I can't put into words how helpful this video is! I'm a working adult taking a data analytics certificate course. I did not do well in high school math because I needed the visual representation and context to make maths meaningful and memorable. If I had videos like this, I might have gone to college. Sincerely, thank you!
After thinking and ruminating about it for a solid half an hour straight, I think I finally understand. Basically you can’t have the STD (standard deviation) without getting the variance first. It all goes back to squaring each variables and their variation from the mean individually because we want the variance to be a positive value since a variance of 0 indicates all values are identical, thus you cannot have a variance with a negative value since it would entail nothing but rather a possible error in calculation. You could just change each individual value where the variation is a negative value, meaning that value falls below the mean value, to a positive value without having to square them to get the STD that way that would be tedious and counterintuitive especially when dealing with much larger sample sizes. In the end, squaring all values allows every value to be on the same level of measurement, whether with negative or positive variations, prevents having to do tedious guesswork of figuring out which values fall below the mean and changing those that do that have negative variation to positive. Please tell me I finally got it right 😢
thanl you so much! also your accent actually is very soothing so it makes learning easier, the slow pronunciations and stuff. my best math teacher in high school had a similar accent and I was an ace in that class :D!
Thanks!! Nice explanation. Just a query, shouldn't it be µ as opposed to x bar as it's being calculated for a population. Doesen't affect the logic though but still thought I should bring it to your attention.
Thanks for this, great explanation. Can elaborate on the reasoning for 1/(n-1)? Why is it (n-1) for sample populations? Why not, say, (n-2) or (n-3.56)? Thanks
When we draw samples from a population, we assume that the samples are drawn in such a way that their mean "s" is a true estimate for the population mean "μ". But there might be situations where we end up drawing samples from the extreme end of the population and their mean is either way less or way more from the actual population mean thereby resulting in a significant deviation of the sample variance from the actual population variance. To counter that, we use "n-1" in the denominator which provides an unbiased estimate for the true population variance. There are arguments as to why only "n-1" is used and not "n-2" or "n-3" "n-x" etc. but researchers after doing so may studies came to the conclusion that "n-1" tends to bring the sample variance much closer to the population variance as compared to "n-2", "n-3" "n-x" and so on.
thank you. Do you have a preferred resource or reference for statistics questions such as the one I had asked? I appreciate your response, thanks. @@pradeepkandpal7523
Hello Data Tab Team. Thank you so much for this wonderful explanation. I now understood it nicely. I have one question regarding the calculation of the position of the median. Why is n/2 used in continuous series and (n+1) /2 is used in discrete series to find median? I will highly appreciate for your explanation
Hi, thank you for your feedback! I'm afraid I don't know exactly what you mean! If the measured values of a variable are ordered by size, the value in the middle is the median. The median is therefore the "middle value" of a distribution. It leads to a division of the series into two parts: one half is smaller and one is larger than the median. If there is an odd number of characteristic values, then the median is a value that actually occurs. If there is an even number of characteristic carriers (persons), the two middle characteristics are added together and their sum is divided by two. I hope that helps! Regards Hannah
@@datatab Thank you so much it helped a lot. Sometime I see in formula to calculate median class for group data like class intervals (0-10, 10-20, 20-30, 30-40) it includes n/2 instead of (n+1)/2. So why there is differene in the formula?
Why can’t we calculate the standard divination right away ? By adding the differences in height and mean and just make them all positive and divide by n ?
You’re right. You could just change each individual value where the variation is a negative value without having to square them to get the STD that way but that would be tedious and counterintuitive especially when dealing with much larger sample sizes, especially when we are dealing with sizes in the thousands. In the end, squaring all values allows every value to be on the same level of measurement, whether with negative or positive variations, prevents having to do tedious guesswork of figuring out which values fall below the mean and changing those that do that have negative variation to positive.
The use of values (individual readings or Xi) to calculate variance is confusing. Just wondering from where those readings 173, 147, 161 etc came from? Kindly watch at 3:41 mins of vedio length. The values could have been (18-155)2 +(8-155)2+(15-155)2 etc...
In one case you square the distance between each point and the mean value and in the second case you just use the absolute value. Both therefore say something similar, but it is mostly the variance that is used. Regards Hannah
If I pull a card to try for the ace of spades (1/52 chance) and I win $500 each time I do, and $0 all the other times, how do I calculate the variance? each pull would be 1/52 (not 1/52 then 1/51 then 1/50...)
I have a doubt. We are squaring only the numerator in the variance but why do we put square root for the entire equation while calculating Standard Deviation. Shouldn't it be only the numerator that has to be square rooted?
You cannot really say that the variance is the standard deviation squared. That implies you can calculate the std to find the variance. This you cannot. Can you?
Hi, yes of course you can say that! It only says that you can calculate the variance from the standard deviation! Let's say you compare different research results. Four reported the results with the variance and one with the standard deviation, then nobody stops you from squaring the standard deviation to calculate the variance so you can compare all 5 results. And this is possible because the variance is the standard deviation squared! You can transform mathematical equations as you like!
![Descriptive Statistics [Simply explained]](http://i.ytimg.com/vi/FzujIYo9GYo/mqdefault.jpg)
I can't put into words how helpful this video is! I'm a working adult taking a data analytics certificate course. I did not do well in high school math because I needed the visual representation and context to make maths meaningful and memorable. If I had videos like this, I might have gone to college. Sincerely, thank you!
I love all your videos. The way you teach with patience, clarity and examples. You are a great teacher.
Glad you like them! Regards, Hannah
I came across many explanation of variance and SD.
None took the unit in to account. Great explanation.
Thanks!
I wanted to know about variance from TH-cam. Only your video served my purpose.
Many thanks for the nice Feedback!!!!
Very long and boring
Anyway she didnt explain what variance is
Thank you so much for your clear explanation! Self-learning is made easier for us international students with your help
It's amazing how easy to understand the teaching is. Thank you
Glad it was helpful! Regerds, Hannah
After thinking and ruminating about it for a solid half an hour straight, I think I finally understand. Basically you can’t have the STD (standard deviation) without getting the variance first.
It all goes back to squaring each variables and their variation from the mean individually because we want the variance to be a positive value since a variance of 0 indicates all values are identical, thus you cannot have a variance with a negative value since it would entail nothing but rather a possible error in calculation.
You could just change each individual value where the variation is a negative value, meaning that value falls below the mean value, to a positive value without having to square them to get the STD that way that would be tedious and counterintuitive especially when dealing with much larger sample sizes.
In the end, squaring all values allows every value to be on the same level of measurement, whether with negative or positive variations, prevents having to do tedious guesswork of figuring out which values fall below the mean and changing those that do that have negative variation to positive.
Please tell me I finally got it right 😢
Yes, that sounds good!!!
Excellent! I'm using the materials for my students. I'm from Uruguay. Thanks a lot!
Happy to help and thanks for the nice feedback : )
Variance in its simplest form👏👏
Many thanks!
thanl you so much! also your accent actually is very soothing so it makes learning easier, the slow pronunciations and stuff. my best math teacher in high school had a similar accent and I was an ace in that class :D!
Glad I could help and many thanks for your nice feedback!!! Regards, Hannah
Thanks!! Nice explanation. Just a query, shouldn't it be µ as opposed to x bar as it's being calculated for a population. Doesen't affect the logic though but still thought I should bring it to your attention.
Bunch of respect for you, my teacher❤
Your channel is extremely nice. Beautiful explanation.
Again many thanks!
Very nicely and simply explained
Thank you so much and now I understand the diferencence between variance & std😀
Very helpful and clear! Thanks so much!
Thank you for the easy explanation.✌️😊
My pleasure 😊
Thanks for this, great explanation. Can elaborate on the reasoning for 1/(n-1)? Why is it (n-1) for sample populations? Why not, say, (n-2) or (n-3.56)? Thanks
Hmm, the answer is a bit complicated and depends on the degrees of freedom. I will try to make a video about it!
When we draw samples from a population, we assume that the samples are drawn in such a way that their mean "s" is a true estimate for the population mean "μ". But there might be situations where we end up drawing samples from the extreme end of the population and their mean is either way less or way more from the actual population mean thereby resulting in a significant deviation of the sample variance from the actual population variance. To counter that, we use "n-1" in the denominator which provides an unbiased estimate for the true population variance.
There are arguments as to why only "n-1" is used and not "n-2" or "n-3" "n-x" etc. but researchers after doing so may studies came to the conclusion that "n-1" tends to bring the sample variance much closer to the population variance as compared to "n-2", "n-3" "n-x" and so on.
thank you. Do you have a preferred resource or reference for statistics questions such as the one I had asked? I appreciate your response, thanks.
@@pradeepkandpal7523
Thank for this awesome explanation.
Glad it was helpful!
the denominator when we calculate variance has to be n-1(i.e., 5 in this case) not n(i.e., 6)
Gracias ❣️
Hello Data Tab Team. Thank you so much for this wonderful explanation. I now understood it nicely. I have one question regarding the calculation of the position of the median. Why is n/2 used in continuous series and (n+1) /2 is used in discrete series to find median? I will highly appreciate for your explanation
Hi, thank you for your feedback! I'm afraid I don't know exactly what you mean! If the measured values of a variable are ordered by size, the value in the middle is the median. The median is therefore the "middle value" of a distribution. It leads to a division of the series into two parts: one half is smaller and one is larger than the median. If there is an odd number of characteristic values, then the median is a value that actually occurs. If there is an even number of characteristic carriers (persons), the two middle characteristics are added together and their sum is divided by two. I hope that helps! Regards Hannah
@@datatab Thank you so much it helped a lot. Sometime I see in formula to calculate median class for group data like class intervals (0-10, 10-20, 20-30, 30-40) it includes n/2 instead of (n+1)/2. So why there is differene in the formula?
@@rajanvk939 Unfortunately, I cannot give you an answer here. Perhaps it is because of whether it is an even or odd number of values.
Thank you! Excellent explanation!
Glad it was helpful!
Why can’t we calculate the standard divination right away ? By adding the differences in height and mean and just make them all positive and divide by n ?
You’re right. You could just change each individual value where the variation is a negative value without having to square them to get the STD that way but that would be tedious and counterintuitive especially when dealing with much larger sample sizes, especially when we are dealing with sizes in the thousands.
In the end, squaring all values allows every value to be on the same level of measurement, whether with negative or positive variations, prevents having to do tedious guesswork of figuring out which values fall below the mean and changing those that do that have negative variation to positive.
So good
Thanks : )
how would you tell a story using both? variance and MAD? what would the story look like? thanks trying to put these in reporting words
The use of values (individual readings or Xi) to calculate variance is confusing. Just wondering from where those readings 173, 147, 161 etc came from? Kindly watch at 3:41 mins of vedio length. The values could have been (18-155)2 +(8-155)2+(15-155)2 etc...
It is just the mean plus the difference in the picture. 18+155=173, 155-8=155,.... Regards, Hannah
thanks!
Welcome! : )
Why don’t we just square root the square unit only ?
It would be interpretable after that right ?
Excuse me. For the definition, not the formula, what are the differences between variance and mean deviation?
In one case you square the distance between each point and the mean value and in the second case you just use the absolute value. Both therefore say something similar, but it is mostly the variance that is used. Regards Hannah
@@datatab Thank you 🙏
Hi,can you tell how the value 133 you got should be interpreted in reference to the data .
Is the quadratic mean important to know? I didn't really understand that tbh.
Thanks Nice video
Welcome!
Thanks. This helped :)
If I pull a card to try for the ace of spades (1/52 chance) and I win $500 each time I do, and $0 all the other times, how do I calculate the variance? each pull would be 1/52 (not 1/52 then 1/51 then 1/50...)
I have a doubt. We are squaring only the numerator in the variance but why do we put square root for the entire equation while calculating Standard Deviation. Shouldn't it be only the numerator that has to be square rooted?
I think you're having an order of operations error.
Other than that, really great video! Thanks
Clear
1000000000 likes and thanks
You cannot really say that the variance is the standard deviation squared. That implies you can calculate the std to find the variance. This you cannot. Can you?
Hi, yes of course you can say that! It only says that you can calculate the variance from the standard deviation! Let's say you compare different research results. Four reported the results with the variance and one with the standard deviation, then nobody stops you from squaring the standard deviation to calculate the variance so you can compare all 5 results. And this is possible because the variance is the standard deviation squared! You can transform mathematical equations as you like!
it's pronounced hite like kite, not hate.