Buenos días estimado Doctor James Gaskin. Tengo una consulta, espero pueda ayudarme: si tengo una variable dependiente medido reflexivamente con 5 items ordinales que va de 1 al 7 y una variable independiente medido reflexivamente con 5 items dicotomicos. Mi pregunta son: a) se puede utilizar PLS-SEM para determinar la influencia de la variable independiente en la variable dependiente?, b) en caso de ser así, los indicadores del modelo de medida para la variable con ítems dicotómicos serían los mismo que para la variable con ítems ordinales?. Gracias
Es muy poco común que un factor reflectivo esté compuesto por ítems dicotómicos. Me sorprendería si este factor pasara las pruebas de AVE y CR. Pero si por algún milagro lo hiciera, entonces sí, puedes estimar esta relación en SmartPLS.
Hello James. Thank you for sharing these videos which i am using to improve my skills at research. It came a doubt to me: which one is better for Multi-group analysis, AMOS or SMART PLS4? Kind regards.
Hi James. I try to develop a formative-formative scale for a variable, and then use this variable with this scale (as an independent variable) to predict other variables. However, the significance of indicators and dimensions in this scale changes greatly when a dependent variable changes to be a different variable. Some indicators and dimensions of the scale (independent variable) change from being significant to being insignificant. My question is to determine a formative-formative scale, what results should I base on to determine which indicators and dimensions I should retain? Thank you for kindly sharing your lecture videos with the public. I have learned a lot from your videos.
Here you go: ruclips.net/video/sGqp6hRktI8/видео.htmlsi=mLt-vhULQCCXj4xV (the method is the same for IVs) And here it is for the CB-SEM side: ruclips.net/video/Y2qsapJki6s/видео.htmlsi=HqTmYCi_gXIuHoTz
Dear Prof., Is it correct to take a control variable as a moderator? Because in my research model, I have taken a construct training as a moderator. And in the past research I found that one study has taken it as a control variable? Could you suggest any resource reference to justify it?
Whether it is a control or a moderator depends on your theory. If your theory includes it as a moderator, that is your choice. No need to further justify it.
Please what is the difference between path and factor analysis? Because when I do path analysis, I encounter the problem of the single matrix, while when I do factor analysis, I do not find the problem. What is the solution?
Path analysis is to examine the structural paths between constructs. Factor analysis is to examine the measurement quality of constructs. Path analysis is for structural modeling. Factor analysis is for measurement modeling. A singularity matrix occurs when there is insufficient variance on a variable included in a model. Please check that all of your variables vary (e.g., don't include a job title variable if all workers were accountants).
@@faroukpsy105 Look at it in Excel or SPSS or even in SmartPLS (data view). See if there is kurtosis > 10. If so, there is probably an insufficient variance problem.
@@Gaskination Thank you. Most of the kurtosis values are 1, there are two values of 3, and one value of 5. For your information, the sample is 235 and it has been confirmed that there is no complete relationship between the variables. Is there a solution please?
@@Gaskination Thank you. Most of the kurtosis values are 1, there are two values of 3, and one value of 5. For your information, the sample is 235 and it has been confirmed that there is no complete relationship between the variables. Is there a solution please?
how we choose different type of data set? as you set service provider is categorical? does it make any difference? can you explain it and tag me to the video?
A variable is categorical if the numerical values are not meaningful, but simply identify a category. For example, gender, job category, industry sector, religion, sports team, etc. This does make a difference. In SmartPLS, you can use these variables in multigroup analysis (MGA). In regression, it makes no sense to use these as predictors or outcomes without first converting them to a set of dummy variables. This is because regression predicts numeric values, but the numeric values for categorical variables are meaningless other than for identification of the group.
Linearity is testing whether the independent variable and dependent variable are linearly related, rather than curvilinearly. One way to test this is by running curve analysis in SPSS. You can do this under analysis, regression, curve estimation.
If it is binary or text or a formula in Excel, it might disallow changing. If it is a formula in Excel, open it up again in Excel and copy and paste only the values (not the formula).
I'm not sure why that might happen. Make sure to expand the project so that you can see all parts of it. Also, make sure you've specified a workspace that is a folder you have administrative rights to edit.
Sir hope so you are good and well. Sir, I have downloaded your plugin folder and downloaded 28+ for IBM Amos 29 version. I am facing the issue that every plugin has now a question mark. I have also gone through your fix syntax error file but it is still not resolved. can you make a video? I am very grateful to you for this favor.
I have not been able to try it for version 29 yet. I don't have access to version 29. My best guess would be that it needs to be unblocked before it is moved into the plugins folder. If that doesn't work, I'm very sorry.
Hello James, is it possible to create a PLS-SEM model where the dependent latent variable is categorical (e.g. choice of a brand) and the rest of the variables are either categorical (mostly dummies) or Likert scale?
If the dependent variable is binary, then yes. If it is multinomial, then you would need to break it up into dummy variables. I think a different analytical approach might be simpler. Perhaps a chi-square test: www.mash.dept.shef.ac.uk/Resources/MASH-WhatStatisticalTestHandout.pdf
If it is a continuous number, with each increment representing an additional participant, then it is metric. If it is binned/bucketed into several equally sized groups, then it is ordinal. If it is split into a few large groups (e.g.,
I've been making updates for version 28. Those updates can be found in the Estimands and Plugins folder on Google Drive (linked from the StatWiki). Check the Version 28 folder.
Thank you! Your videos are very informative for students also working on their thesis.
this is most simple and easy and effective to guide my first time SmartPLS operating. thk u so much
Thank you so much. Your teaching is simple and understandable. 💯
Buenos días estimado Doctor James Gaskin. Tengo una consulta, espero pueda ayudarme: si tengo una variable dependiente medido reflexivamente con 5 items ordinales que va de 1 al 7 y una variable independiente medido reflexivamente con 5 items dicotomicos. Mi pregunta son: a) se puede utilizar PLS-SEM para determinar la influencia de la variable independiente en la variable dependiente?, b) en caso de ser así, los indicadores del modelo de medida para la variable con ítems dicotómicos serían los mismo que para la variable con ítems ordinales?. Gracias
Es muy poco común que un factor reflectivo esté compuesto por ítems dicotómicos. Me sorprendería si este factor pasara las pruebas de AVE y CR. Pero si por algún milagro lo hiciera, entonces sí, puedes estimar esta relación en SmartPLS.
GREAT VIDEO!
I imported data it was in blue clr but there is not licenser word could be written what can i do 👀
I don't know what you mean by "licenser word". So I'm not sure what problem you are facing. Can you explain?
My smart pls doesn’t show process icon
And also it refuses to import excel files only CSV which I can’t do the regressions😢
check you have the latest version SmartPLS 4 (not version 3)
Thank you so much
Hello James. Thank you for sharing these videos which i am using to improve my skills at research. It came a doubt to me: which one is better for Multi-group analysis, AMOS or SMART PLS4?
Kind regards.
SmartPLS makes it easier I think. But both work. Here is a video for how to do it in SmartPLS: ruclips.net/video/L2LdAgMKmBo/видео.html
Hi James. I try to develop a formative-formative scale for a variable, and then use this variable with this scale (as an independent variable) to predict other variables. However, the significance of indicators and dimensions in this scale changes greatly when a dependent variable changes to be a different variable. Some indicators and dimensions of the scale (independent variable) change from being significant to being insignificant. My question is to determine a formative-formative scale, what results should I base on to determine which indicators and dimensions I should retain? Thank you for kindly sharing your lecture videos with the public. I have learned a lot from your videos.
If it is truly formative-formative, then all indicators should be kept unless they are exhibiting high multicollinearity or nearly zero effect.
can anyone plz guide ... how to add Independent variable with sub diementions /sub section
Here you go: ruclips.net/video/sGqp6hRktI8/видео.htmlsi=mLt-vhULQCCXj4xV (the method is the same for IVs)
And here it is for the CB-SEM side: ruclips.net/video/Y2qsapJki6s/видео.htmlsi=HqTmYCi_gXIuHoTz
Dear Prof., Is it correct to take a control variable as a moderator? Because in my research model, I have taken a construct training as a moderator. And in the past research I found that one study has taken it as a control variable? Could you suggest any resource reference to justify it?
Whether it is a control or a moderator depends on your theory. If your theory includes it as a moderator, that is your choice. No need to further justify it.
@@Gaskination Sure Prof. Thanks for clearing my doubt.
Please what is the difference between path and factor analysis? Because when I do path analysis, I encounter the problem of the single matrix, while when I do factor analysis, I do not find the problem. What is the solution?
Path analysis is to examine the structural paths between constructs. Factor analysis is to examine the measurement quality of constructs. Path analysis is for structural modeling. Factor analysis is for measurement modeling. A singularity matrix occurs when there is insufficient variance on a variable included in a model. Please check that all of your variables vary (e.g., don't include a job title variable if all workers were accountants).
@@Gaskination thanks .so how can we know that there is not enough variance in a variable?
@@faroukpsy105 Look at it in Excel or SPSS or even in SmartPLS (data view). See if there is kurtosis > 10. If so, there is probably an insufficient variance problem.
@@Gaskination Thank you. Most of the kurtosis values are 1, there are two values of 3, and one value of 5. For your information, the sample is 235 and it has been confirmed that there is no complete relationship between the variables. Is there a solution please?
@@Gaskination Thank you. Most of the kurtosis values are 1, there are two values of 3, and one value of 5. For your information, the sample is 235 and it has been confirmed that there is no complete relationship between the variables. Is there a solution please?
how we choose different type of data set? as you set service provider is categorical? does it make any difference? can you explain it and tag me to the video?
A variable is categorical if the numerical values are not meaningful, but simply identify a category. For example, gender, job category, industry sector, religion, sports team, etc. This does make a difference. In SmartPLS, you can use these variables in multigroup analysis (MGA). In regression, it makes no sense to use these as predictors or outcomes without first converting them to a set of dummy variables. This is because regression predicts numeric values, but the numeric values for categorical variables are meaningless other than for identification of the group.
Thanks alot
Hiii james before proceeding to cfa there is an assumption named linearity can we show it through qq plots
Thanks in advance
Linearity is testing whether the independent variable and dependent variable are linearly related, rather than curvilinearly. One way to test this is by running curve analysis in SPSS. You can do this under analysis, regression, curve estimation.
Please must I always upload from csv file, how do I upload from SPSS to SMART PLS directly?
In SmartPLS 4, you can import .sav files (spss). It should be one of the default filetypes included in the Open dialog window.
I can't change the scale, anyone know why???
If it is binary or text or a formula in Excel, it might disallow changing. If it is a formula in Excel, open it up again in Excel and copy and paste only the values (not the formula).
Respected sir, what maximum number of variables and items can smart pls 4 take in student free version
I'm not sure. The website doesn't say. I think you'll need to reach out to SmartPLS support email to find out.
please ihave problem, after i import data i don't find it in the program,
I'm not sure why that might happen. Make sure to expand the project so that you can see all parts of it. Also, make sure you've specified a workspace that is a folder you have administrative rights to edit.
@@Gaskination I found the problem. The problem was with the anti-virus. When I disactivated the anti-virus .The data was imported. Thank you.
Sir hope so you are good and well. Sir, I have downloaded your plugin folder and downloaded 28+ for IBM Amos 29 version. I am facing the issue that every plugin has now a question mark. I have also gone through your fix syntax error file but it is still not resolved. can you make a video? I am very grateful to you for this favor.
I have not been able to try it for version 29 yet. I don't have access to version 29. My best guess would be that it needs to be unblocked before it is moved into the plugins folder. If that doesn't work, I'm very sorry.
@@Gaskination Thank you Sir for your kind and quick response. I really appriciate your efforts Sir.
define endogenous variables plz
Variable that has arrows pointing into it. Sometimes called a dependent variable.
Hello James, is it possible to create a PLS-SEM model where the dependent latent variable is categorical (e.g. choice of a brand) and the rest of the variables are either categorical (mostly dummies) or Likert scale?
If the dependent variable is binary, then yes. If it is multinomial, then you would need to break it up into dummy variables. I think a different analytical approach might be simpler. Perhaps a chi-square test: www.mash.dept.shef.ac.uk/Resources/MASH-WhatStatisticalTestHandout.pdf
Is participants metric or categorical?
If it is a continuous number, with each increment representing an additional participant, then it is metric. If it is binned/bucketed into several equally sized groups, then it is ordinal. If it is split into a few large groups (e.g.,
@@Gaskination thank you
Hi James. Any news on your super valuable add-ons to AMOS?
I've been making updates for version 28. Those updates can be found in the Estimands and Plugins folder on Google Drive (linked from the StatWiki). Check the Version 28 folder.
Hello professor, I am using AMOS 26. How do we get Standard coefficient in the table?
@@akhilarao87 In AMOS, go to the analysis properties, then the output tab, then check the box for standardized estimates.