Structural equation modeling (SEM) with lavaan

Quite disorganized, regularly jumps over lines of code and halts explanations mid sentence. Additionally refers to "last week" and other indications that this course was copied from a different source and not designed for Udemy. Overall, I was disappointed.

Excellent !!!, the best virtual course I have ever taken in my life. Clear explanations, concrete and didactic examples. I recommend it 100%.

I watched about half of the course over 2 days (I used 1.75 speed and skipped parts by fast-foprwarding 5sec at a time, since the instructor speaks slowly and moves around with the mouse/scrolls up and down and does things that aren't necessary to follow step-by-step). Overall, I would not recommend this course. It leaves a lot of open questions about SEM and does not explain SEM thoroughly enough. The more I watched, the less I seemed to learn and to understand... It is clear that the instructor knows some stuff about SEM and lavaan() in R. He at least knows how to run SEM and interpret the parameters at a basic level (how to interpret results; what is good or bad, what is significant or not). But there are many things the instructor does not seem to know or understand properly. And when it's clear he doesn't know or understand what he is talking about, he simply skips those parts and moves on to a different topic. That is pedagogically useless and a very bad example of presenting or studying anything. Similarly, the instructor corrects himself so often that at some point, I started doubting whether any of what the instructor said is actually correct. All these self-corrections confused me. It is also clear that this course was not prepared to be a Udemy course ("We looked at this a week ago", "Which we will look at in two weeks, which is our last session", "Someone asked a question" ??). It's most likely an online course that was recorded (e.g., as part of a university course) and then simply cut into pieces and uploaded to Udemy. Smart way, in some way, to make extra money. But this approach really is inferior and might, in the end, only reflect badly upon the instructor because it's clear that for this Udemy online course, no extra effort or work was expanded. Now, a few things I liked: The instructor explained everything that he explained in detail and fairly clearly. I like that we were able to download all course material (R scripts, data). It's great for people who know very little about R, because the instructor shows lots of things about how to use/navigate RStudio. And lots of things I did not like: Many things are mentioned without any explanation for why this might be useful at all (e.g., why name different parementers the same, why force different parameters to be the same, why or when to fix the latent variable to 1 or one of the parameters that estimate the latent variable to 1?). As an Intro to SEM, one should much more clearly explain all the different model fit parameters -- what's the difference and why use so many different fit indices (akaike, BIC, CFI, TLI and whatever else there is...) -- and what are all the diffferent options/arguments to the sem/cfa functions used for, what do they do? Also, certain terms are used that someone without any knowledge about SEM wouldn't understand. What's LISREL? What are all the results of the inspect() function, how do I interpret this? Why rotate a factor analysis, what does this mean and do, and what do the different rotation types (varimax, promax, cfq, bifactor) do? The instructor did not seem to know himself... The course shows so many things unrelated to SEM, i.e., how to use R, what R can do, what standardization is, how to write R code etc. A course like this "Intro to SEM in R" should assume and require that people already know R and understand R code, otherwise SEM is way too advanced anyways... And with this assumption, the course could've been cut down by one fourth or so. What I also did not like was the repetition. The instructor explained the semPaths function and how to interpret this plot graph three or four times. This was not very useful when you don't even have the basic understanding of the raw SEM output (paremeters) yet. The course structure was also confusing. First, there is something about factor analysis, then there is something about a full SEM model, then again something about (confirmatory) factor analysis.

Slides are presented too much information all at once. Too much stuttering. Needs concrete examples to accompany the abstract topics

This course is aweful. The instructor on multiple occasions starts to explain a concept, realises he can't and moves on to something else. Don't waste your money

Not a well structured course. I have a good background of SEM and have worked with many GUI based software in the past, I came here to understand Lavaan better. Was disappointed by his lack of articulation ability and scattered way of jumping around topics

I found the notes on the R files useful. However, there is a great deal of improvisation and going in circles during the lectures which makes it difficult to follow sometimes. It would be easier for students to follow if the lectures were linear, with as few improvisations and self-corrections as possible. Some suggestions that might imrpove the experience. At the very beginning of the course (Section 1, video 2) it would be nice to define/explain keywords (e.g., PLS, exogenous/endogenous variables, reflective measures). It would be nice to suggest further readings and references (e.g., reference for cut-off threshold in modification indices).

The slides that professor incorporates with the video makes it easier to follow along. The reviewer can learn and understand the concepts better.

The versions of the libraries shown have changed. Author says what items are but fails to tell the "so what" about them.

Worst course I ever bought on Udemy. Many commands don't work. Very poor didactics. In short, very bad course.

This course has taught be so much about the Lavaan package, and it definitely did helped me a lot in my Msc thesis. The latter lectures seem abit rushed, but I would still recommend this to others who are keen to experiment with covariance based SEM modelling in R :-)

The instructor provided very hands-on skills using lavaan while conveying theoretical aspects of SEM. Very Helpful!

I am particularly interested in how and to what extent modification indices should be used. This was treated quite well although a bit briefly (maybe there will be more in subsequent lectures). I also wanted to see how experts in lavaan deal with situations when one gets a warning that some covariances are negative. The lecturere had a very good example on this. Right now I am going quickly through imputation of missing values. My data sets did not have many missing variables so this does not interest me so much at present, but the treatment was very good and I'll surely be coming back to it when I would have data-sets with more missing values.

So far this course is exactly what I need - intro to R and SEM. Given that I'm working on a project that requires SEM & R and I haven't used either [I have used SAS and done CFAs etc] this is a wonderful blend of instruction and application.

I found this course very useful and practical. I've learned a lot. Professor Hubona explains everything clearly and patiently. He also answers questions very quickly. I totally recommend this course!