Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

JASP: mediation/moderation analysis & structural equation modeling for Bayes?

Hello everyone!
I teach postgrad psychology stats and I am wanting to include JASP and Bayes beyond a basic level in the course. The programme director is a big fan of MM/SEM and bootstrapping, s/he seems to think that this in combination with bootstrapping is the ONLY way forward in the replication crisis. I‘ve struggled to find papers or textbooks supporting this and I‘d quite like my students to learn Bayesian modeling...is there a way of running th above analysis in JASP?

Thank you so much,

eniseg2

Comments

  • Hi eniseg2,

    Right now we only have "classical" SEM in JASP; we will of course include the Bayesian echo at some point (through blavaan, probably).

    Cheers,
    E.J.

  • Hi E.J.,
    thanks for the quick response! How do I do SEM in JASP? I've had a brief look but was unsure how to run the analysis. I don't have questionnaire data in my data set so I can't run the factor analyses. How does it work?

    Thanks so much!
    eniseg2

  • Basically you just copy-paste Lavaan code. We are working on a video and gif to demonstrate the process.

  • I've never used Lavaan before. Would it be copy-pasting this code but substitute with my own variable names?
    http://lavaan.ugent.be/tutorial/sem.html

    I'm teaching Advanced Quantitative Methods and I've been using JASP for almost everything so far. The SEM module for SPSS doesn't run on Mac which most of my students use. And R isn't an option, sadly.

  • Hi eniseg2,
    you have already looked here?
    forum.cogsci.nl/index.php?p=/discussion/3507/sem#latest

    Cheers,
    MAgoJ

  • Hi, MAgoJ,
    the lavaan tutorial I posted is exactly the one provided in your other pos, haha. Thank you!

  • Unfortunately, it seems to not be working. I copy-pasted the lavaan code for the Bollen example and it is not running or providing results. :(

  • Hi, eniseg2.

    In the textbox you have to insert only the construct that will be passed to the model type chosen in model name.

    Do not insert:
    model <- ' '

    Insert:
    ind60 =~ x1 + x2 + x3
    dem60 =~ y1 + y2 + y3 + y4
    dem65 =~ y5 + y6 + y7 + y8

    dem60 ~ ind60
    dem65 ~ ind60 + dem60

    y1 ~~ y5
    y2 ~~ y4 + y6
    y3 ~~ y7
    y4 ~~ y8
    y6 ~~ y8

    At the end press CTRL + Enter and you will see the results based on what you have selected below.
    Attached image.

    Best regards.
    MAgoJ

  • Hi MAgoJ,
    I'm really sorry, but I copy-pasted what you wrote and it still does this:

    SEM.png 188.1K
  • Hi, eniseg2.

    I'm sorry, but the copy-paste in this object (listbox) I found that it can create problems (report to @EJ).
    I do not copy-paste, but I prefer to type the syntax for the selected model directly in the listbox.
    Do this by typing in the listbox and let me know.

    Best regards.
    MAgoJ

  • Hi,
    I have typed it out and no results. It doesn't change anything at all. I am not sure what is going on? :((

    sem.png 350.1K
  • Hi eniseg,
    I use jasp with win7 x64, so I asked a colleague to try with his MAC.
    A colleague who has MAC, brought me the same problem.

    It could be a bug related to the version of jasp for MAC.

    Best regards.
    MAgoJ

  • Thanks for bringing this to our attention. We now have a team member working to upgrade this functionality.
    Cheers,
    E.J.

Sign In or Register to comment.