Although I feel more comfortable with SPSS-AMOS, this video surely changes my perspective about SmartPLS. I will explore it further. Thanks Prof for sharing your experience
Wow! Your tour of the introductory features has me jumping for joy. It will be so nice to use this for teaching non-coding students how to do SEM modeling of both PLS and CB-based methods. I'm excited!
Hi Dr. Gaskin I read a paper that had a discriminant validity issue between two constructs according to Fornell & Larcker’s inter construct correlation table. To further test discriminant validity the authors stated that they “compared the covariances between models that either freely estimated the covariance between the two constructs or constrained it to one (Wright et al. 2012). The chi-square differences between these models were significant. These results support the discriminant validity of the constructs.” I'm struggling to comprehend this (perhaps a blonde moment). Could you share your thoughts?
The most current approach to discriminant validity is proposed by Ronkko and Cho 2022. This approach requires estimating 95% confidence intervals for latent factor correlations during CFA. If the absolute value of the upper limit (or lower limit for negative correlations) of the correlation with an-other factor is < 0.800, < 0.900, < 1.00, or = 1, it indicates no discriminant validity concern, or there is a marginal, moderate, or severe concern, respectively. Rönkkö, M., & Cho, E. (2022). An updated guideline for assessing discriminant validity. Organizational Research Methods, 25(1), 6-14. doi.org/10.1177/1094428120968614
Prof. Gaskin, thank you very much for your effort and very constructive videos. I have one question and your help would be much appreciated. Could you please explain how to perform moderation analysis with moderating slope? Unlike PLS-SEM model variant, CB-SEM does not have this built-in option (or at least I could not find it). Thank you!
I've asked Christian Ringle and he says they're planning to include these extra features over the next few rollouts. I've specifically requested simple slopes plots, J-N plots, and MGA. So, for now though, you'll still need to use the PLS side to do it.
@@Gaskination Thank you very much for your answer. Is it possible to use the same approach that you demonstrated in your video on Interaction moderation in Amos (introduction of the new interaction variable)? If so, could it be used when we have one dependent, one independent, three mediators and two moderators (should moderators be analyzed separately in two models, or jointly in one)?
@@aleksadokic1872 Yes, you could create a new interaction variable outside of SmartPLS and then model it in SmartPLS. The moderators can be included in the same model.
Hi Dr. Gaskin I am wanting to know if SmartPLS provides modification indices? I am working on my measurement model and I have been reading to look beyond standardized factor loadings and to also consider modification indices as I have read that they can be indicative of spurious correlations with other non-related measurement items and decrease model fit. Any idea where I could find this or how I could work it out? Thanks!
@@Gaskination Thanks. I have been reading about modification indices and variance inflation factor (VIF). I see that VIF is possible to inspect when running a PLS model. My question is this: When running CB-SEM, do we need to inspect VIF as well, and if so, how can we do that?
Thanks for the video. I have a sample composed of two categories, say males and females, when I test a certain path using SmartPls cb-sem it gives insignificant results in both categories, surprisingly, when I test the same path for the whole sample it gives significant result!!! This doesn't appear when I use Amos (in Amos the path is insignificant for the whole sample too). If testing structural model is not yet reliable in SmartPls cb-sem, can I use it for CFA only and then test hypotheses using Amos? Given that the two softwares give very close results regarding CFA, but SmartPls is just more user-friendly
If the path results are non-significant when split by groups, it may be due to sample size issues. You can use different software for different parts of the analysis as long as it is reported. Some software is certainly more userfriendly or useful than others.
Thank you Dr. Gaskin! Your videos are really helpful. I have a question if you don't mind. Could I use the CB-SEM feature in SmartPLS to conduct an EFA? I should also mention that my model may potentially have second order factors. I'm not sure how I would test the need for second order factors in an EFA. Any insight you can provide would be greatly appreciated. Thank you.
You can actually do it in the PLS model (rather than the CB-SEM). Look at the "crossloadings" matrix in the Discriminant validity section of the output when running the regular pls algorithm. As for higher order factors, I have a few videos on validating those. Here is a playlist with those videos: ruclips.net/p/PLnMJlbz3sefIJoPbCfTDwzOgMKR3IHL5x&si=7R3guB1XAMEkaVVj
@@Gaskination Dr. Gaskin I'm sorry to bother you again but I want to be sure I understand. I can perform an EFA and subsequently test hypotheses in smartpls?
The CFA should include all latent reflective factors, even mediators and moderators (if they are reflective and latent). The CFA should include all these factors at once, rather than one at a time. This allows us to test for model fit and discriminant validity.
One more question. I would like to use SmartPLS to calculate my CB-SEMs. How do I set my estimation procedure (preferably maximum likelihood)? Or is that always preset?
Hi this result is looking great really! I used Amos and diagram cant understand. Can I draw the amos diagram by hand? Amos created result is very unclear and I cant read properly many numbers
The primary criteria are whether you have formative factors (choose PLS) or low sample size (choose PLS) or exceptionally complex model with interactions (choose PLS). Otherwise, it is currently considered best practice to choose CB-SEM.
SmartPLS is working on moderation right now. They hope to have it completed by January. In the meantime, you can toggle over to the PLS side for interactions, or you can produce those interactions in Excel/SPSS yourself (standardize variables and then multiply), and then bring the dataset back into SmartPLS, or you can run them as grouping moderators instead (high/low if necessary) and run the model with the two different datasets (one for each group) and then compare those models. It is tedious and manual, but should work. I'm anxiously awaiting the update to SmartPLS to include these features on the CB-SEM side.
Although I feel more comfortable with SPSS-AMOS, this video surely changes my perspective about SmartPLS. I will explore it further. Thanks Prof for sharing your experience
Wow! Your tour of the introductory features has me jumping for joy. It will be so nice to use this for teaching non-coding students how to do SEM modeling of both PLS and CB-based methods. I'm excited!
Thank you, James, for presenting this excellent presentation on the CB-Sem in SmartPLS4.
Your videos are extremely useful and your enthusiasm is contagious! Thank you, Sir.
Prof. Gaskin, You are just amazing with your videos. Thanks
Powerful one Prof. Gaskin. Thank you for this video
Wow. after this version of PLS it will be number one SEM software
Thank you for the update.
Hi Dr. Gaskin
I read a paper that had a discriminant validity issue between two constructs according to Fornell & Larcker’s inter construct correlation table.
To further test discriminant validity the authors stated that they “compared the covariances between models that either freely estimated the covariance between the two constructs or constrained it to one (Wright et al. 2012). The chi-square differences between these models were significant. These results support the discriminant validity of the constructs.”
I'm struggling to comprehend this (perhaps a blonde moment). Could you share your thoughts?
The most current approach to discriminant validity is proposed by Ronkko and Cho 2022. This approach requires estimating 95% confidence intervals for latent factor correlations during CFA. If the absolute value of the upper limit (or lower limit for negative correlations) of the correlation with an-other factor is < 0.800, < 0.900, < 1.00, or = 1, it indicates no discriminant validity concern, or there is a marginal, moderate, or severe concern, respectively.
Rönkkö, M., & Cho, E. (2022). An updated guideline for assessing discriminant validity. Organizational Research Methods, 25(1), 6-14. doi.org/10.1177/1094428120968614
Prof. Gaskin, thank you very much for your effort and very constructive videos. I have one question and your help would be much appreciated. Could you please explain how to perform moderation analysis with moderating slope? Unlike PLS-SEM model variant, CB-SEM does not have this built-in option (or at least I could not find it). Thank you!
I've asked Christian Ringle and he says they're planning to include these extra features over the next few rollouts. I've specifically requested simple slopes plots, J-N plots, and MGA. So, for now though, you'll still need to use the PLS side to do it.
@@Gaskination Thank you very much for your answer. Is it possible to use the same approach that you demonstrated in your video on Interaction moderation in Amos (introduction of the new interaction variable)? If so, could it be used when we have one dependent, one independent, three mediators and two moderators (should moderators be analyzed separately in two models, or jointly in one)?
@@aleksadokic1872 Yes, you could create a new interaction variable outside of SmartPLS and then model it in SmartPLS. The moderators can be included in the same model.
Hi Dr. Gaskin
I am wanting to know if SmartPLS provides modification indices? I am working on my measurement model and I have been reading to look beyond standardized factor loadings and to also consider modification indices as I have read that they can be indicative of spurious correlations with other non-related measurement items and decrease model fit. Any idea where I could find this or how I could work it out? Thanks!
That is a feature SmartPLS is working on with the CB-SEM side of the software. As of the current version 4.1.0.7, it is not available.
@@Gaskination Thanks. I have been reading about modification indices and variance inflation factor (VIF). I see that VIF is possible to inspect when running a PLS model. My question is this: When running CB-SEM, do we need to inspect VIF as well, and if so, how can we do that?
@@TUPhDude as of 4.1.0.8, there is no VIF feature on the CB-SEM side. You could do it on the PLS side though.
@@Gaskination Yes, I thought so too! Thank you, I appreciate it.
Thank you.
Thanks for the video. I have a sample composed of two categories, say males and females, when I test a certain path using SmartPls cb-sem it gives insignificant results in both categories, surprisingly, when I test the same path for the whole sample it gives significant result!!! This doesn't appear when I use Amos (in Amos the path is insignificant for the whole sample too). If testing structural model is not yet reliable in SmartPls cb-sem, can I use it for CFA only and then test hypotheses using Amos? Given that the two softwares give very close results regarding CFA, but SmartPls is just more user-friendly
If the path results are non-significant when split by groups, it may be due to sample size issues. You can use different software for different parts of the analysis as long as it is reported. Some software is certainly more userfriendly or useful than others.
Thank you Foto your tips! :) Do the data need to be normally distributed for CB-SEM in SmartPLS?
Yes, good question. The data should be normally distributed for the CB-SEM analysis.
@@Gaskination Thank you! :)
Thank you Dr. Gaskin! Your videos are really helpful. I have a question if you don't mind. Could I use the CB-SEM feature in SmartPLS to conduct an EFA? I should also mention that my model may potentially have second order factors. I'm not sure how I would test the need for second order factors in an EFA. Any insight you can provide would be greatly appreciated. Thank you.
You can actually do it in the PLS model (rather than the CB-SEM). Look at the "crossloadings" matrix in the Discriminant validity section of the output when running the regular pls algorithm. As for higher order factors, I have a few videos on validating those. Here is a playlist with those videos: ruclips.net/p/PLnMJlbz3sefIJoPbCfTDwzOgMKR3IHL5x&si=7R3guB1XAMEkaVVj
@@Gaskination thank you so much!!!!
@@Gaskination Dr. Gaskin I'm sorry to bother you again but I want to be sure I understand.
I can perform an EFA and subsequently test hypotheses in smartpls?
Yes. You can also just do the EFA in SmartPLS by checking the crossloadings matrix.
@@Gaskination Thank you again. I will do the EFA using the crossloadings matrix in SmartPLS and will also test the hypotheses in SmartPLS
❤❤❤
Very nice sharing, Prof.
Btw, how to handle formative constructs in CB-SEM with SmartPLS 4?
If you have formative constructs, please use the PLS features, rather than cbsem
Dear sir, I just have a query as to whether we can run a reflective model in CB-SEM, given in Smart PLS SEM 4 software?
Regards
as shown in this video, you can run a reflective model in CB-SEM in SmartPLS.
H, can you do the the CFA with meditation variable and why we test for more than one model , I see that in some research.
Thans
The CFA should include all latent reflective factors, even mediators and moderators (if they are reflective and latent). The CFA should include all these factors at once, rather than one at a time. This allows us to test for model fit and discriminant validity.
One more question. I would like to use SmartPLS to calculate my CB-SEMs. How do I set my estimation procedure (preferably maximum likelihood)? Or is that always preset?
Yes, Maximum Likelihood is the default estimation procedure.
👏👏👏👏👏👏👏
I have a question, can we do the correlation for the exogenous and endogenous variables at once? or we need to separate them.
It is best to do them all at once so that you can ensure discriminant validity.
Very nice
Hi this result is looking great really!
I used Amos and diagram cant understand. Can I draw the amos diagram by hand? Amos created result is very unclear and I cant read properly many numbers
I'm not sure if I understand what you're asking. Here is a basic model drawing in AMOS:
ruclips.net/video/efC81f-Z22Q/видео.htmlsi=JBLU5G5i7qeLMcWd
Thanks 😊😊😊😊😊😊
does this mean we don't need AMOS anymore?
How to choose between PLS SEM and CB SEM
The primary criteria are whether you have formative factors (choose PLS) or low sample size (choose PLS) or exceptionally complex model with interactions (choose PLS). Otherwise, it is currently considered best practice to choose CB-SEM.
is it available in 4 ?
In the most recent update. Make sure to update to at least 4.0.9.3
@@Gaskination Is it available in the students versions?
@@amalwijenayaka410 I'm not sure. I would think so. But if you don't see it there, then it must not be there yet. Hopefully in the next release.
@@Gaskination Thank you very much
How could I create moderater
SmartPLS is working on moderation right now. They hope to have it completed by January. In the meantime, you can toggle over to the PLS side for interactions, or you can produce those interactions in Excel/SPSS yourself (standardize variables and then multiply), and then bring the dataset back into SmartPLS, or you can run them as grouping moderators instead (high/low if necessary) and run the model with the two different datasets (one for each group) and then compare those models. It is tedious and manual, but should work. I'm anxiously awaiting the update to SmartPLS to include these features on the CB-SEM side.
RUclips videos always have perfect values. But in reality we face horrible values and never get any reference to deal with it...
Do you have a question? Here is an example of a youtube video with horrible values: ruclips.net/video/oeoTpXiSncc/видео.htmlsi=stj0zfBWOGycU-S7