Greetings sir Can we take demographic variables such as age,gender,education as control variables in pls sem? Also is there a need to perform mga as these are control variables?
@@anugrewal9790 Inclusion of CVs is already explained in this lecture. Please note that structural techniques don't provide detailed CV analysis as in HLR. MGA should only be performed if moderation is hypothesized.
@@ChMahmoodAnwar thank you sir In case a control variables comes significant Do I need to perform mga ? Because it’s not my hypothesis part to check the impact of demographic variables? I will be very thankful
Thanks for your video. I’d like to ask 1) if I have more control variables such as gender, age, type of job, how do we include in the model? 2) If I have more than 2 types of job, how do I record it as control variable? Should it be like 1=doctor, 2=accountant, 3=teacher ?
Thank you for the video. Can I use consistent PLS algorithms if my IVs are reflective but I include gender, age, income etc as control variables? or do I need to consider the model as including formative elements when I include control variables?
Thanks for the video. I was wondering did you extract control variable from spss? If yes how did you do it? you have latent variable here. I need to do the same. I appreciate your help. Thanks
Nominal or categorical controls are not considered as latent variables. Watch my lecture "How to prepare and upload Data file for PLS Path Modeling in SmartPLS" to get an idea how to prepare data file for SmartPLS.
Hi Sir, thank you for your email. I would like to ask something. What is the difference between testing control variables and ANOVA? If I am not familiar with testing control variables in your way, may I use Smart PLS for assessing measurement and structural model and use ANOVA to test the impact of control variables, such as income level ? Thanks.
Control variables are analyzed using Stepwise Regression Analysis, whereas ANOVA is used to assess mean differences between 3 or more sample means. For control variables analysis in SPSS, please search and watch my lecture on it. Thanks
Hi Ch. Mahmood Anwar your video helps me a lot but may I ask how do you code gender. for example there are male and femal and how do you make them in one variable? thx
if we are using SMart-PLS, do we need to find out the edogeneity? (2) can we calculate ENDOGENEITY USING SMART-PLS? WHICH SOFTWARE IS THE BEST FRO CALCULATION OF ENDOGENEITY? HOPE YOU WILL ANSWER IN DETAIL
Endogeneity is a critical CLRM assumption, regardless of analysis technique you are using i.e., OLS, ML, PLS etc. It should be tested, especially, in causal models. It can easily be detected with Durbin-Wu-Hausman test (augmented regression test) in STATA. Once you get an evidence of Endogeneity, you need to use 2SLS to get unbiased and consistent results asymptotically. Hope this will help!
Salam, I am just a learner, but you are an expert. As per my little knowledge, gender as control variable has significant effect on loyalty because both BCLL and BCUL values are negative. What you say?
W.Salam. For control variables, don't look at CIs, simply observe t-statistic and its significance (in this example it is less than 1.96, insignificant at 0.059-Red colour). Hint: Run the PLS Algo for model with control variables than run Bootstrapping for same model. You will find that outcome is similar for control variables.
Please explain how to handle 2 or more than 2 Dependent variable while writing equation in Eviews and interpret the results so obtained or share any link which explains this issue. Thanks
You can write a loop in Eviews like this and interpret results in standard way: for %dep dep1 dep2 dep3 equation eq{%dep}.ls {%dep} c x1 x2 x3 next Hope it'll work for you.
@@ChMahmoodAnwar the mentioned above did not work. assume I have 3 independent variable and multiple independent variables, how to write equation. please suggest
@@asarwaribit It really works. May be you are not entering the code properly. You should read the programming chapter of the EViews Command and Programming Reference. Thanks.
@@reynreyn9900 Controls and moderators lie at same level. Controls need to be controlled in the model, whereas moderators regulate IV-DV signal strength and/or relationship direction.
Salam, so SmartPLS could just basically tell you whether significant or not, but can it tell you whether a male or a female has a higher mean value? Or this sort of test should be done in ANOVA using SPSS. BTW If the gender or other control variables are significant in smartpls , is it possible to be insignificant in SPSS since they were actually using different methods? English ain't my native tongue, hope you could actually understand what I'm talking about
W. Salam. Actually Path Models and SEM were not developed to analyze controls. Regression is the classic model that can handle controls and other confounding variables. As far as SmartPLS is concerned, you can run a Multi Group Analysis (MGA), but will only tell you whether a group significantly influence the outcome or not?
Gender is not a dummy variable although it looks like. You should code gender as 1-M, 2--F. For more details watch "How to Code, Report and Interpret Control Variables in SPSS: Example of Gender and Age" on ruclips.net/video/EQq1hTbu-6I/видео.html
@@ChMahmoodAnwar First of all, thanks for replying. Sometimes a control variable (gender, age, education) is added as dummy, but I do not understand why it is done. Which one is correct? I am confused about it. By the way, could you send an example of control variable reported on the paper, especially with amos or smart? I do not know how to report. Thanks for your time.
@@corinacrn7684 To understand this you need to know why controls are added to the model? Basically, Hierarchical Regression is best to analyze controls, AMOS and SmartPLS were not developed to analyze controls. You can consult Jabbour et al. (2014) to have an idea.
Very good explanation of Smart-PLS, reasonable speed.
👍
Greetings sir
Can we take demographic variables such as age,gender,education as control variables in pls sem?
Also is there a need to perform mga as these are control variables?
@@anugrewal9790 Inclusion of CVs is already explained in this lecture. Please note that structural techniques don't provide detailed CV analysis as in HLR. MGA should only be performed if moderation is hypothesized.
@@ChMahmoodAnwar thank you sir
In case a control variables comes significant
Do I need to perform mga ? Because it’s not my hypothesis part to check the impact of demographic variables?
I will be very thankful
@@anugrewal9790 You can simply report these significant results. MGA can also be run as an auxiliary test.
@@ChMahmoodAnwar ok thank you sir
Thanks for your video. I’d like to ask 1) if I have more control variables such as gender, age, type of job, how do we include in the model? 2) If I have more than 2 types of job, how do I record it as control variable? Should it be like 1=doctor, 2=accountant, 3=teacher ?
Hi. You may include other control variables in similar way as briefed in this lecture. Code controls as you mentioned. Best.
Thank you for the video. Can I use consistent PLS algorithms if my IVs are reflective but I include gender, age, income etc as control variables? or do I need to consider the model as including formative elements when I include control variables?
You should apply PLS Algo. Thanks.
Thanks for the video. I was wondering did you extract control variable from spss? If yes how did you do it? you have latent variable here. I need to do the same. I appreciate your help.
Thanks
Nominal or categorical controls are not considered as latent variables. Watch my lecture "How to prepare and upload Data file for PLS Path Modeling in SmartPLS" to get an idea how to prepare data file for SmartPLS.
@@ChMahmoodAnwar thanks i will check it
Hi Sir, thank you for your email. I would like to ask something. What is the difference between testing control variables and ANOVA? If I am not familiar with testing control variables in your way, may I use Smart PLS for assessing measurement and structural model and use ANOVA to test the impact of control variables, such as income level ? Thanks.
Control variables are analyzed using Stepwise Regression Analysis, whereas ANOVA is used to assess mean differences between 3 or more sample means. For control variables analysis in SPSS, please search and watch my lecture on it. Thanks
Hi Ch. Mahmood Anwar
your video helps me a lot but may I ask how do you code gender. for example there are male and femal and how do you make them in one variable? thx
Gender is always nominally/dummy coded.
if we are using SMart-PLS, do we need to find out the edogeneity? (2) can we calculate ENDOGENEITY USING SMART-PLS? WHICH SOFTWARE IS THE BEST FRO CALCULATION OF ENDOGENEITY? HOPE YOU WILL ANSWER IN DETAIL
Endogeneity is a critical CLRM assumption, regardless of analysis technique you are using i.e., OLS, ML, PLS etc. It should be tested, especially, in causal models. It can easily be detected with Durbin-Wu-Hausman test (augmented regression test) in STATA. Once you get an evidence of Endogeneity, you need to use 2SLS to get unbiased and consistent results asymptotically. Hope this will help!
@@ChMahmoodAnwar Thanks for response
Sub sample 500 is the sample size? I have 251 as my sample size, so should I need to change the sub sample size into 251? Please explain
You may use any sample size for PM.
Salam, I am just a learner, but you are an expert. As per my little knowledge, gender as control variable has significant effect on loyalty because both BCLL and BCUL values are negative. What you say?
W.Salam. For control variables, don't look at CIs, simply observe t-statistic and its significance (in this example it is less than 1.96, insignificant at 0.059-Red colour). Hint: Run the PLS Algo for model with control variables than run Bootstrapping for same model. You will find that outcome is similar for control variables.
sir plz guide threshold of the control variable, and their interpretation.
Could you please elaborate?
Please explain how to handle 2 or more than 2 Dependent variable while writing equation in Eviews and interpret the results so obtained or share any link which explains this issue. Thanks
You can write a loop in Eviews like this and interpret results in standard way:
for %dep dep1 dep2 dep3
equation eq{%dep}.ls {%dep} c x1 x2 x3
next
Hope it'll work for you.
@@ChMahmoodAnwar the mentioned above did not work. assume I have 3 independent variable and multiple independent variables, how to write equation. please suggest
@@asarwaribit It really works. May be you are not entering the code properly. You should read the programming chapter of the EViews Command and Programming Reference. Thanks.
What if I have more than one control variables? should I separate them?
Simply connect each CV to DV.
what if the control variable is a latent construct (like trust in your paper)?
Theoretically, SEM and PM were not designed to analyze control variables. Use regression if you want to analyze or work with controls.
@@ChMahmoodAnwar thank you so much for the answer Professor. I have one more question, are control variable and moderator variable the same?
@@reynreyn9900 Controls and moderators lie at same level. Controls need to be controlled in the model, whereas moderators regulate IV-DV signal strength and/or relationship direction.
@@ChMahmoodAnwar thank you so much Professor! I understand better now
Wonderful. Sir how can I Contact you.
Salam, so SmartPLS could just basically tell you whether significant or not, but can it tell you whether a male or a female has a higher mean value? Or this sort of test should be done in ANOVA using SPSS.
BTW If the gender or other control variables are significant in smartpls , is it possible to be insignificant in SPSS since they were actually using different methods?
English ain't my native tongue, hope you could actually understand what I'm talking about
W. Salam. Actually Path Models and SEM were not developed to analyze controls. Regression is the classic model that can handle controls and other confounding variables. As far as SmartPLS is concerned, you can run a Multi Group Analysis (MGA), but will only tell you whether a group significantly influence the outcome or not?
Hey, I am doing analysis similar to yours. Can you help how you did that?
@@Austresi Hi. You can find my email on About section and send me details.
Do we need to convert gender into dummy variables?
Gender is not a dummy variable although it looks like. You should code gender as 1-M, 2--F. For more details watch "How to Code, Report and Interpret Control Variables in SPSS: Example of Gender and Age" on ruclips.net/video/EQq1hTbu-6I/видео.html
To be frank, I do not understand that well. I need help
What you didn't understand? Please mention.
@@ChMahmoodAnwar First of all, thanks for replying. Sometimes a control variable (gender, age, education) is added as dummy, but I do not understand why it is done. Which one is correct? I am confused about it. By the way, could you send an example of control variable reported on the paper, especially with amos or smart? I do not know how to report. Thanks for your time.
@@corinacrn7684 To understand this you need to know why controls are added to the model? Basically, Hierarchical Regression is best to analyze controls, AMOS and SmartPLS were not developed to analyze controls. You can consult Jabbour et al. (2014) to have an idea.