Given a data.frame containing the item parameters from a differential item functioning (DIF) analysis, the rules for A-, B-, and C-DIF implemented by the Educational Testing Service (ETS) can be applied to the items here. For each item, the program indicates whether its DIF parameter meets the A, B, or C criterion. Users generally do not need to use this function directly, as it is implemented within getResults. The exception is DIF models performed as part of partial credit analyses.

lord1980(dat, absDifCol, seDifCol, lbCol, ubCol, adb=NULL, sdb=NULL)

Arguments

dat

A data.frame with DIF item parameters. The most common use case here would be to use the "item*step" sheet from the Conquest showfile, which is re-imported using get.shw or get.shw.v5.

absDifCol

The column or column number in dat that contains the absolute DIF value.

seDifCol

The column or column number in dat that contains the standard error of the DIF parameter.

lbCol

The column or column number in dat that contains the lower bound of the confidence interval.

ubCol

The column or column number in dat that contains the upper bound of the confidence interval.

adb

The absolute value of the DIF, which is to be used as a threshold. This is only necessary if the user wishes to use a custom criterion for the DIF.

sdb

The significance value of the DIF, which is to be used as a threshold. This is only necessary if the user wishes to use a custom criterion for the DIF.

Details

The user-specific DIF indicator is 1 if the absolute DIF value is greater than adb at a significance level of sdb.

Value

The function returns the original data.frame, which is expanded by two columns: an indicator column that, if desired, contains the DIF criterion specified by the user; and a column containing the DIF according to the ETS criteria (A, C, and C).

Examples

# see example 11 in the help file of defineModel()