|
|
Different output using parameter labels |
|
Message/Author |
|
|
Hi, Before performing a Wald test I have assigned labels to my parameters, however after doing this my output changed (e.g. different intercept and slope values). I believe the output should remain the same as I am only assigning names. Who can tell me what I am doing wrong? My original syntax: MODEL: i s qs | t1bc_es@-3.75 t5bc_es@-3.5 t9bc_es@-3.25 t11bc_es@-1.625 t12bc_es@0; i s qs ON t0bc_es conovk conspx conos; My syntax assigning parameter labels: MODEL: i s qs | t1bc_es@-3.75 t5bc_es@-3.5 t9bc_es@-3.25 t11bc_es@-1.625 t12bc_es@0; i s qs ON t0bc_es (p1) conovk (p2) conspx (p3) conos (p4); Thanks! |
|
|
i s qs ON t0bc_es (p1) says that i ON t0bc_es, s ON t0bc_es, and qs ON t0bc_es are equal. I think you want to say i ON t0bc_es (p1); s ON t0bc_es (p5); qs ON t0bc_es (p6); |
|
|
Thank you! That solved the problem! However, when I now perform the wald test, it does not appear in the output. I think this might be because I am testing too many relationships at once. My syntax: MODEL: i s qs| t1bc_es@-3.75 t5bc_es@-3.5 t9bc_es@-3.25 t11bc_es@-1.625 t12bc_es@0; i ON t0bc_es (p1) conovk (p2) conspx (p3) conos (p4); s ON t0bc_es (p5) conovk (p6) conspx (p7) conos (p8); qs ON t0bc_es (p9) conovk (p10) conspx (p11) conos (p12); MODEL TEST: 0 = p1; 0 = p2; 0 = p3; 0 = p4; 0 = p5; 0 = p6; 0 = p7; 0 = p8; 0 = p10; 0 = p11; 0 = p12; I want to test if adding the condition variables (conovk, conspx and conos) together explain significant variation. Is there a way to do this? I am using multiple imputed data. |
|
|
Please send the output and your license number to support@statmodel.com. |
|
Back to top |
|
|