Simulation study of activities of daily living functions using online computerized adaptive testing

Table 3 Correlation coefficients (left lower triangle) and different number ratios (right upper triangle) between scales’ estimated measures

Estimation methods	A	B	C	D	E	F
MAP
A. CADL	0.61	0.40 %
B. CADL_CAT	*0.77*
C. Male ADL	0.67	0.68	0.90	0.00 %
D. Male ADL_CAT	0.64	0.64	*0.93*
E. Female ADL	0.67	0.68	1.00	0.93	0.90	0.00 %
F. Female ADL_CAT	0.64	0.64	0.93	1.00	*0.93*
EAP
A. CADL	0.67	0.10 %
B. CADL_CAT	*0.76*
C. Male ADL	0.78	0.91	0.80	0.00 %
D. Male ADL_CAT	0.75	0.87	*0.95*
E. Female ADL	0.78	0.97	0.94	0.90	0.74	0.00 %
F. Female ADL_CAT	0.76	1.00	0.91	0.87	*0.97*
MLE
A. CADL	0.49	0.00 %
B. CADL_CAT	*0.76*	0.00
C. Male ADL	0.77	0.91	0.50	0.00 %
D. Male ADL_CAT	0.75	0.87	*0.95*
E. Female ADL	0.77	0.97	0.94	0.90	0.48	1.20 %
F. Female ADL_CAT	0.76	1.00	0.91	0.87	*0.97*

ADL-CAT activities of daily living-Computerized Adaptive Testing, CADL comprehensive activities of daily living-Computerized Adaptive Testing, MLE maximum likelihood estimation, EAP expected a posteriori estimation, MAP maximum a posteriori estimation
Scale reliability coefficients shown on the diagonal line from left to right in the upper MAP table; dimension coefficients displayed in the middle EAP table; model’s residual dimension coefficients displayed in the bottom MLE table
Bold-italic values are correlation coefficients (CC) between NAT and CAT. Bold ones show the binary NAT CADL has significant lower CCs (<0.80) than those CCs (>0.90) between polytomous scales

ISSN: 1472-6947