I am just beginning to work on longitudinal research and I am using Mplus for latent growth modelling.
I am trying to understand the meaning of standardized intercepts (I) and slopes (S) in my model. What are they indicating and what are they pertaining to?
Could you please shed some light on this matter?
Thank you very much.
Witold,
First, don't use standardization with this model if you can avoid it.
The Intercept and Slope latent variables have means, variances, and a covariance (or a path). If you plot from the means using the old middle-school y=mx+b, you'll be the average (across respondents) model-implied best-fit line. The variances indicate how much individual variation there is in the person-specific intercepts and slopes of the best-fit line, assuming the latent intercept and slope are normally distributed. These are in the units of the original measurement.
Hope that helps,
Pat
You might find this tutorial helpful: http://oregonstate.edu/dept/hdfs/papers/lgcgeneral.pdf
Witold,
First, don't use standardization with this model if you can avoid it.
The Intercept and Slope latent variables have means, variances, and a covariance (or a path). If you plot from the means using the old middle-school y=mx+b, you'll be the average (across respondents) model-implied best-fit line. The variances indicate how much individual variation there is in the person-specific intercepts and slopes of the best-fit line, assuming the latent intercept and slope are normally distributed. These are in the units of the original measurement.
Hope that helps,
Pat
@Patrick - Why should I avoid standardization? What is the drawback?
Thanks for clarifying
Witold
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics