Need help in deciding between SmartPLS and AMOS??
Afzal Moshadi
While checking the assumptions of AMOS I found out that my data is not fulfilling multivariate normality. Even though it is Uni-variate normal data. I am left with an option of leaving Maximum Likelihood and adopt Asymptotically distribution free (ADF) or run Bootstrapping technique. Third option is Satorra and Bentler(1988) which I dont know how to run and get softwars (no worries). I think if I have to go for a non-parametric technique than I should do SmartPLS instead. How would we support the results of SmartPLS when it is said that it is not a theory building technique whereas covariance based technique (AMOS) is consider onw. Does this has any implication on my PhD study regarding theoretical contribution. Kind suggestions are required.
Further I have data from two different surveys i) Online survey data (n-338) ii) Field survey data (n-388). Can I combined both samples or should I treat them seperately. What are the chances of intermode biased? Do we have a statistical technique to provide guidance. Also any helping material would be appreciated...
Sanjay Kumar
Dear Afzal,
I used SmartPLS for my PhD work in 2015. I had to move to use Smart PLS because of two reasons:
1. Theoretical.
2. Each of normality in m yu Dara.
Theoretical reason has to be the main basis. While constructing your SM and MM, theory will dictate the type and nature of the relationships between your Latest Variables and the Manifest Variables. AMOS can handle Reflective indicators only, whereas my theory pointed towards Formative relationships.
So, think clearly about your theoretical basis. It will help you to decide.
Best wishes.
Sanjay Kumar
--
Join us at the PLS Applications Symposium: http://plsas.net
---
You received this message because you are subscribed to the Google Groups "PLS-SEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pls-sem+unsubscribe@googlegroups.com.
To post to this group, send email to pls...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Sanjay Kumar
Sorry. Point no 2 should read:
Lack of normality in my data.
Sorry for the typo.
Schuberth Florian
I would say it mainly depends on the way you modeled your constructs: common factor (reflective measurement) vs. composite (formative measurement). If you have only common factors I would go for a covariance-based technique since PLS leads to biased estimiates. Of course you can use PLSc to get consistent estimates. There can be situations where PLS has an advantage in terms of sample size, but to my opinion this is rarely the case.
If you include composites in your model, in particular in an endogenous position, go for PLS.
It is not true that PLS cannot be used for confirmatory purposes, since there is also an overall-model fit. I recommend you to read Henseler et al. (2016) for an current view on PLS
Best regards,
Florian
>>> Afzal Moshadi <afzal....@gmail.com> 10.11.16 6.26 Uhr >>>
While checking the assumptions of AMOS I found out that my data is not fulfilling multivariate normality. Even though it is Uni-variate normal data. I am left with an option of leaving Maximum Likelihood and adopt Asymptotically distribution free (ADF) or run Bootstrapping technique. Third option is Satorra and Bentler(1988) which I dont know how to run and get softwars (no worries). I think if I have to go for a non-parametric technique than I should do SmartPLS instead. How would we support the results of SmartPLS when it is said that it is not a theory building technique whereas covariance based technique (AMOS) is consider onw. Does this has any implication on my PhD study regarding theoretical contribution. Kind suggestions are required.
Further I have data from two different surveys i) Online survey data (n-338) ii) Field survey data (n-388). Can I combined both samples or should I treat them seperately. What are the chances of intermode biased? Do we have a statistical technique to provide guidance. Also any helping material would be appreciated...
Join us at the PLS Applications Symposium: http://plsas.net
---
You received this message because you are subscribed to the Google Groups "PLS-SEM" group.
Arunkumar Balasubramaniam
Arun
From: Afzal Moshadi <afzal....@gmail.com>
To: PLS-SEM <pls...@googlegroups.com>
Sent: Thursday, November 10, 2016 10:56 AM
Subject: [pls-sem] Need help in deciding between SmartPLS and AMOS??
While checking the assumptions of AMOS I found out that my data is not fulfilling multivariate normality. Even though it is Uni-variate normal data. I am left with an option of leaving Maximum Likelihood and adopt Asymptotically distribution free (ADF) or run Bootstrapping technique. Third option is Satorra and Bentler(1988) which I dont know how to run and get softwars (no worries). I think if I have to go for a non-parametric technique than I should do SmartPLS instead. How would we support the results of SmartPLS when it is said that it is not a theory building technique whereas covariance based technique (AMOS) is consider onw. Does this has any implication on my PhD study regarding theoretical contribution. Kind suggestions are required.
Further I have data from two different surveys i) Online survey data (n-338) ii) Field survey data (n-388). Can I combined both samples or should I treat them seperately. What are the chances of intermode biased? Do we have a statistical technique to provide guidance. Also any helping material would be appreciated...
Ned Kock
The most difficult challenge to PLS-SEM made by methodologists is that it does not deal with factors, but with composites, thus not properly accounting for measurement error. Several doctoral students have raised this issue in the past here. The video clip linked below, titled “Conduct a Factor-Based PLS-SEM Analysis with WarpPLS”, discusses the differences in results you are likely to see in factor-based analyses.
As noted in the video, this challenge has been addressed by WarpPLS (see link below) through the inclusion of factor-based algorithms, in addition to the classic PLS Mode A and PLS Mode B algorithms. Authors of papers and dissertations have included results employing factor-based algorithms in their reports, often in addition to classic PLS-SEM results, to overcome this challenge and other methodological challenges.
In addition to that, WarpPLS provides what is probably the most extensively set of results of any software, including: univariate and multivariate measures of normality, several model fit and quality indices, Q-squared predictive validity coefficients, indirect effects for paths with multiple segments, causality assessment coefficients, various 2D graphs and 3D graphs etc.
Best regards, Ned
From: pls...@googlegroups.com [mailto:pls...@googlegroups.com] On Behalf Of Afzal Moshadi
Sent: Wednesday, November 09, 2016 11:26 PM
To: PLS-SEM
Subject: [pls-sem] Need help in deciding between SmartPLS and AMOS??
Dear Group Members,
While checking the assumptions of AMOS I found out that my data is not fulfilling multivariate normality. Even though it is Uni-variate normal data. I am left with an option of leaving Maximum Likelihood and adopt Asymptotically distribution free (ADF) or run Bootstrapping technique. Third option is Satorra and Bentler(1988) which I dont know how to run and get softwars (no worries). I think if I have to go for a non-parametric […]
|
This email has been checked for viruses by Avast antivirus software.
|
Afzal Moshadi
I dont see an issue on this aspect as my theory suggests reflective rather than formative relationship. Hence that is quite do able in both. Major concern is the Multivariate normality. By the way I am very thankful for your kind suggestion
Regards
SAM Shah
To unsubscribe from this group and stop receiving emails from it, send an email to pls-sem+u...@googlegroups.com.
Afzal Moshadi
Many thanks for your valuable comments. I will try to get hold of this article, however I might not be able to secure it because I am from an underdeveloped country and our universities don't have an access to latest articles i.e. 2016. I appreciate your help
Best Regards
SAM Shah
Afzal Moshadi
Thanks for relevant suggestions. Yes I tried transformations. However no luck. I checked standardized z, checked the log10 and even Mahalanobis distance and ommitted around 138 outliers to reach 6.5 c.r value. I checked the missing values and found none.
Regarding group differences. I checked Chi square tests earlier on and did not find an evidence that they are significantly different. I might try Mann-Whitney U test on your demographic and latent variables and will see what is the outcome.
Thanks for your valuable suggestons
Regards
SAM Shah
Jörg Henseler
Dear SAM Shah,
The article mentioned by Florian is open access and thus accessible for everyone.
Best regards,
Jörg
--
Prof. Dr. Jörg Henseler
Chair of Product-Market Relations
Department of Design, Production and Management
Faculty of Engineering Technology
University of Twente.
P.O. Box 217, 7500 AE Enschede, The Netherlands
Phone: +31 (0)53 489 2953
E-mail: j.hen...@utwente.nl
Afzal Moshadi
Thanks for your help. Yes I have downloaded that paper and am currently getting benefited.
Regards
SAM
Jose L. Roldan
Sarstedt, M., Hair, J. F., Ringle, C. M., Thiele, K. O., & Gudergan, S. P. (2016). Estimation issues with PLS and CBSEM: Where the bias lies! Journal of Business Research, 69(10), 3998–4010. http://doi.org/10.1016/j.jbusres.2016.06.007
--
>>> Afzal Moshadi <afzal....@gmail.com> 10.11.16 6.26 Uhr >>>Dear Group Members,--
While checking the assumptions of AMOS I found out that my data is not fulfilling multivariate normality. Even though it is Uni-variate normal data. I am left with an option of leaving Maximum Likelihood and adopt Asymptotically distribution free (ADF) or run Bootstrapping technique. Third option is Satorra and Bentler(1988) which I dont know how to run and get softwars (no worries). I think if I have to go for a non-parametric technique than I should do SmartPLS instead. How would we support the results of SmartPLS when it is said that it is not a theory building technique whereas covariance based technique (AMOS) is consider onw. Does this has any implication on my PhD study regarding theoretical contribution. Kind suggestions are required.
Further I have data from two different surveys i) Online survey data (n-338) ii) Field survey data (n-388). Can I combined both samples or should I treat them seperately. What are the chances of intermode biased? Do we have a statistical technique to provide guidance. Also any helping material would be appreciated...
Join us at the PLS Applications Symposium: http://plsas.net
---
You received this message because you are subscribed to the Google Groups "PLS-SEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pls-sem+u...@googlegroups.com.
To post to this group, send email to pls...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Join us at the PLS Applications Symposium: http://plsas.net
---
You received this message because you are subscribed to the Google Groups "PLS-SEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pls-sem+u...@googlegroups.com.
To post to this group, send email to pls...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
__________________________________________________________ Dr. José L. Roldán Associate Professor of Business Administration Senior Editor, The DATA BASE for Advances in Information Systems http://sigmis.org/the-data-base/ Department of Business Administration and Marketing University of Seville Ramon y Cajal, 1. 41018 Seville (SPAIN) Voice: (34) 954 554 458 / 575 Fax: (34) 954 556 989 Skype: jlroldan67 <mailto:jlro...@us.es> URL: http://personal.us.es/jlroldan
http://orcid.org/0000-0003-4053-7526
Google Scholar: https://goo.gl/PPY32K __________________________________________________________