@Vivek-np9vm Hair et al (2019, p 642) noted in Multivariate Analysis book that " Problems with the X2 Test Perhaps the most clear and convincing evidence that a model’s fit is adequate would be a x2 value with a p-value indicating no significant difference between the observed and estimated covariance matrices. Further they discussed that for example, if a researcher used an error rate of five percent, then a p-value greater than .05 would suggest that the researcher’s model is capable of reproducing the observed variables’ covariance matrix-a “good” model fit."
Great one indeed. Have one important question. The fit indices shown here, are they structural model fit? Or measurement model fit? I am under the impression that these are probably the structural model fit. If that is the case, then where can I get the measurement model fit indices? Thank you in advance.
Dear Prof., Somewhere around the 26:00 mark you mention that the residual correlations in the upper triangle may not be larger than 0.5. Can you please tell me the source for that cutoff? In some literature I found that when we look at the standardized residuals (z scores) than we look for values that are greater than |2|. In his book, if im not mistaken, Kline uses the cutoff of 0.1 but there I am not sure do I look for values greater than 0.1 in the residual covariances or residual correlations. Your help would be greatly appreciated. If you do not have time to respond, I understand. Kind regards, Miljan
Hi, thanks for the video! At 6:03, you mention the triangles represent errors, but aren't the little arrows pointing to measured variables the errors? The triangles seem to represent the intercept, if i'm not mistaken.
I'm following the video and I get a message saying "this analysis has ended, probably due to reaching the resource limit" and it doesn't show anything. Can you help me?
Dear Prof., Thanks for nice Video. I've a query. In path diagram, there is a value shows in error term in your video. But in my practice, value of error terms only shows 0.00 each. Why sir? And other values are not same as result by AMOS. What happened Sir? Could share any idea for this problem?
Professor, I thought that the p-value of 0.232 does not indicate that the user model fits the data really well. In fact, a p-value of 0.232 suggests that there is not enough evidence to reject the null hypothesis, which typically means that the user model does not fit the data significantly better than the baseline model. AM i missing something prof? 28:30
here a chi-square p-value less than 0.05 means the model is not appropriately fit. A p-value greater than 0.05 is required to consider the model appropriately fit
But generally when the sample size is large (>200) the chi square end up with p-value less than 0.05 and so not reliable and therefore other model fit indices should be considered
This is so great! Clear and simple and helpful. Thank you so much for all your videos about SEM!
Look forward to the 2nd part.
Coming soon! Please stay tuned in.
Thank you more and more for all the very knowledge sharing that you actually do, Vahid.
@Vivek-np9vm
Hair et al (2019, p 642) noted in Multivariate Analysis book that " Problems with the X2 Test Perhaps the most clear and convincing evidence that a model’s fit is adequate would be a x2 value with a p-value indicating no significant difference between the observed and estimated covariance matrices. Further they discussed that for example, if a researcher used an error rate of five percent, then a p-value greater than .05 would suggest that the researcher’s model is capable of reproducing the observed variables’ covariance matrix-a “good” model fit."
Great one indeed. Have one important question. The fit indices shown here, are they structural model fit? Or measurement model fit? I am under the impression that these are probably the structural model fit. If that is the case, then where can I get the measurement model fit indices? Thank you in advance.
HI Dr Vahid, is there a video for multigroup analysis in Jamovi?
Dear Prof.,
Somewhere around the 26:00 mark you mention that the residual correlations in the upper triangle may not be larger than 0.5. Can you please tell me the source for that cutoff?
In some literature I found that when we look at the standardized residuals (z scores) than we look for values that are greater than |2|.
In his book, if im not mistaken, Kline uses the cutoff of 0.1 but there I am not sure do I look for values greater than 0.1 in the residual covariances or residual correlations.
Your help would be greatly appreciated.
If you do not have time to respond, I understand.
Kind regards,
Miljan
what are the characteristics of your computer?, in my machine it consumes all the resources
Hi, thanks for the video! At 6:03, you mention the triangles represent errors, but aren't the little arrows pointing to measured variables the errors? The triangles seem to represent the intercept, if i'm not mistaken.
@@muhammadrizqisiregar6202 they are not. They are the manifest variable intercepts. Good video, by the way!
Hello.Can PLS-SEM data analysis be used in Jamovi?
I'm following the video and I get a message saying "this analysis has ended, probably due to reaching the resource limit" and it doesn't show anything. Can you help me?
Dear Prof., Thanks for nice Video. I've a query. In path diagram, there is a value shows in error term in your video. But in my practice, value of error terms only shows 0.00 each. Why sir? And other values are not same as result by AMOS. What happened Sir? Could share any idea for this problem?
I am not exactly sure, but it could be due to multicolinearity, or very low variance in your variable.
Dear Professor, Can I ask a copy of the dataset you used in this example so I can follow your demonstrations? Thank you very much.
The dataset is the property of the university.
Professor, I thought that the p-value of 0.232 does not indicate that the user model fits the data really well. In fact, a p-value of 0.232 suggests that there is not enough evidence to reject the null hypothesis, which typically means that the user model does not fit the data significantly better than the baseline model. AM i missing something prof? 28:30
here a chi-square p-value less than 0.05 means the model is not appropriately fit. A p-value greater than 0.05 is required to consider the model appropriately fit
But generally when the sample size is large (>200) the chi square end up with p-value less than 0.05 and so not reliable and therefore other model fit indices should be considered
@@albertbanunle5849 Thank you so much!
Are you going to use moderators? How can I set those variables? Exogenous?
Same question..please share to us
Hi, professor
Can i use categorical variables in SEM analises on Jamovi?
Thank you
Yes, you can.
One of my beta coefficient exceeded 1. What happened? :(
It may be a Heywood case.
see: web.pdx.edu/~newsomj/semclass/ho_improper.pdf
Hi, which version of Jamovi you have, I don't have sem module available in my library
You might want to get the latest version.
@@VahidAryadoust Thank you
Software alternatif for spss and amos please expaline