Very helpful, thank you very much. In my research I have 5 latent variables, and each one was measured by a questionnaire includes certain number of domains and each domain include several numbers of questions. So if you please I need to know what do you mean by "indicators" ? does this mean the questions or the domains, as it yields completely different sample size. If I use domains, it yields 400 participants, and if I use questions, it yields more than 1300. I am very confused. Thank you in advance for your help.
Hi, nice video. I wonder how i can calculate sample size for a path model (i mean a SEM with only observable variables) with medietors and two dependent variables? As far as i know this calculator needs at least one latent variable. Thanks in adavance
Dear Dr. Fawad, Thank you for the informative explanation in the video. I have a question: Since we calculate the sample size before conducting the study, how do you determine the number of observed variables to include in the analysis?
Thank you! I am wondering what is the difference between “Minimum sample size to detect effect” and "Minimum sample size for model structure". Could you please elaborate
Although, this is not much explanation of this on the site and on the internet, what i could grasp is that the sample size to detect an effect is sample size based on power to get significant results, while the sample size required for model structure is based variables to produce more reliable results that is the reason it is more higher in terms of the size.
Thank you so much for your help! It is so helpful. Can I ask two questions? What about the model without the latent variable? My model has only observed variables. Also, how can I cite this software?
In my model, I have second order constructs, so those sub-dimensions will also be counted into latent variables while entering the value in the latent variables box ??
If you are assessing the interrelationships between lower order constructs, then use count them. If it is only the higher order construct that are used for hypotheses testing, use higher order.
@@researchwithfawad So , Sir is it OK to calculate sample size and write in my thesis that , the calculations are based on "A priori same size calculator " ?
@ fawad Sir i have finite population and my sample size comes 750 sir i dont have any supporting refrence where minimum criteria is 600 or 700 i found many refrence klines ( 2002) he said minimum sample size should be 200 hair et al (2010)said no of items in q *10 Sir if u can give me any supporting refrence in social sciences where this 700 is taken plz i am in immnese need coz i couldnt find any one or if u have any research paper where there has supporting refrence
My study involves 6 latent constructs and 136 indicators, minimum sample size to detect effect is shown to be 161 and the recommended minimum sample size is 16,589...Kindly suggest what should I do now? I am a bit confused regarding the ideal sample size...I have considered 400 respondents for my research...
Good explanation. However, I was expecting you to show how and why a larger or smaller sample may be needed by making changes in desired effect size and number of variables. In general, a larger sample is needed to detect a small effect.
Dear Mr. Fawad, First of all, your posts have been really helpful for my academic studies. For this reason, I thank you very much. In the video, you chose the effect size as the medium effect size (0.3). I wonder if there is a reference study in social sciences that we can accept effect size as medium effect size?
Thank you, you have really saved the day. Many of us have not know this. A study base will help us to make a research robust.
Thanks. I am glad you liked it
Thank you sir, can you please help me to find best method to determine the sample size for Thesis in both Infinite and Finite population case.
Please watch
th-cam.com/video/JNQ99Bg2Og4/w-d-xo.html
Very helpful, thank you very much. In my research I have 5 latent variables, and each one was measured by a questionnaire includes certain number of domains and each domain include several numbers of questions. So if you please I need to know what do you mean by "indicators" ? does this mean the questions or the domains, as it yields completely different sample size. If I use domains, it yields 400 participants, and if I use questions, it yields more than 1300. I am very confused. Thank you in advance for your help.
Consider each statement as an indicator. So if i am measuring 5 latent variable and each has 5 items, so there are in total 25.
Hi. I have a question. Do your indicators shown in your video refers to the items? Thank you. From Peru.
Thanks for watching
Yes, indicators and items are used as interchangeable terms
nicely explained, thank you very much!
Glad it was helpful!
Hi, nice video. I wonder how i can calculate sample size for a path model (i mean a SEM with only observable variables) with medietors and two dependent variables? As far as i know this calculator needs at least one latent variable. Thanks in adavance
Thanks. I am glad you liked it. In such case you can use inverse square root method. Please watch
th-cam.com/video/yvBAjFiHZGE/w-d-xo.html
Dear Dr. Fawad,
Thank you for the informative explanation in the video. I have a question: Since we calculate the sample size before conducting the study, how do you determine the number of observed variables to include in the analysis?
This is based onnthe scale i use
Thank you! I am wondering what is the difference between “Minimum sample size to detect effect” and "Minimum sample size for model structure". Could you please elaborate
Although, this is not much explanation of this on the site and on the internet, what i could grasp is that the sample size to detect an effect is sample size based on power to get significant results, while the sample size required for model structure is based variables to produce more reliable results that is the reason it is more higher in terms of the size.
Thank you so much for your help! It is so helpful. Can I ask two questions? What about the model without the latent variable? My model has only observed variables. Also, how can I cite this software?
Thanks for watching. Use the count of observed variables.
very helpful. thank you very much
Glad it was helpful!
Thank you so much for your video
Thanks. I am glad you liked it
In my model, I have second order constructs, so those sub-dimensions will also be counted into latent variables while entering the value in the latent variables box ??
If you are assessing the interrelationships between lower order constructs, then use count them. If it is only the higher order construct that are used for hypotheses testing, use higher order.
@@researchwithfawad Okay thanks
Sir A- priori sample size calculator is acceptable to defend my Ph.D Thesis, if I do PLS-SEM in my thesis ?
Why not
@@researchwithfawad So , Sir is it OK to calculate sample size and write in my thesis that , the calculations are based on "A priori same size calculator " ?
What is effect size?
Can we consider taking sample size referring to Hair et al.(2010) instead of this
Yes, you can.
Hey, do the residual errors of observed variables count as latent variables (when you enter the number of each)?
Thanks for watching. No only the latent variables.
@@researchwithfawad Ok, thank you for your help!
@ fawad Sir i have finite population and my sample size comes 750 sir i dont have any supporting refrence where minimum criteria is 600 or 700 i found many refrence klines ( 2002) he said minimum sample size should be 200 hair et al (2010)said no of items in q *10
Sir if u can give me any supporting refrence in social sciences where this 700 is taken plz i am in immnese need coz i couldnt find any one or if u have any research paper where there has supporting refrence
You are over the minimum range, the references are good.
To whom we should ciatation this process
Daniel Soper website please
Are the Indicators the same as Items?
Thanks for watching. Yes they are
Thank you!
What if you have not worked on your indicators?
Better to have your questionnaire finalized. Otherwise, if it is a perception based study, a sample between over 200 is good enough (Roscoe, 1975).
@@researchwithfawad Thanks Fawad. I have been following your channel, its made my doctorate journey much easier.
How did you get that model....
The model is from one of my studies. The analysis was performed in SmartPLS.
My study involves 6 latent constructs and 136 indicators, minimum sample size to detect effect is shown to be 161 and the recommended minimum sample size is 16,589...Kindly suggest what should I do now? I am a bit confused regarding the ideal sample size...I have considered 400 respondents for my research...
Please go for Recommended Minimum Sample Size.
Thank you very much dear
Most welcome 😊
Good explanation.
However, I was expecting you to show how and why a larger or smaller sample may be needed by making changes in desired effect size and number of variables. In general, a larger sample is needed to detect a small effect.
Thanks. I am glad you liked the video.
Thanks for the worthy suggestion, i will surely try to do a more detailed video soon.
Assalamualaikum...sir If my population size is 152..what would be its sample size?
WAS. Please refer to the following table
home.kku.ac.th/sompong/guest_speaker/KrejcieandMorgan_article.pdf
That's great!
Thanks. I am glad you liked it.
Dear Mr. Fawad,
First of all, your posts have been really helpful for my academic studies. For this reason, I thank you very much.
In the video, you chose the effect size as the medium effect size (0.3). I wonder if there is a reference study in social sciences that we can accept effect size as medium effect size?
Thanks. I am glad you liked it.
You will need to see relevant studies in your area and see what is the normal effect size.