A restricted latent class model with polytomous attributes and respondent-level covariates

Table 3 Results of second simulation study (contd.)

\(N\)	\(J\)	\(K\)	\(L\)	\(\rho\)	\(\beta\)	\(\delta\)	\(\delta ^0\)	\(\delta^1\)	\(\beta^0\)	\(\beta ^1\)
3000	45	4	3	0.000	0.046	0.984	0.987	0.972	0.014	0.182
3000	70	5	2	0.000	0.021	0.994	0.993	0.997	0.007	0.077
3000	70	5	3	0.000	0.033	0.990	0.994	0.965	0.007	0.199
3000	45	4	3	0.250	0.060	0.974	0.975	0.967	0.027	0.199
3000	70	5	2	0.250	0.017	0.997	0.996	0.999	0.004	0.067
3000	70	5	3	0.250	0.047	0.979	0.984	0.952	0.018	0.230
3000	45	4	3	0.500	0.136	0.921	0.922	0.919	0.091	0.327
3000	70	5	2	0.500	0.040	0.978	0.975	0.991	0.025	0.098
3000	70	5	3	0.500	0.085	0.966	0.972	0.928	0.047	0.328
5000	45	4	3	0.000	0.040	0.986	0.987	0.980	0.013	0.153
5000	70	5	2	0.000	0.020	0.993	0.991	0.998	0.008	0.066
5000	70	5	3	0.000	0.025	0.993	0.995	0.979	0.005	0.155
5000	45	4	3	0.250	0.052	0.976	0.977	0.975	0.024	0.168
5000	70	5	2	0.250	0.012	0.997	0.996	1.000	0.003	0.049
5000	70	5	3	0.250	0.042	0.979	0.981	0.968	0.019	0.191
5000	45	4	3	0.500	0.137	0.916	0.914	0.926	0.093	0.319
5000	70	5	2	0.500	0.041	0.972	0.968	0.988	0.029	0.091
5000	70	5	3	0.500	0.075	0.964	0.967	0.946	0.041	0.297

Values displayed for \(\beta\) parameters are the average, taken over all elements of the parameter, of the mean absolute error of estimation of each element over all replications. Values displayed for \(\delta\) parameters are the average, taken over all elements of the parameter, of the recovery accuracy of each element over all replications