NORMSINV and LEGACYNORMSINV in Calligra Sheets

• Introduced: First release not established. The name is registered in Calligra 2.4.0, which proves historical presence, not introduction.
• Availability: Supported in Calligra Sheets 26.08.2, according to the release-pinned registration.
• Scope: Desktop Sheets with the statistical functions module available. Names and separators can depend on locale.
• Checked: October 14, 2026. The NORMSINV worksheet examples are independent mathematical references. 19 selected LEGACYNORMSINV expressions ran through the 25.04.2 engine library, as detailed below; 26.08.2 was source-checked only, with no GUI-cell or import verification.

NORMSINV maps a cumulative left-tail probability to a standard-normal z-score. Its output is measured in standard deviations from the mean. It is an inverse CDF, not the reciprocal of NORMSDIST. KDE function handbook.

The mathematical relationship is z=Φ⁻¹(p)=√2×erf⁻¹(2p−1). Φ denotes the standard-normal cumulative distribution. These formulas describe an assumed probability model; successful calculation does not establish that a normal distribution fits a real dataset.

=NORMSINV(Number)

All listed arguments are required; there is no documented default to substitute for an omitted input. Use semicolons as shown in the English KDE examples, decimal points in the sample data, and the separators accepted by your installation. Release-pinned function help.

Enter the numeric inputs into the stated columns, then enter the indicated result formula. Fill through row 12. The expected results are mathematical references rounded to twelve decimal places; a displayed probability used as a new input can differ slightly from its unrounded original.

Choose the correct tail probability

A one-sided upper 5% cutoff leaves 95% below it: =NORMSINV(0.95) gives 1.644853626951. A central 95% interval leaves 2.5% in each tail, so its positive endpoint is =NORMSINV(0.975), approximately 1.959963984540. Passing 0.05 instead asks for a negative lower-tail cutoff.

Convert the z-score to measurement units

For a normal model with mean 100 and SD 10, the 90th-percentile measurement is:

=100+10*NORMSINV(0.9)

The expected result is 112.815515655446. This agrees mathematically with =NORMINV(0.9;100;10). If mean and SD are stored in B2 and C2, replace the literals with those cells and keep SD positive.

To check an interior probability, put 0.975 in A2, =NORMSINV(A2) in B2, and =NORMSDIST(B2) in C2. C2 should recover approximately 0.975.

The pinned release requires 0<p<1 and returns an internal value-error object otherwise. Neither endpoint has a finite standard-normal quantile. Use a numeric percentage or a fraction, not 95 when you mean 95%. The median p=0.5 is valid and returns the mathematical target zero. Release computation and guards.

Tail quantiles are sensitive to probability rounding. A probability copied from a displayed cell may be less precise than the stored value; reference the original cell when possible. An inverse/CDF round trip is useful for consistency, but it does not establish uniform accuracy throughout the tails.

An internal Value::errorVALUE() in the source identifies the error category, not a guarantee of the same visible token in every package or locale. Validate numeric input cells and units rather than depending on implicit text or blank-cell conversion.

=LEGACYNORMSINV(probability) is an active alias for NORMSINV. The 26.08.2 alias registration and the canonical registration use the same one-argument contract, argument-handling settings and complete callback. The alias is already present in Calligra 2.4.0 source; its first introduction is not established. The callback uses the same strict 0<p<1 probability guard and inverse-normal helper as NORMSINV. The LEGACY prefix does not identify a retired function.

The following 19 selected expressions ran through the official Calligra 25.04.2 engine library. Their exact function spelling is retained: the earlier NORMSINV worksheet remains a mathematical recipe, while these alias expressions provide the engine evidence. Version 26.08.2 was source-checked only. There is no GUI-cell or import claim.

Alias values and independent references

The engine and mathematical columns are separate. Values are rounded to twelve decimal places for presentation, not as a guarantee for every input.

Alias boundaries

Both p=0 and p=1 return #VALUE!; mathematically their standard-normal quantiles are infinite. Values outside [0,1] are also rejected. This is a deliberate strict endpoint guard, unlike some numerical searches elsewhere in the same module.

Numeric text “0.5” was accepted. Nonnumeric text, TRUE and FALSE failed the resulting strict domain check. Missing and extra arguments are errors. The engine evidence covers these formulas in 25.04.2; later execution and GUI-cell behavior remain untested.

Alias consistency check

The executed inverse/CDF check below recovers approximately 0.975. Its mathematical reference uses the inverse error function; a successful round trip is a consistency check rather than a guarantee throughout the tails.

The executed advanced check was:

=LEGACYNORMSDIST(LEGACYNORMSINV(.975))

Its engine result was 0.975.

Source and precision evidence

• Function-specific implementation
• Normal Inverse helper
• Conversion helper
• Function Default helper

The statistical and math modules and value-calculation helpers are byte-identical between the inspected 25.04.2 and 26.08.2 sources. That comparison supports the source discussion but does not establish later-version execution, GUI-cell behavior or other locales.

NORMDIST supplies the general normal CDF or density; NORMINV reverses its CDF. NORMSDIST and NORMSINV use mean zero and standard deviation one. Follow each function’s own argument contract instead of adding or omitting a flag by analogy.

Other spreadsheet programs may use dotted names such as NORM.DIST or NORM.S.INV. A related mathematical name does not prove that an imported formula, error display or locale separator transfers unchanged. Validate the actual target workbook when exchanging files.

• NORMSINV handbook entry
• Calligra 26.08.2 syntax and argument help
• Release-pinned calculation and domain guards