Please explain chi squared in really simple terms

Debate and discussion of any biological questions not pertaining to a particular topic.

Moderators: honeev, Leonid, amiradm, BioTeam

Post Reply
wier9
Garter
Garter
Posts: 1
Joined: Tue Jan 27, 2009 2:11 am

Please explain chi squared in really simple terms

Post by wier9 » Tue Jan 27, 2009 2:14 am

Can someone please explain the chi squared test in really simple language, im just not getting this and when i search WIKI about it they explain it in too much detail for me. I just need the basic concept and what it tells us.
Can someone explain using a simple example? Thanks!

csdambros
Garter
Garter
Posts: 1
Joined: Tue Jan 27, 2009 2:30 pm

Post by csdambros » Tue Jan 27, 2009 3:57 pm

Hi Wier9,
Chi squared is a simple test that compare two groups.
intuitively people do this. For exemple, if you see one boy fishing
and one girl plaing a game, you have not a suficient sample for a significative correlation.
If you use chi squared test the result is not significant at 5%. But quote that intuitively you know that one
observation it's little to assert boys fish and girls not. One observation about something it's not representative.
A boy fishing not mean that boys fish (as a whole).
But if you see 30 boys fishing independently and girls not, so you can assert that boys fish and girls not.
And if you use the test the result is significant at 5% (the result is less than 0,05).
So, the test improve our abilities of prediction based on a pattern.
The chi squared test tell us that the groups are different, based on our criteria, this is the 5%, but we can be more discriminative using 1%, or less, if you use 10%.

This numbers represent that there is 5% of chance of you are wrong (one error in 20), but the choice depend of the case.

Post Reply

Who is online

Users browsing this forum: Google Adsense [Bot] and 9 guests