FINV in WPS Spreadsheet: Inverse F Right-Tail Thresholds
Find an F threshold from a right-tail probability, preserve numerator/denominator order and check the result using independent survival equations.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
FINV inverts the right-tail F probability described by FDIST. Given p and two degrees of freedom, it seeks a threshold x with P(F>x)=p. The output is a statistic threshold, not a probability or a reciprocal of the original statistic.
Syntax and error-documentation conflict
=FINV(probability, degrees_freedom1, degrees_freedom2)
The first degrees-of-freedom argument belongs to the numerator and the second to the denominator. Original controls use strictly interior probabilities and small positive integer degrees.
The formal invalid-probability inequality is incomplete in the lesson, and its narrative assigns #VALUE! to out-of-range probability while the formal error section refers to #NUM!. Exact invalid-input tokens and endpoint support remain unresolved here. No error prediction is derived from another spreadsheet product.
The lesson’s body names WPS 2019, but this is historical context rather than a numeric build or an introduction date.
Original exact inverse
For degrees (2,2), right-tail probability is 1/(1+x). Solving for x gives 1/p−1.
- Enter 0.25 in A2 and 2 in B2:C2.
- Use
=FINV(A2,B2,C2)in D2. - The mathematical result is 1/0.25−1=3.
- Check the original tail: 1/(1+3)=0.25.
A 25% right tail is a 75% lower tail. Passing 0.75 would produce a different threshold, one third in this example.
Independent controls
| Formula | Independent expected threshold |
|---|---|
=FINV(0.5,2,2) |
1 |
=FINV(0.25,2,2) |
3 |
=FINV(0.75,2,2) |
1/3 |
=FINV(0.1,2,2) |
9 |
=FINV(0.25,2,4) |
2 |
=FINV(0.75,4,2) |
0.5 |
For degrees (2,4), solving (1+x/2)^−2=p gives 2*(1/sqrt(p)−1). For degrees (4,2), solving 1−(2x/(1+2x))^2=p gives the final control. These provide independent known roots without spreadsheet execution.
Validate the probability, not only printed digits
Substitute a proposed threshold into the corresponding survival equation. The result should reproduce the requested probability within a tolerance appropriate for the calculation. Numerical inverse routines may return a nearby value whose printed digits differ while its probability residual is small.
A threshold is only one component of a statistical procedure. It does not calculate sample variances, select a one- versus two-sided test, or verify distributional assumptions. Choosing numerator and denominator order after looking at results can also change the inferential question; define the procedure before interpreting a threshold.
Source corroboration and limitations
The vendor example uses p=0.02 with numerator degrees 5 and denominator degrees 3, displaying 17.42876388. That supports its inverse right-tail meaning. The original controls remain at moderate probabilities with elementary survival formulas; they do not assert a convergence limit, extreme-tail precision or current-platform parity.
The lesson discusses fractional-degree truncation, but no original formula uses fractional degrees. Check that percentages are stored as fractions and that the degrees remain in the intended order. See NORMINV for a contrasting lower-tail normal quantile, and VARP for population-variance calculation under a separate contract.
Availability and evidence scope
The official WPS lesson is dated 2019-05-06 and explicitly labels WPS Office for Windows Personal. Its numeric build is unspecified. This documents the function by that date; it does not establish first introduction, the last functioning build, other-platform parity or a retirement/deprecation history.
The official lesson was rechecked on October 19, 2026. Original examples are documentation-derived independent calculations, not WPS or other spreadsheet-engine executions. Numeric reference values do not certify application precision or statistical model suitability.