f disrtibution

[Visakha]

can someone explain how to calculate F(15,15) at 0.2 level
15,15 aint in the tables so how do we go about it?
this is freaking me out

 

[John Lee]

can someone explain how to calculate F(15,15) at 0.2 level
15,15 aint in the tables so how do we go about it?
this is freaking me out

It shouldn't be in a CT3 exam as it's not possible (at least not with any accuracy) to obtain it from the Tables provided.

 

[Visakha]

It's a question from Nov 2008 of the Indian Institute exam.

 

[John Lee]

It's a question from Nov 2008 of the Indian Institute exam.

Q13a - fascinating. I can only assume that the Indian students have different tables or just the people who set the exam used different tables (and hence there was a mistake in the exam). Worth pursuing with the IAI I think...

 

[Visakha]

no we have the same tables..
anyways i just had a question.. suppose we need to calculate F(14,11) at 5% so can we use interpolation. If not how to we go abt it?

 

[Calum]

It could be done (I think; I haven't tried!) by considering the F(15,15) distribution as the ratio of two chi-squared variates and using that to derive the density - messy, but looking at the equations it looks analytically possible.

Not sure this is what the examiners were driving at though!

 

[John Lee]

no we have the same tables..
anyways i just had a question.. suppose we need to calculate F(14,11) at 5% so can we use interpolation. If not how to we go abt it?

Yup - though it won't be very accurate.