Hi Dr. Fawas, I have a question. According to Hair et al. (2017), shouldn't we start with reporting VIF values, R square, f square, and q square values before the hypotheses testing for structural model tesing?
Prof. if the measurement model is established for the whole relationships in the complete model and we still get insignificant levels in the structural model for most of the hypothesis. in this case, what should we do?
Hello Sir, I have one question: I have made four hypotheses that are directly related to higher-order constructs, and each higher-order construct has four lower-order constructs. So, the question is, should I also create hypotheses related to lower-order constructs?
I have tested on 2-tailed test and my t-value=1.891 and my p-value is 0.029 which is less than .05 but as you mentioned in your earlier video that t-value should be greater than 1.96 for two tailed
how can I run an moderation analysis, by using pls, for a model including 3 independent variables, 1 moderator (which influence the relation IV-DV for each IV) and 1 dependent variable ? and how to interpret because R square of DV depends to 3 independent simultaneously if I model all together.
Sir,i have a confusion in calculating p value,which provide 2 type value. In spss, when I drop IV(only one variable) to DV it provide p value 0.0000, But when I put All the IV(all the variables) to DV,it provide different result,p value 0.0689.. Which one is right sir,using One IV to DV or All variable to DV? This p value I used in hypothesis testing,I want to clear my confusion..
Sir, I have one confusion, when I am running my whole model including mediators - the result between iv and final dv became insignificant whereas if I run the model separately measuring direct relationship only no mediator relations - the result is significant. what does it mean? should I run the model 1st without mediators and then again with mediator to show SEM result and Mediation result separately? Please Help
@@researchwithfawad Sir I understand the concept, but my issue is that in my model IV and main DV are insignificant because of full mediation (serial). But my confusion is how can I claim full mediation as per my knowledge full mediation is only possible when one has checked all the possible mediators which is not possible then how should I justify my results.
Your videos are really helpful I want to know that while doing the structure model analysis do we use the sum of the all item to represent a variable and at this stage there is no need to consider formative or reflective constructs seperately
Sir,i have a confusion in calculating p value,which provide 2 type value. In spss, when I drop IV(only one variable) to DV it provide p value 0.0000, But when I put All the IV(all the variables) to DV,it provide different result,p value 0.0689.. Which one is right sir,using One IV to DV or All variable to DV? This p value I used in hypothesis testing,I want to clear my confusion..
Good Evening Sir, It was a great session. But in the Structural assessment table (time stamp 2.41) there is a column that mentions Standard Error (SE). Sir, where we will find the value of SE in smart PLS? Sir under the menu button (Result-->path coefficient--->) we only have values of Original sample (B), Sample Mean (O), Standard deviation, T statistics and P-values. Values of Standard Error is not there. Where I can find SE value?
Thanks for your interest. This video may help on how to add control variables in SmartPLS. I am yet to make on in SmartPLS4. This is using SmartPLS3 (Overall the procedure doesn't differ much) ua-cam.com/video/AjUt5FsHZiw/v-deo.html
@@researchwithfawad Sir I m using Hayes Approach for measurement model CB-SEM and OLS for hypothesis testing using R TO test mediating and moderation analysis.Is that right for (Likertscalae data)
Does negative beta coefficients(with insignificant results) mean that there is a negative relationship? We will need to explain the result that why our relationship is coming negative?
Negative means inverse relationship, negative beta that has p value over 0.05 means that the relationship between the variable is negative/inverse but insignificant.
@@researchwithfawad thank you Sir. Do we have to provide any reasons for why the relationship is coming negative even though its insignificant? I hypothesized it as a significant positive relationship but it is coming insignificant and negative.
Asalamualaikum, i am planning to use smartpls for data analysis in thesis. Some researchers are suggesting that i should do EFA in pilot study using spss. is it a correct approach. please suggest how should i go about in pilot study. what analysis should i do and what should i report in pilot study.
Hi Dr. Fawas, I have a question. According to Hair et al. (2017), shouldn't we start with reporting VIF values, R square, f square, and q square values before the hypotheses testing for structural model tesing?
You can do that.
Prof. if the measurement model is established for the whole relationships in the complete model and we still get insignificant levels in the structural model for most of the hypothesis. in this case, what should we do?
Measurement model assessment has no contribution in significance or lack of significance of relationships.
Thank u sir u resolved many issues .. God bless u
Pleasure. I am glad you liked it.
Hello Sir, I have one question: I have made four hypotheses that are directly related to higher-order constructs, and each higher-order construct has four lower-order constructs. So, the question is, should I also create hypotheses related to lower-order constructs?
This depends on the study objectives. But normally we don't do it at the lower order level.
You are great sir, thank you
You are most welcome
If beta is -ive have i write in hypothesis it has negative impact should i use word negative for my thesis
Yes.
Hi sir, when results show that hypothesis is proved, what does a negative beta coefficient mean?
Its mean a negative relationship.
Very clear, thanks a lot.
Thanks for watching
Dear Sir, can we use this way of reporting for both thesis and paper?
You can make it more detailed for thesis.
Sir what we call a structural SEM model where we have a relation like, A(Some variables) effect B and B effect C without any mediation?
This is possible. In this case you are only interested in assessing direct relationships.
@@researchwithfawad yes i want to check the direct relationship.
You will use the same procedure as in this video, just assess the direct relationships
ua-cam.com/video/khAH9f_xpRY/v-deo.html
I have tested on 2-tailed test and my t-value=1.891 and my p-value is 0.029 which is less than .05 but as you mentioned in your earlier video that t-value should be greater than 1.96 for two tailed
Share the results on kh.fawad83@gmail.com
how can I run an moderation analysis, by using pls, for a model including 3 independent variables, 1 moderator (which influence the relation IV-DV for each IV) and 1 dependent variable ? and how to interpret because R square of DV depends to 3 independent simultaneously if I model all together.
Please watch these video. These will help
ua-cam.com/video/53O2hOnVWY8/v-deo.html
ua-cam.com/video/isVAxbA5DOE/v-deo.html
Sir,i have a confusion in calculating p value,which provide 2 type value.
In spss, when I drop IV(only one variable) to DV it provide p value 0.0000,
But when I put All the IV(all the variables) to DV,it provide different result,p value 0.0689..
Which one is right sir,using One IV to DV or All variable to DV?
This p value I used in hypothesis testing,I want to clear my confusion..
Put all variables at once in the analysis.
Be blessed my dear 🙏🙏🙏🙏🙏
Thanks. I am glad you liked it.
Sir, is this one tailed, as t statistics of OC on OP is 1.873 and 1.708, which is greater than 1.64 @ 5%.
Did you select One Tailed from the Options
Sir, I have one confusion, when I am running my whole model including mediators - the result between iv and final dv became insignificant whereas if I run the model separately measuring direct relationship only no mediator relations - the result is significant. what does it mean? should I run the model 1st without mediators and then again with mediator to show SEM result and Mediation result separately? Please Help
Thanks for watching. Please watch
ua-cam.com/video/8lDwmMF9aN8/v-deo.html
@@researchwithfawad Sir I understand the concept, but my issue is that in my model IV and main DV are insignificant because of full mediation (serial). But my confusion is how can I claim full mediation as per my knowledge full mediation is only possible when one has checked all the possible mediators which is not possible then how should I justify my results.
Your videos are really helpful I want to know that while doing the structure model analysis do we use the sum of the all item to represent a variable and at this stage there is no need to consider formative or reflective constructs seperately
No need to take mean scores. All types of constructs must be analyzed together.
@@researchwithfawad many thanks for your reply
Thanks for the video
Thanks for watching.
Fantastic Session sir!❤
Glad you liked it!
Sir,i have a confusion in calculating p value,which provide 2 type value.
In spss, when I drop IV(only one variable) to DV it provide p value 0.0000,
But when I put All the IV(all the variables) to DV,it provide different result,p value 0.0689..
Which one is right sir,using One IV to DV or All variable to DV?
This p value I used in hypothesis testing,I want to clear my confusion..
Good Evening Sir, It was a great session. But in the Structural assessment table (time stamp 2.41) there is a column that mentions Standard Error (SE). Sir, where we will find the value of SE in smart PLS? Sir under the menu button (Result-->path coefficient--->) we only have values of Original sample (B), Sample Mean (O), Standard deviation, T statistics and P-values. Values of Standard Error is not there. Where I can find SE value?
Thanks for watching. Standard Error is the same as Standard Deviation.
@@researchwithfawad Thank You So much Sir, really wonderful to learn from you....
Thanks a lot
dear sir how to find out impact independent categorical variables on depend variable ..
Thanks for your interest. This video may help on how to add control variables in SmartPLS. I am yet to make on in SmartPLS4. This is using SmartPLS3 (Overall the procedure doesn't differ much)
ua-cam.com/video/AjUt5FsHZiw/v-deo.html
So informative
Glad it was helpful!
Hello Sir,Can we apply OLS regression for likert scale data using R
Yes you can. R has its own libraries and functions.
@@researchwithfawad Sir I m using Hayes Approach for measurement model CB-SEM and OLS for hypothesis testing using R TO test mediating and moderation analysis.Is that right for (Likertscalae data)
Hayes macro can test Mediation and Moderation and if you are using Lavaan in R, all of this analysis can be done using Lavaan.
Does negative beta coefficients(with insignificant results) mean that there is a negative relationship? We will need to explain the result that why our relationship is coming negative?
Negative means inverse relationship, negative beta that has p value over 0.05 means that the relationship between the variable is negative/inverse but insignificant.
@@researchwithfawad thank you Sir. Do we have to provide any reasons for why the relationship is coming negative even though its insignificant? I hypothesized it as a significant positive relationship but it is coming insignificant and negative.
Asalamualaikum, i am planning to use smartpls for data analysis in thesis. Some researchers are suggesting that i should do EFA in pilot study using spss. is it a correct approach. please suggest how should i go about in pilot study. what analysis should i do and what should i report in pilot study.
No need to go for EFA. If your measures are adapted from existing reliable and validated scales.
pls,, sir could you share this word template
Thanks for your interest. I am sorry but i lost much of my resources to a ransomware virus. You will have to rely on the video.
Salam Alaikum brother, the video is not valid anymore. Kindly re-upload it.
WAS. Thanks for watching. It is working. Please check this link
ua-cam.com/video/rCMWcNJLUr4/v-deo.html
I found u little fast in this video 😅
Agreed. I am sometimes a bit faster than expected.
@@researchwithfawad anyway thank you so much