|
|
 |
| Multiple group analysis within GGMM |
 
|
| Message/Author |
|
|
|
|
Hi,
I'm trying to conduct a multiple group analysis for gender within GGMM.
When I ran separate GGMM's for males and females, the best-fitting solution for males was 4-class model, and the best-fitting solution for females was 3-class model.
What I want to do is to compare two analogous(similar-looking) trajectories from males and females in a multiple group analysis.
For example, from 2 separate GGMM's, class #4 of male and class #3 of female seem to be very similar in developmental pattern, and I want to examine whether their intercept and slope are significantly different.
I tried to start from example 8.8(GMM with known classes), but couldn't figure out how to write an input when I have different number of classes for males and females.
Would you please provide tips for writing up the input? |
|
|
|
|
| I would just look at these qualitatively and not try to do a statistical test unless both males and females have the same number of classes and the same classes. |
|
|
| Carolin posted on Friday, July 06, 2012 - 12:33 am
|
|
|
Hello,
I have a GMM with 4 classes for girls and 3 classes for boys. I would like to compare them with regard to significant differences of intercept and slope.
With regard to your answer above: will it never be possible to do a multiple group analysis with different numbers of classes?
If so, do I have other possibilities to compare them?
Thanks,
Carolin |
|
|
|
|
| From a statistical point of view your analysis plan is very tricky and I, personally, have never seen it in a publication. You would need the same number of classes and essentially the same trajectory classes (with regard to intercept, growth rate, and number of growth factors) in each subgroup respectively. If I were you, I would stick to qualitative descriptions for each subgroup (i.e., male/female) and I would try to explain (on a sound empirical basis) why you found 3 trajectory classes for boys and why there were 4 for girls. Especially the latter finding sounds very interesting and may have the potential for a nice discussion. See Wiesner, Weichold, & Silbereisen, 2007 for a similar examination of gender-specific GMMs for alcohol use. They also compared gender specific GMMs on a qualitative basis. |
|
|
|
|
| uupps, I meant "theoretical basis" not "empirical basis")... |
|
|
| Carolin posted on Friday, July 06, 2012 - 7:22 am
|
|
|
| Thanks a lot for your opinion. That was my idea before submitting my manuscript... but the reviewers critizised that I did only qualitative descriptions for girls versus boys and suggested to do a multiple group analyses... That is the reason I asked the question... |
|
|
| Carolin posted on Monday, July 09, 2012 - 5:01 am
|
|
|
One related question:
... so I can answer to the reviewer that a multiple group analysis is not possible (or not advisable) with different numbers of classes in girls and boys?
Thanks a lot again! |
|
|
|
|
| If girls and boys have classes that are very similar, you could compare those classes using multiple group analysis. But if the classes for girls and boys are different substantively, it would not make sense to compare them. |
|
|
| Carolin posted on Thursday, July 12, 2012 - 5:15 am
|
|
|
What do you mean when you say "very similar"? Does that mean the values of mean intercept and slope should be similar?
And what about the different number of trajectories in girls(4) and boys (3)? Will that be a problem?
Thanks |
|
|
|
|
| I mean that the trajectories of the classes have the same meaning for boys and girls, for example, start high and decline by a similar amount. You would compare only the trajectories that are similar for boys and girls. You would not compare all trajectories. |
|
|
|
|
Dear Carolin,
you might find the following paper useful:
Schaeffer, C.M., Petras, H., Ialongo, N., Masyn, K.E., Hubbard, S., Poduska, J., & Sheppard, K. (2006). A comparison of girl's and boy's aggressive-disruptive behavior trajectories across elementary school: Prediction to young adult antisocial outcomes. Journal of Consulting and Clinical Psychology, 74, 500-510.
download paper contact first author show abstract
Abstract
"Multiple group analysis and general growth mixture modeling was used to determine whether aggressive– disruptive behavior trajectories during elementary school, and their association with young adulthood antisocial outcomes, vary by gender. Participants were assessed longitudinally beginning at age 6 as part of an evaluation of 2 school-based preventive programs. Two analogous trajectories were found for girls and boys: chronic high aggression– disruption (CHAD) and stable low aggression– disruption (LAD). A 3rd class of low moderate aggression– disruption (LMAD) for girls and increasing aggression– disruption (IAD) for boys also was found. Girls and boys in analogous CHAD classes did not differ in trajectory level and course, but girls in the CHAD and LAD classes had lower rates of antisocial outcomes than boys. Girls with the LMAD trajectory differed from boys with the IAD trajectory."
hide abstract
Best,
Hanno |
|
|
| Carolin posted on Friday, July 13, 2012 - 7:54 am
|
|
|
| Thanks a lot Linda and thanks a lot Hanno, you are very helpful! |
|
|
| Back to top |
|
|
