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]

Re: st: Mokken scale on mi data

From   Nick Cox <>
Subject   Re: st: Mokken scale on mi data
Date   Mon, 25 Feb 2013 13:37:15 +0000

-msp- (SSC) is not an e-class program. Its help documents that it is
r-class, i.e. leaves things in its wake with names wrapped in "()".

-mi- evidently expects to work with a e-class program that emits a
vector of coefficient estimates and a variance-covariance matrix.
There may be more requirements, but -msp- fails at the first hurdle.
Even if -msp- could be written or recast in those terms, some
programming is needed. If that makes no sense in principle, the road
ahead is totally blocked. I don't know more than this, unfortunately.
Others should know (much) more.

Please do remember to explain where user-written programs you refer to
come from, a longstanding and prominent request in the FAQ. To answer
this question, I had to spend some time finding where -msp- is.


On Mon, Feb 25, 2013 at 1:22 PM, Laura Maria Schwirz <> wrote:

> I have a dataset with missing values and used Stata's mi command to
> impute them. I had a look at the mi estimate commands and msp for
> Mokken Scale Procedure unfortunately is not one of them. I also tried
> using mi estimate, cmdok: msp, but it returned an error. Below
> is my syntax and the error message:
> . mi estimate, cmdok: msp strong1 knowo1 intel1 inspi1 moral1 com1
> decent1 depend1 rel1 sens1
> matrix e(b) is not set
> matrix e(V) is not set
> r(301)
> I had a look at some of the comments in previous threats but am unsure
> how to proceed from here.
> Also, running a factor analysis on imputed data is also not feasible.
*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index