NORMINV in WPS Spreadsheet: Normal Quantiles and Location-Scale

=NORMINV(probability, mean, standard_dev)

Use a positive standard deviation and a probability strictly between zero and one for the examples here. The WPS narrative describes probability with inclusive brackets, but a formal probability-error inequality is incomplete in the published lesson. This guide therefore does not claim support or an exact error token at 0 or 1. Mathematically, finite normal quantiles correspond to strictly interior probabilities.

The lesson states the defining equation: the normal cumulative distribution at the returned x should equal probability. It also describes iterative searching and #N/A after 100 iterations without convergence. Those are documented historical statements, not a measured guarantee for arbitrary inputs.

Enter 0.75 in A2, 10 in B2 and 2 in C2. Use =NORMINV(A2,B2,C2) in D2.

The standard-normal 75th percentile is approximately 0.674489750196082. Scale by 2 and shift by 10: expected x is 11.3489795003922. It lies above the mean because the requested lower-tail probability exceeds one half.

To verify the mathematics, standardize the result back to z=(x−10)/2 and evaluate the independent normal CDF. It returns 0.75 up to numerical rounding. This probability-space check is more informative than demanding that two implementations print identical final digits.

The median equals the mean in this symmetric model. The lower and upper quartiles in the first distribution are equally far from 10. Doubling the location and scale doubles the corresponding quantile, as the last row shows.

For mean 10 and standard deviation 2, the central 95% model interval uses lower-tail probabilities 0.025 and 0.975. Symmetry gives approximate endpoints 6.08007203091989 and 13.9199279690801. Their difference is about 7.83985593816022.

That is an interval for values drawn from a specified normal distribution. It is not automatically a confidence interval for an estimated mean or a prediction interval accounting for parameter uncertainty. Selecting a probability does not validate the model or provide a decision policy.

The vendor example probability 0.088, mean 4.8 and standard deviation 2.8 displays 1.011112367. Independent normal-quantile arithmetic corroborates that rounded result. The lesson notes that inverse accuracy depends on the forward distribution calculation.

If a quantile falls on the unexpected side of the mean, check whether the input is a lower-tail probability or an upper-tail probability. For example, a 5% upper-tail threshold uses lower-tail probability 0.95. Keep the probability as 0.95, not 95. Nonnumeric arguments and nonpositive standard deviation have documented error remarks; exact endpoints and extreme-tail precision remain unresolved here. See PERCENTILE for an empirical data threshold, which answers a different question from a fitted normal quantile.

The official WPS lesson is dated 2019-05-20 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 complete lesson and its formula images and walkthrough were inspected on October 18, 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.