Confidence Intervals for the Ratio of Population Variances

I develop the appropriate confidence interval for the ratio of population variances (when sampling from normally distributed populations), then work through an example. Note that these methods do not work well when the populations are not normally distributed.

[samestate]

I am irritated yet thankful that you clearly explained this in eight minutes when my prof couldn't in two hours.

[HazieRox]

if it wasn't for this, and the 20 other video's i've watched today, id probably fail my final friday. thanks so much! yours are seriously the best stats videos on youtube.

[jbstatistics]

You're welcome, and thanks for the kind words!

[dursung_]

Super vid for beginners. I was quite terrified like "did I understand correctly". I see I did. Thx

[jbstatistics]

Finding the left value is a little tricky if you're using a table. If you click on one of the annotation links at that point I go into finding that values in detail. (In the video "Finding areas and Percentiles for the F distribution...")

[puzzleman]

@İsmet Evren, for that part, F_v, u where v =3 and u = 14 (look up f tables). Take the inverse of this number and you get 0.0700..

[purplecoathanger]

You make fantastic stats videos but this gets my like just for the amount of time you must have spent weighing all those boxes of cereal to get the variances of their weights

[yusufahmed]

without you, I would be working at McDonald's

[elsyarahmawati]

how you define the left tail 0, 0700 ? because i got with the formula (1/(f 14, 3, 1-a/2) is (1/26, 99= 0, 037) please correct me if im wrong


thank you

[wellnoturchingu]

where did those values come from ? 4.2417 and 0.0700

[extramiles]

what if means of the two population are known? do the degrees of freedom remain same as n1-1 and n2-1?

and F distribution is not symmetric can we say this interval is shortest?

[rugbywowsorted]

Is there an easy way to prove that this ratio follows an F distribution? Brilliant vid btw

[nyashadzashechayambuka]

thank you so much. this is really helpful for my statistics course. Can't wait for the exams 🤓. Also do we always divide alpha by 2 when looking at the F distribution I have the same question for the t distribution and the chi square distribution

[ismetevren]

How did you find 0, 07 ? I think it should be 1 / 4.241 = 0.236. Please correct me if I'm wrong. Thanks