Wikipedia:Reference desk/Archives/Mathematics/2014 February 25

= February 25 =

Pearson's chi square for null values in the "expected" range
I wanted to use MS Excel to calculate p-values using Pearson's (new and improved!) Chi square formula. The problem I've run into is that some of the values in my "expected" range are zero. Excel therefore gives the error #DIV/0. How can I proceed? Do I need to find an alternative statistical test or could I just put very low values (e.g 0.001). In case that was incoherent, I made an example to see if the red-blue berry ratio was significantly different in my garden compared to my friend's garden; the relevant cells are highlighted. — Preceding unsigned comment added by 129.215.47.59 (talk) 12:00, 25 February 2014 (UTC)
 * Actually, nevermind. There are a load of promising results after searching Google for "pearson chi square zero". 129.215.47.59 (talk) 12:04, 25 February 2014 (UTC)