Good presentation. I don't agree with your point of "more information is always better" though. There are studies showing that after certain circumstances a bigger sample is not necessarily better for SEM.
I see there are lots of cases in which SEM + Artificial Neural network is used as a balanced approach to address real world complexity. Please make a tutorial to understand the procedure, i am not able to understand how Neural network is used since its a supervised learning technique and we do not have the numbers for Latent variable here so how can we train the model?
I'm not 100% sure here exactly what you are describing, but you can get an estimated value for the latent means in lavaan if you wanted to use those to "train" a neural network model, depending on your goals.
@@StatisticsofDOOM Hey, thank you for your reply, could you please let me know functions which can help me to get the estimated latent trait? Yes, we can use it for training our Neural network/machine learning model but why anyone wants to predict those values from ANN/ML model if we can calculate it using our Psychometric Assessments?
Good presentation. I don't agree with your point of "more information is always better" though. There are studies showing that after certain circumstances a bigger sample is not necessarily better for SEM.
A recursion between two or more variables is quite often the result of simultaneity, i.e. the effect is reciprocal
Thanks Dr Erin for all your presentations, very instructional and well designed.
Glad to help!
This is one of the first SEM explanations I actually kind of understand haha thankyou!!!
Glad to help!
Thank you so much for your presentation :) Your video helps me a lot when
revising my SEM !
I see there are lots of cases in which SEM + Artificial Neural network is used as a balanced approach to address real world complexity. Please make a tutorial to understand the procedure, i am not able to understand how Neural network is used since its a supervised learning technique and we do not have the numbers for Latent variable here so how can we train the model?
I'm not 100% sure here exactly what you are describing, but you can get an estimated value for the latent means in lavaan if you wanted to use those to "train" a neural network model, depending on your goals.
@@StatisticsofDOOM Hey, thank you for your reply, could you please let me know functions which can help me to get the estimated latent trait?
Yes, we can use it for training our Neural network/machine learning model but why anyone wants to predict those values from ANN/ML model if we can calculate it using our Psychometric Assessments?
@@sourabhsharma9830 You can use lavPredict(fitted.model) to get the latent means.
SOOOO HELPFUL.