Item analysis using Rasch models confirms that the Danish versions of the DISABKIDS® chronic-generic and diabetes-specific modules are valid and reliable

Table 6 Overview of the reliability of subscales

Domain	Subscales	Cronbach’s Alpha	Reliability^a	Reliability depends on	SEM^b	Observed test-retest correlation
Mental	Independence	0.83	0.72–0.91	Sex and Age	1.5–1.8	0.73
Mental	Emotion	0.82	0.65–0.89	Sex and Age	2.6–3.0	0.64
Social	Inclusion	0.64	0.64–0.66	Sex and Age	2.6	0.57
Social	Exclusion	0.75	0.50–0.84	Sex, Age, and HbA1c	2.2–2.7	0.85
Physical	Limitation	0.71	0.67–0.78	Age	2.4–2.5	0.57
Physical	Treatment	0.80	0.79	None	3.0	0.69
Diabetes module	Impact	0.78	0.77–0.78	Sex	2.4	0.69
Diabetes module	Treatment	0.84	0.82–0.88	Age	1.5–2.0	0.77

This table displays both Cronbach’s Alpha, which is known to provide a lower bound to the true reliability if items are locally independent, and reliability calculated using the Monte Carlo method. The observed test-retest results are provided in the final column
^aReliability [Variance (True score)/Variance (Score)] depends on both the population and on the DIF among items. It is necessary to calculate reliability in subgroups defined by variables with a significant effect on the score. Reliability is therefore reported as an interval from the smallest to the largest degree of reliability in these groups
^aSEM = The standard error of the total score as an estimate of the true score. The SEM depends on the true score and the DIF. SEM is therefore reported as an interval of the largest SEM value in the groups defined by the sources of DIF

ISSN: 1477-7525