Hi, all,
My dependent variable is self-rated health, measured on a 5 point likert
scale. It's measured repeatedly 4 times over a 9 year period.
The IV I am interested in is SES (education, income, and occupational
class). Control variables are age, gender, marital status, health behaviors,
BMI, and chronic conditions.
All those variables are measured 4 times, some of them are variant (income,
age, health behaviors, BMI, and chronic conditions) and some are non-variant
(gender, education, occupational class, marital status).
I am testing how SES predicts health outcome changes. It seems that both
latent class analysis and mixed effects model could answer that questions.
Just wonder which one is better and can handle missing data better.
Thanks
--
Yawen Li
Ph.D candidate
School of Social Work
University of Southern California
MRF 347, 669 W 34th St. Los Angeles
Tel: 213-740-1391
Email:
[hidden email]
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
