Hey Mr. ...!! I recommend another tactic that we can use instead. And that is,to find the percent of our total sample which is 40 out of the total population which is 580, i.e;40/580×100 that results to 20% {0.2}. Then, because our aim is to balance our sample with available finite population (580), we'll be multiplying by 0.2 throughout the four grades (i.e; 120×0.2=8,150×0.2=10,130×0.2=10 and 180×0.2=12. And when we sum up these constants, they'll give us 40 (that is the total subset we're interested to study)). That is what I know Sir. ....!!
I know you said discrete data means we will not get a perfect even 40 with this sampling method. However, wouldn't rounding the amount of people in each strata influence the results of the study because not every strata is perfectly proportionally represented? Is there then a more accurate sampling method, (or perhaps statisticians just go with rough ideas and observe the results with a grain of salt)? Loved the video!! Very simple and helpful for visual learners. Thank you!!
Many times stratified samples actually take larger samples of the small populations and then just weight the results (proportionally scale them down). That way, you avoid some of the issues with really small samples while also being very precise with your proportions.
Jesus fuck! Thank you! I asked 5 people and 3 websites and finally got a simple and easy to understand explanation without the use of annoying jargon 'stratum'. I swear to gods, if I have to read that word one more time, I'm going to kill something.
That's actually 38 persons chosen. This is due to rounding issues as people are discrete and not continuous. You could do the same thing to choose the other 2.
The more precise solution to that is to keep the decimal weights, take too large of random samples in each category, and then multiply proportionally to get to the precise quantity of people per category. The actual calculations are not particularly hard (just a lot of multiplying), but conceptually that throws a lot of people off. This approach is simplistic in nature to convey the essence, but if you're performing samples for real-world applications, you would want to tighten up your approach a bit (though often this level of error is not going to be an issue if your sample is large enough and your strata/groups are somewhat balanced in size).
Hey Mr. ...!! I recommend another tactic that we can use instead. And that is,to find the percent of our total sample which is 40 out of the total population which is 580, i.e;40/580×100 that results to 20% {0.2}. Then, because our aim is to balance our sample with available finite population (580), we'll be multiplying by 0.2 throughout the four grades (i.e; 120×0.2=8,150×0.2=10,130×0.2=10 and 180×0.2=12. And when we sum up these constants, they'll give us 40 (that is the total subset we're interested to study)). That is what I know Sir. ....!!
No need to cross multiply just 120÷580 x 40
I never got the meaning of this until now. Thank you very much!
Thank you. I found your explaination amusing, simple and memorable.
Super quick and easy to grasp
Thank you, this video was pretty helpful
Very helpful! Thank you
I know you said discrete data means we will not get a perfect even 40 with this sampling method. However, wouldn't rounding the amount of people in each strata influence the results of the study because not every strata is perfectly proportionally represented? Is there then a more accurate sampling method, (or perhaps statisticians just go with rough ideas and observe the results with a grain of salt)? Loved the video!! Very simple and helpful for visual learners. Thank you!!
Many times stratified samples actually take larger samples of the small populations and then just weight the results (proportionally scale them down). That way, you avoid some of the issues with really small samples while also being very precise with your proportions.
great explanation for stratified random sampling when population is heterogeneous and you want to represent them in proportions....
thank you
Please answer me, what kind of the formula he explained?? I need the name of this formula😭
it's in the title 🙂
This is very helpful! Thank youuuuuu
Jesus fuck! Thank you! I asked 5 people and 3 websites and finally got a simple and easy to understand explanation without the use of annoying jargon 'stratum'. I swear to gods, if I have to read that word one more time, I'm going to kill something.
What is the name of the formula
When you add the results, the output of person picked was 30 ( 8 + 8 +10+12). N it said to choose 40 students so what happens next
That's actually 38 persons chosen. This is due to rounding issues as people are discrete and not continuous. You could do the same thing to choose the other 2.
The more precise solution to that is to keep the decimal weights, take too large of random samples in each category, and then multiply proportionally to get to the precise quantity of people per category. The actual calculations are not particularly hard (just a lot of multiplying), but conceptually that throws a lot of people off. This approach is simplistic in nature to convey the essence, but if you're performing samples for real-world applications, you would want to tighten up your approach a bit (though often this level of error is not going to be an issue if your sample is large enough and your strata/groups are somewhat balanced in size).
How did you find 10.34???
THATS WHAT IM SAYING I NEED TO FIGURE IT OUT
@@juhsselol283 120/580 x 40
120/580 x 40
Thank u !! Was fruitful tho!
I wonder you