How to Add Control Variables in SmartPLS3? (See Description)
Вставка
- Опубліковано 15 жов 2024
- The focus of the session is on how to add, analyze, and interpret control variables in SmartPLS3.
SmartPLS4 is now released. if you want to learn the new and updated PLS-SEM tool, please refer to the step by step playlist to learn SmartPLS 4.
• SmartPLS4
Dummy Variables in SPSS: • Categorical Predictor/...
Categorical Predictor Variables in SmartPLS: • Categorical Predictor ...
#SmartPLS #PLSSEM #SEM #ControlVariables
Greetings from Spain, sir! I am doing some research and I found your video very helpful.
I have a question, I'll be extremely grateful if you could assist me. As suggested, I created four dummy variables for my 4-category control variable (i.e. income intervals), and added 3 of them to the model, being the unadded (the lowest income interval) the reference category. The bootstrapping results show that two of those dummies were significant to the same dependent variable, the third one was not significant.
The question is: when writing my paper, how do I interpret such results correctly? How should I report that only the two highest income intervals affect the dependent variable? Should I report all path coefficients?
Thank you for your time!
Thanks for your interest.
You will mention that a particular income category has higher/lower ajd significant/insignificant impact on DV in comparison to the reference category.
Dear Sir, Can we add 2 Job Rank dummy variables (middle & seniour) as LOCs and Job Rank acts as a HOC in the SmartPLS model?
Hello, thank you for the video! One question, if our control variable is not categorical data, such as Familiarity with technology, measured by Likert scale, ex. 1-7, how should we treat this variable? I really appreciate if you have any reference that I can read or watch for further explanation. Thank you!
You can treat it as normal variable.
Good day Dr. Fawad. May I ask if I can use process to analyse the effects of control variables. If yes, would the steps be similar as demonstrated? And would I need to create dummy variables for categorical data in Process as well?
Thank you so much Dr for your kind help!
Thanks. I am glad you liked it. You can use covariates, the continuous variables as controls. Or you can run two models one with categorical predictors and other without and see the change in effects.
Thank you for your kind help Dr. Fawad
Thats great much needed, Bundle of thanks Sir I have a few question please.
Can we also have to report the measurement model of control variable as these variable consists of only one dimension so is it possible to get measurement results (reliability and validity of these variable? Isn't including control variable in the model affect the overall outcome of the measurement model?
Do we also have to report the control variable's effect on the endogenous variable or not?
How can we determine the how the effect of control variable effect on overall model
Pleasure. I am glad you liked it. NO, for control variables quality criteria is not assessed.
Dear Prof, if we have several control variables, can we assess their effect seperately or all together?
Not separately.
hi sir thank you for this video. i have a question, if my control variables has 4 categories what value do i use to interpret them? for example for control variabel Country i have 4 categories (USA, UK, Rusia, Australia) is it ( 0 , 1 , 2 , 3) or (1.2.3.4) ?
thank you
Thanks. I am glad you liked it.
Please create dummy variables, and add three of the four countries whereas the one left is your reference category. Each variable will be added separately in the model.
Greetings of the day sir
For control variable categorical like gender, age, qualification do I need to perform micom or mga or can I directly do like this above?
MICOM is for MGA. Control is different.
@@researchwithfawad ok sir thank for your answer, it means control variable analysis will be performed for categorical variables by dummy variables?
HI thanks for your helps, i have some question I can choose more than one option for each how can I model them? for example in what region you serve ? reg 1 to 20. should i add 19 variables to model ?
Pleasure. I am glad you liked it. Yes, you will add 19 dummy variables
Greetings sir
I have 8 demographic variables in my study do I need to add them all as control variable ?
Also some of them have more than 3 categories.
But in literature I came across moderator and I added that in my model so is it fine to have model without control variable
You can have a model without controls.
@@researchwithfawad ok sir thank you
Hello Doctor Fawad
Please i have a 5 category control variable so i created 5 dummy variables and tested 4 and consider the fourth as reference to test the control variable. If i have one insignificant effect of one of the dummy variables tested , does this mean that the effect of the whole control variable is insignificant?
Thanks.
No, if just one is insignificant, you can say it is significant overall.
Can we do this through MGA also?
Why not.
Greetings..
i ran the bootstrapping after coding my variables s binary. but the results show NaH or n/a. what could be the issue. Kindly guide. Thank you.
Hi. Did you add all the dummy variables or left one out as reference.
Hello Sir, do I assess my structural model without control variables first and report it. Then add the control variables and report on that? (I'm using SmartPLS4 and I've been following your series but I'm a bit confused about the reporting procedures)
Yes you can do it this way to see how controls affect the relationship.
@@researchwithfawad I saw that I can use a PROCESS model to interpret control variables, can you provide instructions on that or is it better to follow this regular PLS-SEM procedure?
Hello Sir, could you please tell me if it is possible to split the gender variable an see if women have different intrests than men ?
for example for men enjoyment is important for purchaseintentions but not for woman?
Thanks in advance :)
Please watch
ua-cam.com/video/taw1kMfi0wk/v-deo.html
excellent
Thanks. I am glad you liked it
Asalam alaikum sir
The aim of my study is to see the impact of social media marketing on purchase intention. Sir can i have age as an control factor for both independent as well as dependent variable .ie for both social media marketing and purchase intention??
May allah bless you
WAS. Use Control for the endogenous variables.
@@researchwithfawad thank you so much sir
How about the reference category itself? Why did not we check the effect of it?
That is what the other categories are compared to. It is like a test of differences.
what to do if categorical variables such as gender are significant as control variables. How to interpret them further. In my case I have used gender as control variable and it is significant.
This would mean that there is a confounding effect. A confounder is a variable whose presence affects the variables being studied so that the results do not reflect the actual relationship. This would mean that the effect of other variables may be obscured and that the control variable is contributing to the variable in DV. For example if the impact of gender is significant, this mean that the outcome may be different from male and female.
@@researchwithfawad What would be the next step. Will I show two models in my paper one with control and one without control? How to report this in the paper. Can you please share an article where there is a categorical control variable?
And I would like to thank you for your informative videos. Learnt a lot from you sir.
What if we are supposed to report even the impact of junior level.
There has to be a reference category, we are not reporting impact of categories, rather we are comparing the construct across different categories.
@@researchwithfawad Okay sir. Thank you so much. May allah bless you
AOA. I hope you are doing well. I have a few queries regarding my control variables.
1. If a few control variables are categorical variables like a country (three countries) and job ranks (three levels). When we check the mean, SD, and correlation of the variables, we will add the country and job ranks as a whole or check it separately for every country or only two countries from the three countries and serve one country as a reference and the same goes for job ranks?
2. If control variables have no confounding effect on DV, shall we check its HTMT or not, I saw a few papers where control variables have an insignificant relationship with DV but they have reported control variables in HTMT or Fornel & Larcker as well.
3. If we are using SmartPLS 4, we will check the control variables' effect in PLS-SEM as we did earlier in SmartPLS 3 or in its process tool? Need your kind guideline regarding my three queries.
Regards
Thanks for your interest. Here are my recommendations.
1. If a few control variables are categorical variables like a country (three countries) and job ranks (three levels). When we check the mean, SD, and correlation of the variables, we will add the country and job ranks as a whole or check it separately for every country or only two countries from the three countries and serve one country as a reference and the same goes for job ranks?
We do not check the Mean and SD for control variables. Frequency Analysis is utilized for such variables.
2. If control variables have no confounding effect on DV, shall we check its HTMT or not, I saw a few papers where control variables have an insignificant relationship with DV but they have reported control variables in HTMT or Fornel & Larcker as well.
Control variables are not checked for Discriminant Validity
3. If we are using SmartPLS 4, we will check the control variables' effect in PLS-SEM as we did earlier in SmartPLS 3 or in its process tool?
You can use the normal model to check its effects.
I am highly obliged for your timely and positive response.
Regards
How to report it with full model . Where IV Dependent mediator all variables are there. I have found no video