irt_coef help file


    irt_coef --

               Calculates y* standardized coefficients (discrimination parameters) for

               binary and ordinal IRT models

General syntax

    irt_coef [varlist] , [options]


    irt_coef calculates y* standardized coefficients (discrimination parameters) for

    binary and ordinal IRT models. The raw coefficient, standard error, and p-value are

    also reported alongside the y* standardized coefficient.


    ----+ Required Option for gsem models +----------------------------------------------

    latent(name)  is required if gsem was used to fit the model of interest. This is the

                  name of the latent variable that is the indepenent variable (the items

                  are the dependent variables [i.e. the y's]).  The latent( ) option should

                  not be used if the model was fit with the irt command as irt

                  automatically names the latent variable Theta.


    ----+ Options +----------------------------------------------------------------------

    model(name)   specifies the name of saved model estimates to use.  See estimates store

                  for saving model estimates. By default, irt_coef will use the IRT/GSEM

                  estimates in memory. If the relevant model estimates are not in memory,

                  you must specify their name.

    decimals(#)   changes the number of decimal places reported for the statistics. The

                  default is 3. Any integer between 1 - 8 is allowed.

    sort          orders the rows of the table based on the values of the y* standardized



                  changes the title of the table of results.  A default title is

                  automatically included.

    help          prints footnotes below the table describing what the columns in the table



    webuse charity


    irt grm ta1 ta2 ta3 ta4 ta5


    irt_coef, help


    gsem (Theta -> ta1 ta2 ta3 ta4 ta5, ologit), var(Theta@1)


    irt_coef, latent(Theta)




    For details on y* standardized coefficients generally see Long 1997 (pages 69-71;

    128-130). In the context of IRT models, see Bartholomew et al.  2008 (pages 224-225;



    irt_coef is written by Trenton D Mize (Department of Sociology & Advanced

    Methodologies, Purdue University).  Questions can be sent to


    Bartholomew, David J., Fiona Steele, Irini Moustaki, and Jane I. Galbrath.  2008.

    Analysis of Multivariate Social Science Data. Second Edition. CRC Press.

    Long, J. Scott. 1997.  Regression Models for Categorical and Limited Dependent

    Variables. Sage.