quot;Nick Cox []
Gesendet: Mittwoch, 7. November 2012 11:42
Betreff: Re: AW: st: prgen option quadratic term

When you are fitting a linear term and a quadratic term, the two act
together. It's their combined effect that you need to interpret. This
is usually easier with a graph. For example, one negative coefficient
and one positive coefficient mean a turning point somewhere, but not
necessarily within the range of the data.


On Wed, Nov 7, 2012 at 10:35 AM, Meulemann  Max <> wrote:
> thx, the margins and marginsplot function helped a lot.
> Anyways anybody got a idea how I am to interpret my data here: I find a negativ linear term for capitagdp, but if i include a quadratic gdp term, that one turns out to be positiv and significant with a lrtest.
> Plotting with marginsplot tells me then that for the linear model the probability to agree with a certain issue decreases whereas it increases with capitagdp for the quadratic model, keeping fixed my other covariates
> ________________________________________
> Von: []&quot; im Auftrag von &quot;Maarten Buis []
> Gesendet: Dienstag, 6. November 2012 17:10
> An:
> Betreff: Re: AW: st: prgen option quadratic term
> Also see: <>
> -- Maarten
> On Tue, Nov 6, 2012 at 4:26 PM, Richard Goldstein
> <> wrote:
>> I am not familiar with prgen
>> sounds like you want the -margins- command?
>> On 11/6/12 10:24 AM, Meulemann Max wrote:
>>> I d like to get the predicted probabilites as my contionus predicator changes keeping fixed the other ones as it is done in -prgen- only with a quadratic term of that predicator. I do not just want the predicted probabilites.
>>> ________________________________________
>>> Von: []&quot; im Auftrag von &quot;Richard Goldstein []
>>> Gesendet: Dienstag, 6. November 2012 16:10
>>> An:
>>> Betreff: Re: st: prgen option quadratic term
>>> I don't understand; why don't you start by telling us what is wrong with
>>> just using -predict- after your -ologit-? for help on this, start with
>>> -h ologit-, then click on "also see" and click on "ologit postestimation"
>>> Rich
>>> On 11/6/12 10:07 AM, Meulemann Max wrote:
>>>> I would like to calculate predicted probabilities in an ordinal model, ologit or oprobit. I am using a continous quadratic term, which makes calculating the predicted probablities more cumbesome. I am using stata 12
>>>> I found an old entry in the statalist archive that  proposed oprobpr which does work. I was wondering if another package has been written since 2005.
>>>> oprobpr does work, but some of it its features are a bit irritating like how it saves the predicted probailites in another dataset and so on. The nicest way would be to change prgen and its familiar option in a way that the same feature that works for oprobpr would work for prgen.
>>>> the level options allows oproppr to vary an otherwise fixed covariate with a squared term, which is not possible for the x option of prgen.

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index