# Thread: Goodness of fit and Benford's law

1. ## Goodness of fit and Benford's law

Hello,

for a statistics homework I have to find out if these data pass the goodness-of-fit test. I don't like the results very much though, so I'd appreciate if someone can look at this.
http://michal.sustr.sk/chi-kv-benford-en.xls

For each leading number I found out their frequency, and then calculated chi square, But the resulting chi square is so big, and I don't know how to set up the test so it can pass sometimes. Thank you very much

2. Originally Posted by fiso
Hello,

for a statistics homework I have to find out if these data pass the goodness-of-fit test. I don't like the results very much though, so I'd appreciate if someone can look at this.
http://michal.sustr.sk/chi-kv-benford-en.xls

For each leading number I found out their frequency, and then calculated chi square, But the resulting chi square is so big, and I don't know how to set up the test so it can pass sometimes. Thank you very much
1. You calculate the critical value for acceptance/rejection at the significance level you want and compare the value of $\displaystyle \chi^2$ you have with that to determine acceptance/rejection of the hypothesis.

2. You can use an asymptotic approximation for the complementary cumulative distribution (See Abromowitz and Stegun).

CB