R & F-test

[Anonymous No.4547898]

Does anyone have experience with R (google's statistics language)?

So I'm using R to figure out if the variances of two populations are equal or not.

Here are the results:
> var.test(best2.txt[,1], best.txt[,1])

F test to compare two variances

data: best2.txt[, 1] and best.txt[, 1]
F = 2.0983, num df = 29, denom df = 29, p-value = 0.05039
alternative hypothesis: true ratio of variances is not equal to 1
95 percent confidence interval:
0.9987234 4.4085482
sample estimates:
ratio of variances
2.09831

What the fuck? How can I determine whether or not to reject the null hypothesis and accept the alternative?

[Anonymous No.4547941]

Don't bother.

[Anonymous No.4547952]

>p-value = 0.05039

[Anonymous No.4547962]

you just need p-value (for some confidence level) to be lower than some number (look up in statistics book)

Also use Octave

[Anonymous No.4547978]

>>4547962
>Also use Octave

Octave is not for statistics. R is a way better statistics software than octave will ever be.

[Anonymous No.4547984]

>>4547978

Go on...

[OP No.4547990]

>>4547962
>(look up in statistics book)

So I have to look this shit up? Why doesn't R provide that for me?

[Anonymous No.4547991]

>(google's statistics language)

Lolwut?

[Anonymous No.4547996]

p-value >5% => do not reject Ho
p-value <5% => reject

Don't know what to do in your example...

[OP No.4548017]

>>4547991
Wow I have no idea why I thought that.

In trying to figure out why I thought it was google related, I found this though:
http://rseek.org/

Dunno if it's helpful though.

[Anonymous No.4548024]

>>4547990
Because programs are not meant to think for you.

[Anonymous No.4548032]

>>4547978
Eh, I do all my statistics in MATLAB and I can't see how one is any worse than the other.

[OP No.4548070]

Okay so after listening to what you guys said, and what I can find online, I have two different ways of figuring this out:

Method 1) Compare F value to F-critical values:

First, I need to look up the F-critical values:
> qf(.025, 29, 29)
[1] 0.4759648
> qf(.975, 29, 29)
[1] 2.100996

Reject H0 if F < 0.4759648 or F > 2.100996. F = 2.0983. so I accept the null hypothesis that the variances are equal.

Method 2)
>>4547996
p-value >5% => do not reject Ho
p-value <5% => reject

p-value = 0.05039, so I reject the null hypothesis that the variances are equal and accept the alternative.

..so which is it?

[Anonymous No.4548080]

>>4548070
But your p-value is >5%.
Not a lot but a little.

[Anonymous No.4548082]

>>4548080
*Not by a lot

[OP No.4548094]

>>4548080

Oh fuck me. Yeah, I don't know how I read that wrong.

wtf is wrong with me today.

[Anonymous No.4548115]

>>4548070

Well it looks like they give the same result, so since you dont have to do the extra function calls of qf with Method 2, I'd use that.

[Anonymous No.4548133]

If pee is low, you must reject the Hoe.