Function for bootstrapped confidence intervals of standardized solution?

496 views
Skip to first unread message

David Disabato

unread,
Jun 26, 2021, 3:26:55 PM6/26/21
to lavaan
Hi lavaan group,

I was wondering if `lavaan` or `semTools` has created a function for automating bootstrapped confidence intervals for the standardized solution from SEM models estimated using `lavaan`. I noticed the standardizedSolution() function does not have a `boot.ci.type` argument like the parameterEstimates() function does. The standardizedSolution() help page does not mention bootstrapped confidence intervals and says if the `ci` argument is TRUE, that "simple symmetric confidence intervals are added to the output". Thus, while the standardizedSolution() function can provide confidence intervals of the standardized solution, I assume this means it can not provide bootstrapped confidence intervals. 

I noticed there is a post from 2014 that shows a way to do this using the bootstrapLavaan() function: https://groups.google.com/g/lavaan/c/pIXy7NUa3c0/m/hVPH3zClaYcJ. However, it has been 7 years since then and I was wondering if lavaan, semTools, or any other R package has created a function for automating the calculation of bootstrapped confidence intervals of the standardized solution?

If not, I guess you could treat this post as a feature request. I think this would be of interest to a lot of users as social science researchers are more and more interested in having 1) more interpretable (e.g., standardized) effect sizes, 2) confidence intervals around those effect sizes, 3) robust versions (e.g., bootstrapping) of such confidence intervals.

Best,
-David

Terrence Jorgensen

unread,
Jul 8, 2021, 5:41:23 AM7/8/21
to lavaan
I was wondering if `lavaan` or `semTools` has created a function for automating bootstrapped confidence intervals for the standardized solution from SEM models estimated using `lavaan`

I just added that option to semTools::monteCarloCI().  You can access it in the development version:

devtools::install_github("simsem/semTools/semTools")

Monte Carlo CIs are a parametric bootstrap, which is much less computationally intensive.   The help page has a reference you can read.

Terrence D. Jorgensen
Assistant Professor, Methods and Statistics
Research Institute for Child Development and Education, the University of Amsterdam

HG Wells

unread,
Jan 24, 2022, 11:21:20 AM1/24/22
to lavaan

Hello Terrence,

I am using the semTools::monteCarloCI(). function, it is great (very fast... !) Thank you very much for it.

I was wondering, how to get standardized estimates (the function seems to automatically return non-standardized versions from a lavaan fitted object)?

Emanuele

Christian Arnold

unread,
Jan 24, 2022, 5:30:30 PM1/24/22
to lavaan
Hi Emanuele et al. I will shortly post an application for testing. It should be able to solve this problem and similar problems (hopefully). The coefficients are bootstrapped, but I think you can do what you want to do. Best. Christian 

Emanuele Fino

unread,
Jan 24, 2022, 6:28:54 PM1/24/22
to lav...@googlegroups.com
Thank you, I really appreciate it.

Emanuele


--
You received this message because you are subscribed to a topic in the Google Groups "lavaan" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lavaan/qQBXSz5cd0o/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/a07a6c5b-04d3-4b32-ad99-10d3a7a4b7d3n%40googlegroups.com.

Christian Arnold

unread,
Jan 24, 2022, 7:04:22 PM1/24/22
to lavaan
see: https://groups.google.com/g/lavaan/c/0RSsh4M6zQg

Feedback is very welcome :-)

Ranaivo Rasolofoson

unread,
Jan 24, 2022, 7:17:18 PM1/24/22
to lav...@googlegroups.com
Thank you so much for this Christian. It will be super useful for many.
Ranaivo 

You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/cffae065-6d66-4c32-bc00-b835b40123f4n%40googlegroups.com.
--
Ranaivo A. Rasolofoson, PhD, MS
Nicholas School of the Environment
Duke University

Gund Affiliate, Gund Institute for Environment
University of Vermont  (https://www.uvm.edu/gund)


HG Wells

unread,
Jan 25, 2022, 5:25:10 AM1/25/22
to lavaan
Hello, thank you so much!!!

This is so helpful, thank you for doing this.

I've just run it on some data and it worked well, this is much appreciated :)

Emanuele

Christian Arnold

unread,
Jan 25, 2022, 3:41:42 PM1/25/22
to lav...@googlegroups.com
Thank you for your feedback :-) 

Best

Christian

From: lav...@googlegroups.com <lav...@googlegroups.com> on behalf of HG Wells <det...@gmail.com>
Sent: Tuesday, January 25, 2022 11:25:10 AM
To: lavaan <lav...@googlegroups.com>
Subject: Re: Function for bootstrapped confidence intervals of standardized solution?
 

Terrence Jorgensen

unread,
Jan 25, 2022, 6:24:54 PM1/25/22
to lavaan
I was wondering, how to get standardized estimates (the function seems to automatically return non-standardized versions from a lavaan fitted object)?

In the development version, you can simply set standardized = TRUE.  Find instructions how to install it here: https://github.com/simsem/semTools/wiki

This will be much faster than bootstrapping.

Emanuele Fino

unread,
Jan 25, 2022, 8:06:58 PM1/25/22
to lav...@googlegroups.com
Hello, thank you so much! Very helpful and much appreciated :)
Emanuele

--
You received this message because you are subscribed to a topic in the Google Groups "lavaan" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lavaan/qQBXSz5cd0o/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lavaan+un...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages