Alternatives to Mixture Model Analysis of Correlated Binomial Data

Table 4

Analysis of English sole eggs data: (i) GEE parameter estimates and working correlation matrix using the SAS system GENMOD procedure. (ii) GEE parameter estimates and working correlation matrix using the TIBCO Spotfire S+. (iii) GEE parameter estimates and eigenvalues of the working correlation matrix using STATA version 11. (iv) QLS parameter estimates and correlation estimates using an R program given in the Appendix.

(i) GEE estimates using SAS GENMOD procedure							Working correlation
Parm.	Est.	SE	95% C.I.		Z	Pr>|𝑍|	1.000	0.999	0.999	0.999
Int.	−1.983	1.149	−4.235	0.270	−1.73	0.085		1.000	0.999	0.999
Sal.	−0.017	0.038	−0.091	0.058	−0.43	0.664			1.000	0.999
Temp.	0.378	0.163	0.059	0.697	2.32	0.020				1.000
(ii) GEE estimates using TIBCO Spotfire S+							Working correlation
Parm.	Est.	SE	95% C.I.		Z	Pr>|𝑍|	1.000	1.049	0.890	0.978
Int.	−2.096	1.562	−5.157	0.965	−1.34	0.180		1.000	0.820	0.874
Sal.	−0.010	0.040	−0.089	0.069	−0.25	0.799			1.000	0.754
Temp.	0.379	0.141	0.103	0.656	2.69	0.007				1.000
(iii) GEE estimates using STATA 11							Eigenvalues
Parm.	Est.	SE	95% C.I.		Z	Pr>|𝑍|	3.664	0.259	0.114	−0.038
Int.	−2.096	1.562	−5.157	0.965	−1.34	0.180
Sal.	−0.010	0.040	−0.089	0.069	−0.25	0.799
Temp.	0.379	0.141	0.103	0.656	2.69	0.007
(iv) QLS estimates using R 2.14.1							Working correlation
Parm.	Est.	SE	95% C.I.		Z	Pr>|𝑍|	1.000	0.968	0.920	0.940
Int.	−1.936	1.623	−5.117	1.244	−1.19	0.233		1.000	0.931	0.912
Sal.	−0.018	0.042	−0.100	0.064	−0.42	0.672			1.000	0.874
Temp.	0.370	0.147	0.083	0.657	2.53	0.012				1.000