Hi,
I have a DCE with two alternatives and an opt out option. I have coded the opt out option to have all attributes equal 0. I have also created a varirable called 'Opt Out' where if it is an opt out option, Opt out =1, if its alt A or alt B, opt out = 0. I have included this variable within the MNL model
I have used the following code to run an MNL model:
clogit chosen i.X1 i.X2 i.X3 X4_cont i.X5 X6_cont OptOUT, group(OBSID) vce(cluster RID)
clustered by patient ID, and separate observations created using variable OBSID.
However when I run the model, Opt out is omitted and there is no omitted level for X1, despite being dummy coded
See below:
og pseudolikelihood = -751.87905 Pseudo R2 = 0.1734
(Std. err. adjusted for 69 clusters in RID)
------------------------------------------------------------------------------
| Robust
chosen | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
X1 |
1 | 1.948214 .274358 7.10 0.000 1.410482 2.485945
2 | 2.207028 .3075966 7.18 0.000 1.60415 2.809906
3 | 2.134176 .311178 6.86 0.000 1.524278 2.744073
|
X2 |
1 | -.7446537 .1775933 -4.19 0.000 -1.09273 -.3965773
2 | -.5960045 .2927989 -2.04 0.042 -1.16988 -.0221292
3 | -.3294891 .2655665 -1.24 0.215 -.8499898 .1910117
4 | 0 (omitted)
|
X3 |
1 | -.0525713 .1151439 -0.46 0.648 -.2782492 .1731066
2 | -.2162025 .1621985 -1.33 0.183 -.5341058 .1017007
3 | 0 (omitted)
|
X4_cont | -.0047839 .0009392 -5.09 0.000 -.0066247 -.0029431
|
X5 |
1 | .2856962 .1330388 2.15 0.032 .024945 .5464474
2 | -.0495849 .1862612 -0.27 0.790 -.4146501 .3154803
3 | 0 (omitted)
|
X6_cont | -.0409465 .017253 -2.37 0.018 -.0747617 -.0071312
OptOUT | 0 (omitted)
------------------------------------------------------------------------------
Is there an error in my code? Please help!
Thanks,
Amber
I have a DCE with two alternatives and an opt out option. I have coded the opt out option to have all attributes equal 0. I have also created a varirable called 'Opt Out' where if it is an opt out option, Opt out =1, if its alt A or alt B, opt out = 0. I have included this variable within the MNL model
I have used the following code to run an MNL model:
clogit chosen i.X1 i.X2 i.X3 X4_cont i.X5 X6_cont OptOUT, group(OBSID) vce(cluster RID)
clustered by patient ID, and separate observations created using variable OBSID.
However when I run the model, Opt out is omitted and there is no omitted level for X1, despite being dummy coded
See below:
og pseudolikelihood = -751.87905 Pseudo R2 = 0.1734
(Std. err. adjusted for 69 clusters in RID)
------------------------------------------------------------------------------
| Robust
chosen | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
X1 |
1 | 1.948214 .274358 7.10 0.000 1.410482 2.485945
2 | 2.207028 .3075966 7.18 0.000 1.60415 2.809906
3 | 2.134176 .311178 6.86 0.000 1.524278 2.744073
|
X2 |
1 | -.7446537 .1775933 -4.19 0.000 -1.09273 -.3965773
2 | -.5960045 .2927989 -2.04 0.042 -1.16988 -.0221292
3 | -.3294891 .2655665 -1.24 0.215 -.8499898 .1910117
4 | 0 (omitted)
|
X3 |
1 | -.0525713 .1151439 -0.46 0.648 -.2782492 .1731066
2 | -.2162025 .1621985 -1.33 0.183 -.5341058 .1017007
3 | 0 (omitted)
|
X4_cont | -.0047839 .0009392 -5.09 0.000 -.0066247 -.0029431
|
X5 |
1 | .2856962 .1330388 2.15 0.032 .024945 .5464474
2 | -.0495849 .1862612 -0.27 0.790 -.4146501 .3154803
3 | 0 (omitted)
|
X6_cont | -.0409465 .017253 -2.37 0.018 -.0747617 -.0071312
OptOUT | 0 (omitted)
------------------------------------------------------------------------------
Is there an error in my code? Please help!
Thanks,
Amber
Comment