Differential item functioning (DIF) is investigated using log-odds estimators in Table 30.1Mantel-Haenszel (1959) for dichotomies or Mantel (1963) for polytomies. The sample is divided into difference classification groups (also called reference groups and focal groups) which are shown in Table 30 and specified with DIF=.

 

MHSLICE= specifies the width of the slice (in logits) of the latent variable be included in each cross-tab. The lower end of the lowest slice is always the lowest observed person measure.

 

MHSLICE = 0 bypasses Mantel-Haenszel or Mantel computation.

 

MHSLICE = .1 logits and smaller. The latent variable is stratified into thin slices. This corresponds to the slicing by raw scores with complete data

 

MHSLICE = 1 logit and larger. The latent variable is stratified into thick slices.

 

For each slice, a cross-tabulation is constructed for each pair of person classifications against each scored response level. An odds-ratio is computed from the cross-tab. Zero and infinite ratios are ignored. A homogeneity chi-square is also computed when possible.

 

Example: The sample size is small, but an approximate Mantel-Haenszel estimate is wanted:

MHSLICE=3 ; this probably divides the sample into 3 strata.