Item | A: Rasch model–six items | B: Graphical log linear Rasch modela–five items |
---|
Observed γ | Expected γ |
p
| Observed γ | Expected γ |
p
|
---|
CG1 | .54 | .58 | .48 | .56 | .63 | .24 |
CG2 | .62 | .57 | .39 | .65 | .69 | .41 |
CG3 | .73 | .58 | .017 | .74 | .62 | .051 |
CG4 | .68 | .59 | .12 | .70 | .70 | .99 |
CG5 | .71 | .58 | .052 | .74 | .67 | .17 |
CG6 | .27 | .56 | < .001 | | | |
- The correlations are measured using Goodman and Kruskal's gamma. Goodman and Kruskal's gamma measures rank correlation for ordinal categorical data [37]. When item 6 is excluded, and local dependency and DIF are allowed the observed correlations are the same as expected
-
aThe model assumes that items 1 and 3 are locally dependent and that item five is affected by DIF depending on age