writing help: Chi-square distribution - Help.com

Chi-square distribution

I have random sample - 5000 values of 0 and 1
The null hypothesis - this values are uniformy distributed
Computing x^2 (Chi-square statistic) after that i have no idea how to get the probability that my sample is unifromly distributed… :(

Found nice calculator:
http://stattrek.com/Tables/ChiSquare....
But i do not need web calculator or tables - i need formulas to write my own program…

This open post was written 12 months ago | V/U/S: 549, 6, 3 | Edit Post | Leave a reply | Report Post


Reciprocity (0) Reciprocation Failure -- The poster has NOT helped anyone else yet!

Since writing this post rivne may have helped people, but has not within the last 4 days. rivne is a verified member, has been around for 12 months and has 1 posts and 0 replies to their name.

Post Tags (12)

Replies (6)

Where were you?

Click and drag to move the map around. FAQ: How we place people on this map »
You can also watch events on Help.com as they happen
Mouse over the map for 2 seconds to see an expanded, interactive view

Pocoloco offline Verified User (1 year) Long Term User Shouts: 2 #
An Unknown Location | 12 months ago (29 minutes after post)

x^2 = [ ( n - 1 ) * s^2 ] / σ^2

where σ is the standard deviation of the population, s is the standard deviation of the sample, and n is the sample size.

Quote this reply Report this reply to moderators
Pocoloco offline Verified User (1 year) Long Term User Shouts: 2 #
An Unknown Location | 12 months ago (33 minutes after post)

Oh sorry, that’s the wrong one… that’s to calculate the chi-square statistic.

Quote this reply Report this reply to moderators
Pocoloco offline Verified User (1 year) Long Term User Shouts: 2 #
An Unknown Location | 12 months ago (42 minutes after post)
Quote this reply Report this reply to moderators
Debb offline Verified User (1 year, 4 months) Long Term User Shouts: 2 #
An Unknown Location | 12 months ago (10 hours, 41 minutes after post)

I could be wrong, but I think that all you have to do is compare your chi-square test, using x^2 = [ ( n - 1 ) * s^2 ] / σ^2 with that of your chi-square critical value. I’m thinking you have to use tables for this or just memorise it:

http://www.itl.nist.gov/div898/handbo…

So, if your x^2 chi-square critical value then you do not reject Ho.
Say what you calculated from
x^2 = [ ( n - 1 ) * s^2 ] / σ^2
= 0.8

you compare it with the value you get from the table (here we shall use degrees of freedom of 1 at 0.05 significance (see the table))

so the critical value would be 3.841,

so do not reject Ho and conclude that the values are uniformly distributed.

Ummm I think that youre doing a computing subject right? yea but I’m a maths student and suck at computing but yeahhh I’m thinking that all you have to do is compare all your results with 3.841 or whatever your degrees of freedom is.

yeah so if your x^2 is greater than your critical value, then you reject Ho and conclude that the values are not uniformly distributed.

Hopefully i didn’t get you confused.

Quote this reply Report this reply to moderators
Debb offline Verified User (1 year, 4 months) Long Term User Shouts: 2 #
An Unknown Location | 12 months ago (10 hours, 43 minutes after post)

Oops the second paragraph did not make any sense, here’s what i meant to say

So, if your x^2 chi-square is less than critical value then you do not reject Ho.
Say what you calculated from
x^2 = [ ( n - 1 ) * s^2 ] / σ^2
= 0.8

Quote this reply Report this reply to moderators
bruce102 offline Unverified User #
An Unknown Location | 5 months, 1 week ago (6 months, 3 weeks after post)

if x^2 are vey close to critical value, do wo still stick to the rule, reject or not reject

Quote this reply Report this reply to moderators
This account has been deactivated.

Invite Others to Help

A logged in and verified Help.com member has the ability to setup a Friends List and invite others to help with posts.