Data analysis
Analyses of variance and analyses of covariance (ANCOVA) were
used to compare differences in the PEM-CY community participation
and environment summary scores and between children
with and without disabilities. The child’s age category and family
income category were entered as covariates because of group
differences on these variables in our sample. Partial eta squares
(n2p) were used to examine effect sizes specific to differences
in the disability group according to Kirk’s classification
(smallZ.01e.05; moderateZ.06e.13; large .14).34
ANCOVAs (adjusting for age and income category) and partial
eta squares were also used to compare mean item-level scores for
participation frequency and involvement between children with
and without disabilities. Chi-square analyses were used to
compare the percentage of parents in each group who responded
(a) “never” for participation frequency items; (b) “yes” for desired