TINV Function (Gnumeric)

• Introduced: First release not established from the evidence cited here.
• Supported in: Gnumeric 1.12.62, confirmed by its function registration. No retirement is established.
• Scope: Desktop Gnumeric with the Statistics Functions plugin. Tables use English comma-separated formulas.
• Checked: October 14, 2026. Tables were recalculated in 1.12.57 and finite reference values independently checked. Version 1.12.62 was source-checked, not executed.

TINV returns the nonnegative cutoff t such that the combined probability below −t and above +t is p. It is a two-tail inverse, not a left-tail percentile. Equivalently it uses the right-tail probability p/2. At p=0.05 with 10 degrees of freedom, each outside tail contains 2.5%, producing approximately 2.228138851986. Official function reference.

=TINV(p,dof)

All arguments shown are required. The version-pinned help and evaluator establish the Gnumeric contract.

The table records the tested release’s outputs, with finite mathematical references rounded to twelve decimal places. Keep full precision when using a result in a later calculation.

For a 95% two-sided interval for a normal-population mean with unknown SD, suppose n=11, sample mean=50 and sample SD=12. The degrees of freedom are 10. The margin is =TINV(0.05,10)*12/SQRT(11), approximately 8.061709694020. The interval is approximately 41.938290305980 to 58.061709694020, under the independent random-sample and normal-population assumptions. The margin is not an interval for 95% of individual measurements.

For a pre-specified one-sided upper 5% cutoff, use =TINV(0.10,10), approximately 1.812461122812. TINV’s input is twice the one-tail probability. Using 0.05 would produce a 2.5% upper-tail cutoff instead.

Check an ordinary result with =TDIST(TINV(0.05,10),10,2), which should recover approximately 0.05 when full precision is retained. To obtain the corresponding negative cutoff, negate the result. The function preserves fractional degrees of freedom, as illustrated by the 5.5-degree reference case; it does not automatically round an approximate degree count to an integer.

The tested p=1 result is zero, while p=0 gives #NUM! because no finite cutoff leaves zero combined tail probability. Probabilities outside [0,1] and degrees of freedom below 1 are invalid. Do not substitute a very large arbitrary number for an infinite-boundary error. For confidence intervals, p is the excluded total tail probability, so a 95% confidence procedure uses 0.05, not 0.95. Nonfinite numerical results are converted to worksheet errors.

These additional outcomes were observed in the same native recalculation. F5 was genuinely empty; it is not an omitted argument. Missing inputs still need deliberate validation, even if they are converted to a numeric zero.

The worked calculations do not certify every locale, imported text value, extreme parameter or later binary version. Shared numerical helpers can change even when a function’s wrapper remains the same.

• Gnumeric function reference
• Gnumeric 1.12.62 evaluator

The reference and boundary tables use recorded 1.12.57 command-line recalculation. Further model examples were independently calculated from their stated inputs; they are not additional native test claims.