psib: POLYSIBTEST

In jweese1441/DIFSIB: SIBTEST Suite of Procedures

POLYSIBTEST

Description

This function assesses uniform DIF for polytomous DATA using the POLYSIBTEST procedure (Chang et al., 1996). The function outputs the uniform DIF magnitudes as well as the significance testing.

Usage

psib(
  data_ref,
  data_foc,
  minc = 2,
  cusr = 0,
  idw = 0,
  suspect_items,
  matching_items,
  listwise = 1,
  nch
)

Arguments

data_ref	The dataset that contains the reference group participants
data_foc	The dataset that contains the focal group participants
minc	The minimum number of individuals in a give total test score category. Initialized to 2 per Shealy and Stout (1993) recommendation
cusr	The guess correction factor. This is initialized to 0, but can be changed based on the items used. The cusr value is the average of guess correction across the items in the dataset.
idw	The weighting factor. This is initialized to use the reference and focal group weighting recommended by Shealy and Stout (1993)
suspect_items	the item(s) to be assess for DIF. Listing one item number will assess for DIF on that item, multiple items (e.g. c(2,3,8)) will assess DIF for the bundle of items.
matching_items	the list of items to match on.
listwise	initialized to 1, this inidicates that the procedure will not delete any rows with missing data, changing this to 2 will listwise delete.
nch	a vector indicating the number of categories for each item. Example if there are 5 items and each has 3 catgories nch <- c(3,3,3,3,3). If there are 5 items and items 1 - 3 have 4 categories and items 4 and 5 have 2 then nch <- (4,4,4,2,2)

References

Chang, H. H., Mazzeo, J. & Roussos, L. (1996). DIF for Polytomously Scored Items: An Adaptation of the SIBTEST Procedure. Journal of Educational Measurement, 33, 333-353.

DIF-Pack (2021) Measured Progress. https://psychometrics.onlinehelp.measuredprogress.org/tools/dif/

Shealy, R. & Stout, W. (1993). A model-based standardization approach that separates true bias/DIF from group ability differences and detect test bias/DTF as well as item bias/DIF. Psychometrika, 58, 159-194.

Weese, J. D., Turner, R. C., Liang, X., Ames, A., & Crawford, B. (accepted). Implementing a Standardized Effect Size in the POLYSIBTEST Procedure

Examples

## Not run: 

#perform POLYSIBTEST with one suspect item. Sample data has 20 items with 5 categories each.
nch <- rep(5,20)
psib(data_ref = data_refP, data_foc = data_focP, 
minc = 2,suspect_items = c(20), matching_items = c(1:19),nch = nch)

#perform POLYSIBTEST with a bundle of suspect items 

psib(data_ref = data_refP, data_foc = data_focP, 
minc = 2, suspect_items = c(16,17,18,19,20), matching_items = c(1:15),nch = nch)


## End(Not run)

jweese1441/DIFSIB documentation built on April 3, 2022, 8:58 a.m.