@@centerstat And I am looking forward to learning a lot from you in the coming workshops in May and June. A professor of mine recommended the workshops, I checked it out on your website, then watched all the episodes there, now I know for sure I want those workshops :)
@@ThuyNguyen-ji2qy Thanks for your nice comment. We hope you can join us -- remember we have a free 3-day workshop on SEM on May 10-12...see centerstat.org for details. Take care -- patrick
Wow! I am blown away, for such concise, yet super informative content. It is absolutely clear you have been teaching this for a long time given the smoothness of your lecture and selection of MLM/SEM construct explanations.
Thank you so much for your very informative videos. I do have a question. I need to do a power analysis and have no idea where to start for my growth model. Do you have any recommended resources? Thanks in advance!
I just watched all 9 in one sitting. Quick question, which you kind of answered in this video. I do plan on testing children at annual age points, but what do I do about their exact age, is that not useful information that can account for some variance?
Hi Rain -- thanks for the note. I hope you found the videos to be of some use. You're right that exact age can be used in the model, and that is what is commonly done. Instead of use "wave" or "year" as the metric of interest, simply use the child's age at which they gave their observations. So maybe a kid was assessed at age 8, 9 and 11, while another kid was assessed at 12, 13, 15, and 17 -- these ages would simply be the value of time used in the model. It then becomes a missing data problem (so kid 2 is "missing" data at age 16 -- I use quotes because you might not have even asked him for data at age 16), but any SEM package easily handles this. Hope that helps -- good luck with your project -- patrick
Hi Nadira -- thanks for the note. These terms are used very loosely in the literature, and people can use different terminology to refer to the same thing, or the same terminology to refer to different things. It drives you a bit crazy. That said, more often than not the term "latent trajectory analysis" refers to a structural equation modeling-based growth model, primarily because the growth factors are defined as latent variables. However, as I describe in the video, the random effects from an MLM are often precisely the same as the variance of latent factors in the SEM, so this same term could be used in reference to the MLM. Indeed, all "latent" really means is "unobserved", so it can apply to both analytic approaches. I hope that helps -- patrick
If I need to ask questions about MLM application to my research, whom do I ask or through what do i contact, email? Can anyone guide me here, I have some questions regarding the MLM!!
All 9 episode are just amazing. The teacher makes it so clear and accessible. Thank you and thank you.
Hi Thuy -- thanks so much for your very kind note. I'm so glad you found these of some use. Good luck with your work and stay safe -- patrick
@@centerstat And I am looking forward to learning a lot from you in the coming workshops in May and June. A professor of mine recommended the workshops, I checked it out on your website, then watched all the episodes there, now I know for sure I want those workshops :)
@@ThuyNguyen-ji2qy Thanks for your nice comment. We hope you can join us -- remember we have a free 3-day workshop on SEM on May 10-12...see centerstat.org for details. Take care -- patrick
Wow! I am blown away, for such concise, yet super informative content. It is absolutely clear you have been teaching this for a long time given the smoothness of your lecture and selection of MLM/SEM construct explanations.
Thanks so much for your kind comments -- I hope these are of some use in your own work -- patrick
This is a terrific video to wrap up your previous lectures on MLM and SEM! Thank you so much for a very clear explanation.
Fantastic series from beginning to end. Top notch.
HI Greg -- Thanks for your nice comment. Dan and I are really glad you're finding these useful. Stay safe....patrick
Thanks a lot for sharing your amazing knowledge and teach us how to do a good research.
Thank you so much for your very informative videos. I do have a question. I need to do a power analysis and have no idea where to start for my growth model. Do you have any recommended resources? Thanks in advance!
I just watched all 9 in one sitting. Quick question, which you kind of answered in this video. I do plan on testing children at annual age points, but what do I do about their exact age, is that not useful information that can account for some variance?
Hi Rain -- thanks for the note. I hope you found the videos to be of some use. You're right that exact age can be used in the model, and that is what is commonly done. Instead of use "wave" or "year" as the metric of interest, simply use the child's age at which they gave their observations. So maybe a kid was assessed at age 8, 9 and 11, while another kid was assessed at 12, 13, 15, and 17 -- these ages would simply be the value of time used in the model. It then becomes a missing data problem (so kid 2 is "missing" data at age 16 -- I use quotes because you might not have even asked him for data at age 16), but any SEM package easily handles this. Hope that helps -- good luck with your project -- patrick
Thank you!
What is the difference between these two models and latent trajectory analysis?
Hi Nadira -- thanks for the note. These terms are used very loosely in the literature, and people can use different terminology to refer to the same thing, or the same terminology to refer to different things. It drives you a bit crazy. That said, more often than not the term "latent trajectory analysis" refers to a structural equation modeling-based growth model, primarily because the growth factors are defined as latent variables. However, as I describe in the video, the random effects from an MLM are often precisely the same as the variance of latent factors in the SEM, so this same term could be used in reference to the MLM. Indeed, all "latent" really means is "unobserved", so it can apply to both analytic approaches. I hope that helps -- patrick
@@centerstat thank you Patrick
If I need to ask questions about MLM application to my research, whom do I ask or through what do i contact, email? Can anyone guide me here, I have some questions regarding the MLM!!
There's a great online listserve at www.jiscmail.ac.uk/cgi-bin/webadmin?A0=multilevel -- I'd start there.
@@centerstat Thank you. Is there any forum you recommend Forensic Psychology Postgrads?
@@mish7992 I'm sorry, but I don't....