I was curious if Google and Microsoft Live Search were biasing their results for..well let us just say less than friendly searches.  For example would the results for "google sucks" be different on both search engines.  Well of course the results are different, but so what?  What we really want to know is does the difference hold any signifigance?

Let's find out.

I used four terms and four search engines then ran a chi squared test to see if there was any significance.

Here are the observations I made:

Now these numbers don't mean much on their own, google might return a much higher number of results for all it's results.  Let's continue with the analysis.  Taking the sums and working through the second step of chi squared test, let's look at the expected results.

Some interesting numbers in there.  But nothing looks conclusive.  The only thing I might infer is that "sucks" is being over represented in the observed results and "hate" is being under represented.

Lastly let's look at the t-value and critical value and compare.

Well there isn't enough here to make any conclusive arguments.  So anyone looking for anti-Microsoft ammo or anti-Google ammo is going to have to go somewhere else to find it.