Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Advice Sought: -xtmixed-, growth curves and graphs

    Hi,
    I am having a few difficulties plotting a graph of my growth curve model. I am using the -xtmixed- command to estimate both age and cohort differences (+ covariates) on a continuous variable. The model is fine, but I am having trouble plotting the graph. Specifically, I'm trying to produce cohort-specific age trajectories I have had a few attempts at plotting, such as the example below, but as you can see this does not look right. I used the following commands:
    predict
    separate, by(cohort)
    twoway line X X X X

    It looks better if a replace the dichotomous variables (i.e. replace X = 1) before I use the -predict- and -separate- function, but I worry this produces a graph that is not correct (although looks correct)
    Any advice
    Best
    John

  • #2
    What you proposed by fixing the other explanatory variables is similar to what I proposed here: http://www.maartenbuis.nl/wp/inter_q...ter_quadr.html . I retrackted that submission at the time as a new Stata version came out, which included the margins command. Everything I showed in that Stata tip can also been done by margins, so I thought it was superfluous.
    ---------------------------------
    Maarten L. Buis
    University of Konstanz
    Department of history and sociology
    box 40
    78457 Konstanz
    Germany
    http://www.maartenbuis.nl
    ---------------------------------

    Comment

    Working...
    X