------------------------------------------------------------------------------- help fordifd(April 11, 2011) -------------------------------------------------------------------------------

Detection of differential item function (DIF).

difdvarlist,ABility(varlist)GRoups(varlist)[CATegorical(varlist)RUnname(str)NUL(#)NUP(#)NUPValue(#)UBeta(#)UBP(#)ULPV(#)UP(#)UPPValue(#)ITemsub(#)]where:

varlistis the list of variables (items, blocks) to be tested for DIF

idis the subject id variable.

abilityis the ability variable(s).

groupsis the list of grouping variables. Continuous ‘grouping’ variables are permitted.

Options

categoricalis 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).

runnamenames the log file DIFdRUnname.log. Default is DIFd.log.

nulindicates 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.

nupindicates 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.

nupvalueis the p-value for testing non-uniform DIF. Default is 0.05.

ubetaindicates 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.

ubpis 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.

ulindicates whether the log-likelihood test will be used as a criterion for uniform DIF. Default is no (0). UL(1) will include this criterion.

ulpvalueis the p-value for testing uniform DIF with the log-likelhood method. Default is 0.05.

upindicates 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.

ulpvalueis the p-value for testing uniform DIF with the Wald test. Default is 0.05.

itemsubsubtracts the item value from the ability measure. Default is no (0). ITemsub(1) will include this feature.

RemarksSends DIF results to DIFd

runname.log.DIF results for categorical grouping variables will be in terms of the ordered values of group. For example, if

ethnichas 3 levels, 3 sets of DIF results will be reported:ethnic12,ethnic13, ethnic23, whereethnic12compares the 2 lowest values ofethnic,ethnic13the lowest and highest, etc.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.]

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.

Examplesdifd item1-item13, id(id) ru(gender0) ab(theta0) gr(g)

difd apple - item13, id(id) ab(theta0) gr(eth) cat(eth) nupv(0.01) ul(1) ulpv(.01)

AuthorsPaul Crane, Laura Gibbons, Lance Jolley, and Gerald van Belle. University of Washington, Copyright 2005. Email: gibbonsl@u.washington.edu

Also see

difwithpar