I'm looking to report a chi-squared statistic for a chi-squared goodness of fit test using a 2x2 table.

....Treatment 1.... Treatment 2 Yes 100 120 No 30 50

I'm looking to see if the proportion of Yes/No differs between the two treatments (equal=null, different=alternative hypothesis).

A chi-squared test is easily done by computing the expected values in excel, and using the CHITEST function to compute the p-value. However, I also need to report the chi-squared statistic itself, not just its associated p-value.

According to Microsoft, CHISQ.INV Returns the inverse of the left-tailed probability of the chi-squared distribution, while CHIINV returns the inverse of the right-tailed probability of the chi-squared distribution.

If I'm looking to report the chi-squared statistic for a chi-squared test (e.g. in APA format), which one should I use? Is there another way to have Excel report the chi-squared statistic for a 2x2 table?


  • $\begingroup$ There is not enough information to answer this question. Please explain the circumstances in which you are computing a chi-squared statistic and what test you are doing--its hypothesis and its null hypothesis. $\endgroup$
    – whuber
    Jul 27, 2015 at 15:34
  • $\begingroup$ My apologies - more details have been added $\endgroup$
    – mp4-12c
    Jul 27, 2015 at 18:56

1 Answer 1


The p-value for that chi-square you're doing is the area in the right tail beyond the test statistic.

Help for CHIINV says:

CHIINV(p, df) is the inverse function for CHIDIST(x, df). For any particular x, CHIDIST(x, df) returns the probability that a Chi-Square-distributed random variable with df degrees of freedom is greater than or equal to x.

So that's what you need.

As a check, compute the statistic by hand for an example you already know the statistic for (or one you make up; in either case with large enough expecteds that the chisquare approximation will be fine), and check that the p-value generated for that (using something that calculates a p-value for a chisquare) inverts to give the statistic you had or calculated (or very close to it)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.