Latent growth curve modeling in R (January, 2020)

[marcuscrede]

Very useful and clear! Thank you so much.

[serdarsolmaz]

Thank you very much for the valuable information you have provided. It is indeed very insightful. Could you also explain how to plot multiple graphs for continuous variables? Additionally, is it possible to add another variable that predicts the satisfaction variable and then repeat the study?

[chantellewong]

Thank you so much for all these awesome videos, Dr Crowson! I was wondering if you could also make a video demonstrating LGCM where the growth model is non-linear? what can we do if the data do not fit a linear model, or how can we know if the data fit a linear or non-linear model using R??? Thanks a bunch for your help, really appreciate it

[thomasdudek]

Thanks for this video. Unfortunately, the Google links do not seem to work anymore.

[BloopyBlobBob]

How would you interpret the results, specifically the slope and intercept INTERCEPTS, if you had a continuous predictor? Like age? Would the intercept average be the value of your outcome variable when age is 0 or at mean age (if centred)? And the regression with intercept would be how much the outcome variable changes with every unit (year) increase in age at initial status?

Similarly, would the average slope represent the average change over time for individuals with mean age (if centred), and the regression value would be how much the slope changes with every additional year at initial status?

[adeelafarqan]

Your videos are always very helpful, Mike. Can you please tell me the best approach for two wave mediation model analysis? And whether you have already made a video on any method about two wave analysis already? Thanks.

[QaiserPhD]

Dear Crowson, thanks for making such informative and easy-to learn videos. I have recently started learning multilevel mediation and etc, and I found this subject very difficult to understand. Can you kindly make a video(s) on this topic as there is a great confusion and no one has elaborated the execution of these models such as 2-1-1 in simple ways as you usually do in your videos. I would request you to kindly make a video on this important topic and prepare videos for R & Mplus (both). Thanks.

[nkotbs]

Great video, thank you! Liked and subscribed. I find this example very useful in making quick distinction between indicators and regression in context of LGM. I have a question, though. I need to build a bivariate latent curve model, and I understand what it is from reading bollen and curran book p.199. Basically I have two series, y and w, y as you modeled here and another series w that I also want to model as indicators of a different growth process(different intercept and a different slope). My research question is actually concerned with how the latent slope of y and latent slope of w correlates with each other, or rather, "travel" with each other within individuals. Can I then simply specify two intercepts, two slopes. and in regression part regress one slope on the other? I haven't found an example in lavaan package tutorials itself, or any other sources...

My guess of the syntax would be like
model<- '
intercept_y=~ 1*y1+1*y2+1*y3+1*y4
intercep_w=~ 1*w1+1*w2+1*w3+1*w4
slope_y=~ 0*y1+1*y2+2*y3+3*y4
slope_w=~ 0*w1+1*w2+2*w3+3*w4

slope_y~slope_w
'
And if I also want for example slope_y~intercept_w I just add to the bottom of the specifications? Thank you so much for your time.

[hamzayalcin]

Hello
Thank you for your help.
Where is your codes for initial graph?