latent growth curve

Question 1

• fixed effects → describe …
• random effects→ describe

Answer 1

…the data at the group level, pooled across all subjects

• individual subjects deviate from the average starting point and rate of change
  … individuals’ variability around the average starting point and rate of change

Question 2

We may be interested in additional covariates… (2 types)

Answer 2

• that are constant over time → i.e., added at the aggregated level
• that can vary over time → i.e., added at the subject level

Question 3

Latent growth curve modeling

Answer 3

we treat the intercept and slope as
latent variables (i.e., 𝜂1and 𝜂2) and use the
values of time variable 𝑡 as constraints in
the loadings matrix 𝚲

Question 4

explain the parameters in latent growth curve model

Answer 4

eta1= latent intercept = initial starting point

→ mean = average starting point

→ variance = random effect, variance across intercept

eta2 = latent slope

→ mean = average rate of change

→ variance = individual vary in rate of change

covariance = covariance between the variance of the intercept and slope

Question 5

LGC specification rules

Answer 5

1. Loading latent intercept are fixed to 1
2. Loadings of the latent slope are fixed to increasing values from the baseline time point to another
3. observed intercepts 𝝉 are fixed to zero
   and the latent means 𝜶 become
   identified

Question 6

when LGC similar to multilevel regression?

Answer 6

when adding equality constraints between error variances = residual .337… you replicate what multilevel does.