CHISQINV Function (OpenOffice Calc)
Apache OpenOffice CHISQINV converts cumulative probabilities to chi-square cutoffs. Check the 3.1 history, endpoint rules and inverse examples.
Verification status: Official release source and independently calculated examples for CHISQINV were reviewed. No new Apache OpenOffice Calc application execution is claimed for this guide; the stated version and example limits still apply. How formula examples are checked
Quick answer: =CHISQINV(0.95;2) returns approximately 5.991464547108. It is the cutoff with 95% of the chi-square distribution on its left, not a 95% right-tail cutoff.
Availability and version history
- Introduced: Added in OpenOffice.org 3.1, whose release notes explicitly introduce this function.
- Supported in: Apache OpenOffice 4.1.16; no final supported release established.
- Scope: Desktop Calc. Argument contracts and boundaries below refer to the cited release.
- Checked: October 14, 2026. Documentation/source and independent mathematics checked; Apache OpenOffice Calc was not run for this guide.
What the inverse returns
CHISQINV solves CHISQDIST(x;k;TRUE())=p for x. The output is a chi-square value; the input p is a cumulative probability. For example, 95% and 0.95 represent the same probability, whereas the number 95 is outside the accepted interval. Official reference.
Its degree count describes the distribution, not a sample size that can always be entered unchanged. In a variance calculation with an estimated sample mean, the usual degree count is sample size minus one. Other statistical procedures can require different counts.
Syntax and allowed inputs
=CHISQINV(Probability; Degrees_of_freedom)
Both arguments are required. The release evaluator accepts 0≤Probability<1 and floors Degrees_of_freedom before requiring at least 1. Probability 0 corresponds to the lower boundary 0; probability 1 is rejected instead of being returned as an infinite cutoff.
A displayed 100% may be a rounded interior probability. Inspect the stored value before replacing it with the literal 1. Use integer degrees of freedom deliberately: an input such as 5.9 is evaluated with 5 by this implementation.
Reproducible reference cases
These independent mathematical values are rounded to twelve decimal places. Keep full precision in a worksheet used for an inverse check.
| Formula | Expected result, approximately |
|---|---|
=CHISQINV(0;2) |
0 |
=CHISQINV(.5;1) |
0.45493642312 |
=CHISQINV(.5;2) |
1.38629436112 |
=CHISQINV(.95;2) |
5.991464547108 |
=CHISQINV(.95;5) |
11.070497693516 |
=CHISQINV(.975;9) |
19.022767798642 |
=CHISQINV(.05;10) |
3.940299136119 |
=CHISQINV(.9;5.9) |
9.236356899781 |
Practical worksheet: a 5% upper-tail threshold
Put significance level 0.05 in A2 and degrees of freedom 5 in B2. In C2 use =CHISQINV(1-A2;B2), expected 11.070497693516. The expression 1−alpha matters: CHISQINV takes the left-tail probability, while a 5% upper-tail threshold has 95% on its left.
Check C2 with =CHISQDIST(C2;B2;1). It should be approximately 0.95. A cutoff produced by =CHISQINV(A2;B2) would instead put only 5% on the left and answer a different question.
Advanced example: a central probability interval
For nine degrees of freedom, =CHISQINV(0.025;9) and =CHISQINV(0.975;9) give approximately 2.700389499980 and 19.022767798642. The cumulative probability difference at those unrounded endpoints is 0.95. The interval is asymmetric; a chi-square distribution is not generally symmetric around its mean.
These are distribution quantiles, not automatically a confidence interval for a measured quantity. Any transformation into bounds for a population variance must use the appropriate statistical formula and assumptions. Do not label the raw chi-square cutoffs in the original measurement units.
Fractional shapes and solver limits
If a genuine noninteger gamma-equivalent degree parameter is intended, use GAMMAINV(p;k/2;2) rather than asking CHISQINV to preserve a fractional k. Both products document GAMMAINV separately; this is a model relationship, not a promise that all implementations have identical numerical errors.
The inverse uses an iterative solver, and the implementation can report failure to converge. Inputs near a probability endpoint and very large degree counts deserve additional installed-application testing. The reference cases cover ordinary values only.
CHIINV uses the older right-tail convention and therefore takes a different probability orientation. The absence of a dot in CHISQINV is part of the documented OpenOffice function name.