Hello. I am very new to Rstudio and scripts and I am in the process of adapting a syntax so that it can compute an actor-partner interdependence model (if curious about this: https://psycnet.apa.org/record/2005-01893-002) on my data. Basically, I want to add a covariate to my SEM. I realize that researchers add the covariate as another independent variable in the model, but I'm not sure I can do that because the actor-partner model has a distinct look (but if I'm wrong, do let me know). So, I found syntax to help (found here: https://randilgarcia.github.io/week-dyad-workshop/APIM_SEM.html#introduction_to_lavaan under "Add a Covariate: Length of Relationship"). However, I am stuck because Rstudio does not understand the operation ":=" and neither do I. There are a few other commands it doesn't understand (e.g., cx1, cx2, CovXd, CovXa). What would help me is knowing if I have to define these functions and how. Really, any suggestions about how to proceed. I am totally out of my depth. Many thanks!
RStudio is just a way to interface with R. Which R packages are you wondering about? Packages like lavaan have good vignettes. On the :=, that is something that the lavaan syntax will understand. So think of RStudio as just a place to put your code (and it can do it in a few ways) and send this to R, and then R uses functions from the package lavaan (and others). If you are worried about something not working in RStudio, just run in R. RStudio does allow you to do more and organize things, but if you are worried about it you can just skip it.
Good question! I have been using lavaan, installed and loaded, and it didn't understand (" Error in `:=`(covXd, cx1 - cx2) : could not find function ":=" "). Any idea why that might be happening if lavaan an should understand it? It was suggested to me to use Rstudio because it was more accessible and easier, especially for users who are starting out with this program. I can't say I have a handle on R, but I really appreciate your suggestion and will give it a try if I can't resolve it on Rstudio. Thanks!
I'm assuming the := is something within the model formulation part of lavaan. In any SEM formulation you'll need to specify a few different types of relationships between nodes. As far as using R and/or RStudio, I would start just doing simpler stats, go through some intro books on that. Any SEM running and output will have issues that will assume not only do you understand the statistical issues of these procedures but that you are comfortable, to some degree, with the data language (i.e., S/R ... R being an implementation of the language S). Was R covered in your statistics courses?
No coverage of R in my stats courses whatsoever. Only SPSS. I think you make a good point about starting with simpler stats to familiarize myself with R language.
- Badges
- Science topic
- Similar topics
- Thesis Research