{smcl}
{hline}
help for {cmd:difd} {right:(April 11, 2011)}
{hline}
{title:Detection of differential item function (DIF).}
{p 8 17 2}
{cmd:difd} {it: varlist} {cmd:,}
{cmd:ABility}{it:(varlist)} {cmd:GRoups}{it:(varlist)} {break}
[{cmdab:CATegorical}{cmd:}{it:(varlist)}{cmd:}
{cmdab:RUnname}{cmd:}{it:(str)}{cmd:} {cmdab:NUL}{cmd:}{it:(#)}{cmd:}
{cmdab:NUP}{cmd:}{it:(#)}{cmd:} {cmdab:NUPValue}{cmd:}{it:(#)}{cmd:}
{cmdab:UBeta}{cmd:}{it:(#)}{cmd:} {cmdab:UBP}{cmd:}{it:(#)}{cmd:}
{cmdab:ULPV}{cmd:}{it:(#)}{cmd:} {cmdab:UP}{cmd:}{it:(#)}{cmd:}
{cmdab:UPPValue}{cmd:}{it:(#)}{cmd:} {cmdab:ITemsub}{cmd:}{it:(#)}{cmd:}]
{p 4 12 2}
where:
{p 8 12 2}
{it:varlist} is the list of variables (items, blocks) to be tested for DIF
{p 8 12 2}
{it:id} is the subject id variable.
{p 8 12 2}
{it:ability} is the ability variable(s).
{p 8 12 2}
{it:groups} is the list of grouping variables. Continuous ‘grouping’ variables are permitted.
{title:Options}
{p 8 12 2}
{cmdab:categorical} is the list of any group variables that are categorical and have more than 2 levels. Note that it is simpler to omit dichotomous variables from this list. Default is none (all continuous or dichotomous).
{p 8 12 2}
{cmdab:runname} names the log file DIFdRUnname.log. Default is DIFd.log.
{p 8 12 2}
{cmdab:nul} indicates whether the log-likelihood test will be used as a criterion for non-uniform DIF. Default is yes (1). Nul(0) will omit this criterion.
{p 8 12 2}
{cmdab:nup} indicates whether the Wald test for the interaction term will be used as a criterion for non-uniform DIF. Default is no (0). Nup(1) will include this criterion.
{p 8 12 2}
{cmdab:nupvalue} is the p-value for testing non-uniform DIF. Default is 0.05.
{p 8 12 2}
{cmdab:ubeta} indicates whether the change in the ability coefficient will be used as a criterion for uniform DIF. Default is yes (1). UBeta(0) will omit this criterion.
{p 8 12 2}
{cmdab:ubp} is percent change in the ability coefficient for determining uniform DIF. Default is .10. A positive change indicates an increase in the relationship between ability and the outcome with a higher value of the grouping variable.
{p 8 12 2}
{cmdab:ul} indicates whether the log-likelihood test will be used as a criterion for uniform DIF. Default is no (0). UL(1) will include this criterion.
{p 8 12 2}
{cmdab:ulpvalue} is the p-value for testing uniform DIF with the log-likelhood method. Default is 0.05.
{p 8 12 2}
{cmdab:up} indicates whether the Wald test for the group term will be used as a criterion for uniform DIF. Default is no (0). UP(1) will include this criterion.
{p 8 12 2}
{cmdab:ulpvalue} is the p-value for testing uniform DIF with the Wald test. Default is 0.05.
{p 8 12 2}
{cmdab:itemsub} subtracts the item value from the ability measure. Default is no (0). ITemsub(1) will include this feature.
{title:Remarks}
{p 8 12 2}
Sends DIF results to DIFd{it:runname}.log.
{p 8 12 2}
DIF results for categorical grouping variables will be in terms of the ordered values of group. For example, if {it:ethnic} has 3 levels, 3 sets of DIF results will be reported: {it:ethnic12, ethnic13, ethnic23},
where {it:ethnic12} compares the 2 lowest values of
{it:ethnic}, {it:ethnic13} the lowest and highest, etc.
{p 8 12 2}
Generates an output data set, DIFd.dta, which includes individual model results, with Brant test p-values for ordinal items and Hosmer-Lemeshow p-values for binary items. [The relevance of the fit statistics has not been established for DIF.]
{p 8 12 2}
Displays warning messages when models do not converge, collinearity problems are observed, models are completely determined, standard errors are large, or Brant tests are not possible.
{title:Examples}
{p 4 8 2}
difd item1-item13, id(id) ru(gender0) ab(theta0) gr(g)
{p 4 8 2}
difd apple - item13, id(id) ab(theta0) gr(eth) cat(eth) nupv(0.01) ul(1) ulpv(.01)
{title:Authors}
{p 4 4 2}
Paul Crane, Laura Gibbons, Lance Jolley, and Gerald van Belle. University of Washington, Copyright 2005.{break}
Email: {browse mailto: gibbonsl@u.washington.edu}
{title:Also see}
{p 4 4 2}
{cmd:difwithpar}