vaccinations- combining variables [message #2598] |
Mon, 21 July 2014 10:08 |
Lisa
Messages: 7 Registered: July 2014
|
Member |
|
|
For a multinomial logistic regression analysis, how could one combine 8 separate variables into one main variable, with outcome responses stating they have received full vaccinations, partial vaccinations and no vaccinations? The variables that need to be combined are H2,H3,H4,H5,H6,H7,H8,H9
|
|
|
|
Re: vaccinations- combining variables [message #2604 is a reply to message #2603] |
Tue, 22 July 2014 10:03 |
Bridgette-DHS
Messages: 3203 Registered: February 2013
|
Senior Member |
|
|
Thanks for letting us know. However, following is a response from Senior DHS Specialist, Tom Pullum (as it might help another user):
One possibility would be to construct a variable "sum_imm", for example, which would be the sum of the "yes" responses to these variables. (You may also want to include H0.) The sum could be 0, or "some" yes's, or ALL yes's. You could analyze with multinomial logit or ordered logit (I would recommend multinomial).
|
|
|