This video just made an entire semester of statistics make more sense. Thank you, this information will help out a lot for the final. Much Appreciated!!
I can't believe how simple this was! the brilliance is always in the basics! I've been trying to understand why instead of "thats just how it is" and neither my teacher or my textbook have made sense to me as to why! Thank you so much you made everything click after the coin example!
What an amazing explanation! I banged my head reading quite a lot, still didn't get it, and watching this video I will remember it forever now! Really great job James! Looking forward to watch more such videos!
I’m on Level 2 of the CFA designation and this blew my mind. Thank you for the clarity and simplicity. I feel like I’m walking away with a better understanding of
this is just so great i must appreciate your effor and the simplicity woth which you explained the concept,Thank You so much please continue to make such contributions
I completely understood your explanation. Having calculated the mean of the sample, the DOF reduces by one, as the nth value gets fixed. But, why can't I make the same argument for the whole population? Shouldn't DOF for its variance also reduce by one, since we already have the mean calculated? So, by that effect population variance shouod also be divided by N-1 ?
this is the most intuitive on df, the fact in reality that we never knew value of true population mean, and sd always come after, in most cases n-1 is the reality of our data because as soon as we come up with an estimate of an average' the last piece of information is no longer needed. because that last piece of datapoint violates randomness of data so it has to be tossed out
This video is of virtue! Degree of reedom is a very hard concept to understand and other internet searches do not help much. I cannot thank you enough for this exceptional explanation.
I don't know if you'll ever see this but I have a question that relates to the calculation of sample variance versus population variance, where for population variance calculation, you don't need the n-1 degrees of freedom. Given that you know the mean for the population as well as for the sample and so effectively an n-1 dof, which only 'n-1' not used for the calculation of population variance. And by the way, your explanation of degrees of freedom is absolutely brilliant. Thank you so much for this video
I understood from the degrees of freedom perspective but I still do not understand why the Nth value do not make any contribution to the value of Standard deviation/Variance
Absolutely brilliant and crystal clear explanation of this concept. Top notch! And the use of the drawing visuals is spot on too, instead of being just a gimic. Hats off to you, Sir!
Thanks for clearly explaining. My conclusion about df: Within a formula if you know already other parameters (not include values of samples), how many minimum values of samples do you need to know for imputing all values of entire samples.
I still don't get why the n-th value doesn't contribute to the standard deviation. How can we say that only n-1 pieces of information contribute to it when in the numerator, we calculate the square of the deviation from the mean for all n values? I understand that (x1-xbar) + ... + (xn - xbar) = 0, so if we know what (x1-xbar)^2 + ... + (xn-1 - xbar)^2 is, we can work out (xn-xbar). What I don't understand is why it matters. I mean, we HAVE all n values, we don't have to work anything out. And we use all n values in the formula, so why act as if we didn't?
@@alijulaeerad5258 Could you give an example of that? Are you talking about not knowing all the scores of the population, and estimating the population standard deviation from the sample standard deviation? If so, the sample scores must still all be known right? Or are you talking about not knowing all the scores in a sample? Then what standard deviation are you estimating? Also practically why do statisticians estimate standard deviation?
@@Han-ve8uh we try to estimate the standard deviation of the population because we do not have all the scores In the population we only have all the scores in the sample. So in order to estimate the standard deviation of the population we assume that we know all the scores but one because that one score doesn't contribute to the standard deviation.