PERCENTILE.EXC in Collabora Office: Exclusive Positions and the Small-Sample Window

• Introduced: original Collabora release and launch date are not established by the reviewed sources.
• Availability: currently documented in Collabora Office 26.04 desktop help.
• Platform scope: online and mobile deployments are not inferred from this desktop-help release.
• Checked: October 14, 2026. Function-specific documentation and independent calculations; no native Collabora execution.

=PERCENTILE.EXC(data;alpha)

Both arguments are required. With n numeric observations, the help says values below 1/(n+1) or above n/(n+1) cannot be interpolated and produce an error. The exact displayed error code is not established by this section.

The corresponding position model is (n+1)×alpha in sorted data. Whole-number positions select an observed value; intermediate positions interpolate between adjacent values. The examples use numeric cells and semicolon separators.

An alpha inside 0 to 1 can still fall outside the usable window for a small sample. Percentage formatting cannot fix a missing interpolation position.

Enter 10,30,50,90,150 in A2:A6:

=PERCENTILE.EXC(A2:A6;0.25)
=PERCENTILE.EXC(A2:A6;0.5)
=PERCENTILE.EXC(A2:A6;0.75)

Expected results: 20,50,120. For five values, the positions are 1.5,3,4.5. The first result is halfway from 10 to 30; the third is halfway from 90 to 150.

The median control lands on the middle value 50. The other two cutoffs need not be present in the input range.

For the same five-value range:

=PERCENTILE.EXC(A2:A6;0.1)
=PERCENTILE.EXC(A2:A6;0.9)

The positions are 0.6 and 5.4, outside positions 1 through 5. Both therefore reach the documented error condition. This guide does not substitute Excel’s native error label for Collabora’s unspecified display.

The mathematical usable alpha interval is 1/6 through 5/6 for this sample. Exact native floating-point behavior at expressions representing those fractions has not been tested; the worked success and failure controls stay away from those boundaries.

Enter integers 1 through 19 in C2:C20. Now n+1 is 20:

The same 10th and 90th percentile requests now lie at ordinary observed positions. Adding copied or invented records merely to make a percentile available would change the dataset rather than resolve a statistical limitation.

On the original five observations, the exclusive upper-quartile cutoff is 120. The inclusive PERCENTILE.INC model gives 90.

A rule accepting values at least the exclusive cutoff selects 150 alone. With the inclusive cutoff it selects 90 and 150. Record the function name and comparison operator when a threshold controls inclusion in a report.

For the lower quartile, the exclusive model gives 20 while the inclusive model gives 30. The relationship is about positional conventions; neither is a blanket instruction to prefer the higher or lower threshold.

Double each original reading. The exclusive quartile results become 40,100,240, exactly twice 20,50,120. Adding 5 instead gives 25,55,125.

Keep the original units and transformed units explicit. The positional rule is unchanged by these positive affine conversions, but arbitrary nonlinear conversions require separate analysis.

The function’s COM.MICROSOFT.PERCENTILE.EXC namespace is documented outside the cited OpenFormula 1.3 standard. No universal import/export guarantee follows from that note.

Mixed text, Boolean, empty-string and source-error rules remain unverified for this dedicated section. QUARTILE.EXC restricts the requested position to specific quartile indices and has its own documented sample-size conditions. Do not suppress an unavailable percentile as zero, because zero may be a genuine measurement.

• Collabora Office 26.04: PERCENTILE.EXC.
• Editorial testing standards.

Complete function-specific documentation reviewed. Worked outputs are independent mathematical models or explicitly documented rules, not application screenshots or native test results. Precision, type and error boundaries retain their stated limits.