Intro to Structural Equation Modeling Using Stata

Chuck Huber, PhD with StataCorp presents on conducting statistical analyses using Structural Equation Modeling (SEM) during the USC Interdisciplinary Speaker Series.
Chuck Huber is a Senior Statistician at StataCorp and Adjunct Associate Professor of Biostatistics at Texas A&M. In addition to working with Stata's team of software developers, he produces instructional videos for the Stata UA-cam channel, writes blog entries, develops online NetCourses and gives talks about Stata at conferences and universities. Most of his current work is focused on statistical methods used by psychologists and other behavioral scientists. Dr. Huber currently teaches introductory biostatistics and previously taught introductory biostatistics,
categorical data analysis and statistical genetics at Texas A&M. He has published in the areas of neurology, human and animal genetics, alcohol and drug abuse prevention, nutrition and birth defects.

[@muniirahmbabazi9965]

Thanks a lot. Wonderful introduction to SEM. the best i have watched.

[@johnpumphrey7655]

This is a great presentation by Chuck Huber on SEM and Stata. I'm a novice at both and have been struggling. This two hours was well worth the time investment. Thank you to USC for making this available online and thanks to Chuck for the wonderful presentation. // John in Singapore.

[@kizmanryce]

I5i554eee9u🐰

[@nminhas786]

Hats off for you, Sir, such a nice way of teaching a very complex subject and transformation into an easier way.

[@eiko6224]

Great presentation with clear buildingblock and how-to-do skills

[@doodnathpersad]

Hi Professor Huber. Excellent presentation! I am really impressed with what one can do with the GSEM function, particularly as it relates to multivariate multilevel modeling. My question is: Is it possible to use GSEM to estimate the covariances of the first level residuals (in a 2-level multivariate model) when both dependents are ordinal ....so that the the estimated variance of each of these residuals is pi squared on 3. The dependent variables (students' math and English grades) are correlated so that one would expect the first level residuals to be correlated. Schools represent the second level in the model.

[@beatazukowska4754]

Great workshop, clearly presented practical knowledge.

[@diwang6773]

Thank you. A nice video.

[@macrs8459]

Thank you! VERY helpful!!

[@faridrezanajafi9565]

Very informative. Thanks

[@pegahch1315]

great video thank you so much

[@vikrantkkamble]

how does SEM take care of omitted variable bias? Also simultaneous bias such as, funding affects GPA but the GPA also affects funding?

[@wubishetmengesha6977]

Thank you very much it is helpful

[@abelfekadudadi7003]

Hi Dr. chuker Huber; I am a PhD student at Flinders University, Australia and I want to be clear with one point. I tried to fit SEM to my data but I found a lot of exogenous and mediator variables that made my path diagram very complicated and I want to know if it is possible to fit first an ordinary model and use variables which are only significant in the final model for building my SEM. Or if there are any other options used to reduce these variables.

[@nishakarthakur8540]

Very helpful...than you so much

[@okwaterdog]

Very informative! I would very much appreciate it if someone let me know how I can find the “cair.dta” that is used in this lecture? I’d like to exercise ;)

[@user-wg9un3gb9p]

. .. Thank you very much

[@Katrina86]

If I include only observed variables in simultaneous equations model (no latent variables), can I call it structural equation modelling or just pass analysis?

[@mohrezah4924]

TQVM was useful, if have any video about Moderate &/or Mediate PLS shair it

[@giovanijrodriguez5267]

What is the name of the textbook he suggested on minute 27:21, I barely understood what he said.

[@mayadabenslama9299]

books.google.com.sa/books/about/Principles_and_Practice_of_Structural_Eq.html?id=mGf3Ex59AX0C&redir_esc=y

[@mimosia1191]

Hi ,
I have one dependent variables called ( IWFP billion m3 ) for 3 crops.
and
4 independent variables ( RWr billion m3 year-1 + Ta Celsius +
precipitation mm + productivity kg/ha for 3 crops) Yearly data for a
country
Observation 19 years
I
want to do the analysis and see the effect of independent variables,
Please any suggestions, which is the perfect model I can apply and
dealing with ?? , thank you in advance

[@nesrineb21]

Great lecture sir! is it possible to have access to that big zip file, seems it is nolonger there ..

[@eyohan6944]

Thanks

[@sainaveenbali]

how can i do if my data is panel

[@swtgirlofmine]

make a 'panel data set' and input it on Stata.

[@deeg1385]

also wanted to ask the same question

[@marouaneidmansour8276]

great chuck... ok with @Ellen Almirol, sound a bit of an issue, but all good.

[@ellenalmirol8060]

Great presentation. The guy who scoffs or laughs in the background is sort of annoying, but the speaker did very well on a very complicated matter.