Variance: why square

This video explains why we square deviations when calculating variance. Why does it always sum to zero if we don't square it? Because we calculated the mean right

[@user-kc1gp5mo7k]

Actually the best video and explanation on standard deviation, variance, and mean which uses real-life examples and explains with all simplicity, showing how the formulas were derived. Thanks so so much. Honestly, I am really grateful, and thankful that you made this video. I THINK YOU DESERVE A 5-STAR LIKE BUTTON.

[@kerv15]

Great explanation, thank you! I do wonder why absolute value is not used though, as I would think that would be more appropriate seeing that you wouldn't have to undo it.

[@r0cketRacoon]

why not absoluate value ?

[@marcoglara2012]

Thank you.

[@asapGooby]

You are a riot to watch and very informative thank you

[@user-Revanth791]

Taking square root on whole of variance results in dividing the sum of squares of "Wrong" by the square root of total number of "Wrong". So, why would you do that ?

[@klalrinchhana5680]

Very informative. Thank you

[@mohdharis1981]

I learnt an important point i.e because sum of the wrongs will always turn out to be Zero, therefore, you square them idividually and then take an average of the total which is a mean. And then you undo the squaring effect from the mean by square root.

[@hemanthvippalareddy9861]

DammCourse Won't the square root of a number be positive or negative.... For example square root of 4 is + or - 2? So, wont there be any uncertainty in standard deviation?

[@michelledigdecarvalhoperei144]

Você salvou a minha vida eu estava literalmente morrendo de curiosidade odeio aprender as coisas sem entender o porquê delas serem assim

[@rushi00717]

Very nice

[@Zar4thustr4]

hmmm ... No. I don't think that this is the deeper reason, I think that is just a side product? I could just sum up the absolute values of the means divide it by the amount and get a representation of the average error, and that would be just fine - I mean it could be done like that as a standard (and we do actually it is called "mean deviation"). In my opinion the reason why we use the average of squares is not just convinience (not having to sum up the absolute value of mean error) but that this way the outliers are weighted more, which has advantages in certain statistical problems (not in all). (When I say "outliers I mean valid values far away from the mean)

[@MrMamanDon]

I love you.

[@DaBestAround]

thanks from the United Kingdom 🇬🇧

[@rushi00717]

Love from India

[@juancamiloariasospina55]

I still don't get, why don't you just use absolute values?

[@user-kc1gp5mo7k]

😀

[@kxrlzzz795]

Im sorry, but it doesn't seem like you ever answered the question "why we square." The only difference you gave between a square and an absolute value is "its hard to undo an absolute value." But you cant undo a square either? Also, would "squared dogs" not be to "dogs" what "velocity" is to "acceleration" with respect to time, and inappropriate to use interchangeably?

[@Pratik-lm9gt]

this is cute

[@tonyhaddad1394]

Ye ye ye you just break math if you sum squares and you make square root you get a different logic absolut value is more logic