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: st: impute items and total sum scores of a scale
From
"JVerkuilen (Gmail)" <[email protected]>
To
[email protected]
Subject
Re: st: impute items and total sum scores of a scale
Date
Thu, 6 Dec 2012 10:06:55 -0500
If it runs fine with the items then there's got to be something with
declaring the variables passive. Try just making one and seeing if
that crashes.
With Stata 12 MI it works pretty straightforwardly. You simply
calculate these as passive as they are linearly dependent on the
items.
You mention something about interactions but I'm not 100% sure what
you want. Those linear terms are linearly dependent and add no new
information if you are including them in the model.
On Thu, Dec 6, 2012 at 9:42 AM, Paula Arce <[email protected]> wrote:
>
> Thanks - I originally only imputed the items and the imputation ran correctly.
>
>
> However, that model did not contain the interaction terms I am intending to use in my analysis and I was told that I should include them in the model for a more efficient estimate.
>
>
> could give me any tip on how to best build up from the simpler model to include the sum scores ?
>
> Thanks again
>
>
>
> ----- Original Message -----
> From: JVerkuilen (Gmail) <[email protected]>
> To: [email protected]
> Cc:
> Sent: Thursday, 6 December 2012, 13:22
> Subject: Re: st: impute items and total sum scores of a scale
>
> On Thu, Dec 6, 2012 at 7:25 AM, Paula Arce <[email protected]> wrote:
>>
>
>> I am using mi ice to impute missing data into my data set. I need to impute both single items and total scores of my scales.
>
> What you're doing makes sense, but you might want to use the new MI
> suite, which in Stata 12 has the full conditional specification that
> you get from ice.
>
> If I had to guess you have a collinearity problem and one of the
> imputations is not converging. Try stripping the model down and see if
> you can just impute the items first, or even a subset of them, or use
> linear regression and see if you get a reasonable answer, then build
> up to the model you want.
>
> MI is never smooth.
> *
> * 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/
>
>
> *
> * 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/
--
JVVerkuilen, PhD
[email protected]
"Thus the typical citizen drops down to a lower level of mental
performance as soon as he enters the political field. He argues and
analyzes in a way which he would readily recognize as infantile within
the sphere of his real interests. He becomes a primitive again. His
thinking becomes associative and affective." ---Joseph A. Schumpeter,
Capitalism, Socialism and Democracy, 1950, p. 262.
*
* 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/