Validity of the Play Assessment for Group Settings: an evaluation of differential item functioning between children...

The purpose of this study was to evaluate the validity evidence based on internal structure of the Play Assessment for Group Settings (PAGS).

The study was conducted in day care centres by comparing observations of the free play performance of two groups of children: those with specific language impairments (SLIs) (n=55) and those with typical language development (n=55). The participating children were 3 to 6.5 years of age. Data were subjected to many-faceted Rasch analyses and differential item functioning analysis was conducted to identify possible group-specific items in the PAGS. The effect of differentially functioning items on mean play performance measurements between two subgroups of children was controlled with two different independent samples t-tests.

According to the results, 80% of the test items were stable across the groups of children. Four items were harder for children with SLI, and clearly required language skills. Two items were relatively speaking easier for children with SLI, and reflected explorative functioning. The differential functioning items did not affect the ability of the PAGS to separate the children with SLI in mean play performance from their typically developing peers.

We concluded that the PAGS is a valid tool for identifying children who have challenges in their play performance and can be used for clinical purposes. However, the slight variability supports the fact that differential item functioning should be controlled in research when using the PAGS for comparison of different subgroups in play performance.