simul_confusion/Modules/ado/personal/p/pcmodel.hlp

{smcl}
{* Mars 2012}{...}
{hline}
help for {hi:pcmodel}{right:JFH}
{hline}

{title:Estimation of the parameters of a Partial Credit Model}

{p 8 14 2}{cmd:pcmodel} (varlist) [{cmd:if} {it:exp}], 
[{cmdab:qual:itatives}({it:varlist})
{cmdab:quant:itatives}({it:varlist}) 
{cmdab:dif:ficulties}({it:matrix list}) 
{cmdab:it:erate}(#)
{cmdab:ad:apt} 
{cmdab:ro:bust}
{cmdab:f:rom}(matrix)]

{title:Description}

{p 8 14 2}{cmd:pcmodel} allows estimating the parameters of a random effect
 partial credit model (the item difficulties, and the potential covariates
  that may influence the considered latent trait are considered as fixed effects.
   The individual latent traits are considered as a  normally distributed
    random effect){p_end}
{p 14 14 2}Two situations are possible:{p_end}
{p 14 14 2}- The item difficulties can be
 considered as already known. They do not have to be estimated during the
  analysis.{p_end}
{p 14 14 2}- The difficulties are considered as unknowns, and then require
   to be estimated during the analysis. {p_end}
{p 14 14 2}It is possible to include covariates (that can possibly influence
 the individual latent traits) in the partial credit model. These covariates
  can either be qualitative or quantitative. {p_end}
{p 14 14 2}{cmd:pcmodel} provides assistance for interpreting the estimated
 effects of covariates (estimation of the type III sum of squares, percentage
  of variance explained with the introduction of each covariates). {p_end}
{p 14 14 2}It is finally possible to thest the fit using {cmd:pcmtest} after
 estimating parameters of the model with {cmd:pcmodel}

{title:Options}

{p 4 8 2}{cmdab:qual:itatives} List of the categorical covariates included in
 the partial credit model

{p 4 8 2}{cmdab:quant:itatives} List of the continuous covariates included in
 the partial credit model

{p 4 8 2}{cmd:difficulties} Row vectors containing the considered known values
 of each item difficulty. A row vector must match with each item, and have the
  same name as the corresponding item. The item difficulties should be known for
   all the items. If {cmd:difficulties} is not filled, the dificulties are
    considered as unknown, and are estimated during the analysis. 

{p 4 8 2}{cmd:iterate} specifies the (maximum) number of iterations. With the
 adapt option, use of the iterate(#) option will cause pcmodel to skip the
  "Newton Raphson" iterations usually performed at the end without updating
   the quadrature locations.


{p 4 8 2}{cmd:adapt} causes adaptive quadrature to be used instead of
 ordinary quadrature.

{p 4 8 2}{cmd:robust} specifies that the Huber/White/sandwich estimator of the
 covariance matrix of the parameter estimates is to be used. 


{p 4 8 2}{cmd:from} specifies a row vector to be used for the initial values.
 Note that the column-names and equation-names do not have to be correct. This
  line vector must have exactly the number of parameters to be estimated,
   starting with the difficulties parameters, then the parameters associated
    with the covariates, and finishing with the estimated standard deviation
     of the latent trait.

{title:Outputs}

{p 4 8 2}{cmd:e(lll)}: (marginal) log-likelihood

{p 4 8 2}{cmd:e(cn)}: Condition number

{p 4 8 2}{cmd:e(N)}: Number of observations

{p 4 8 2}{cmd:e(Nit)}: Number of items

{p 4 8 2}{cmd:e(Nqual)}: Number of qualitative covariates

{p 4 8 2}{cmd:e(Nquant)}: Number of quantitative covariates

{p 4 8 2}{cmd:e(items)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(itest)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(datatest)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(mugauss)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(sdgauss)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(cmd)}: program used to implement {cmdpcmtest}

{p 4 8 2}{cmd:e(sigma)}: Estimated standard deviation of the latent trait

{p 4 8 2}{cmd:e(Varsigma)}: Variance of the estimated standard deviation of
 the latent trait

{p 4 8 2}{cmd:e(b)}: coefficient vector of the parameters associated with
 the latent trait covariates (if no covariate is included in the model, value
  of the average latent trait).

{p 4 8 2}{cmd:e(V)}: Covariance matrix for the latent trait covariates.

{p 4 8 2}{cmd:e(delta)}: Estimated difficulty parameters

{p 4 8 2}{cmd:e(Vardelta)}: Covariance matrix for the estimated difficulty
 parameters


{title:Author}

{p 4 8 2}Jean-François HAMEL{p_end}

{title:Also see}

{p 4 13 2}Online: help for {help pcmtest}, {help gllamm}, {help simirt},
 {help raschtest}.{p_end}