STDEVA in WPS Spreadsheet: Sample Standard Deviation with Mixed Values

STDEVA estimates sample standard deviation while including logical and text values under the WPS lesson’s conversion rules. Referenced 2, 4, TRUE, FALSE, pending become 2, 4, 1, 0, 0 in the independent model, giving expected standard deviation 1.673320053.

This is the square root of sample variance 2.8. It is a sample measure, not a population standard deviation.

STDEVA(value1,value2,...)

Supply the first sample value or range and, if needed, additional values or ranges. The source permits a single array or reference in place of separate arguments. It mentions 1 to 30 arguments in its historical description; that is not established as a current limit across builds. These examples use one small range.

The lesson says TRUE is evaluated as 1 and FALSE or text as 0. This guide applies that mapping to referenced logical cells and ordinary text. It does not claim tested coercion for every literal argument form.

Enter Score in A1. Put 2 in A2, 4 in A3, =TRUE() in A4, =FALSE() in A5 and pending in A6. Put =STDEVA(A2:A6) in C2.

Independent expected output: 1.673320053. The five modeled values sum to 7, giving mean 1.4. Their squared deviations sum to 11.2. Divide by n-1 = 4 to obtain 2.8, then take its square root.

For a manual check, explicitly write 2, 4, 1, 0, 0 in a helper column. Do not calculate a mean with a different text/logical exclusion rule and insert it into the formula; that would mix two data sets.

For 2, 4, 6, squared deviations total 8; divide by 2 and take the square root. Identical observations have no spread even when their common value is nonzero.

Enter =TRUE(), =TRUE(), =FALSE(), =FALSE() in H2:H5. Calculate =STDEVA(H2:H5).

Expected output: 0.5773502692. The converted sample is 1, 1, 0, 0, with mean 0.5 and squared deviations totalling 1. The sample denominator is 3, so the result is the square root of 1/3. Dividing by four instead would give a population standard deviation of 0.5.

This numerical spread is not the same as the fraction of TRUE values, which is 0.5 in this example. Decide whether you need a proportion or a standard deviation before interpreting a logical column statistically.

The separate STDEV lesson incorrectly recommends STDEVA for a population in one sentence. The STDEVA lesson itself explicitly says sample and n-1, and points to STDEVPA for a population. This guide follows that explicit contract rather than repeating the conflicting recommendation.

Check the selected range for headings and status text, since included text changes both the mean and sample count. A missing measurement described as pending should not be treated as zero unless that encoding is justified. Keep actual logical values distinct from text that spells TRUE or FALSE.

Fewer than two included observations cannot support the sample denominator. Exact error labels, empty strings, referenced errors and numeric-text coercion remain target-build checks. The source’s phrase about an unbiased n-1 method is not a claim that taking the square root produces an unbiased standard-deviation estimator for every distribution.

The official WPS STDEVA lesson was fetched afresh on October 19, 2026. Its edition banner says WPS Office for Windows Personal. The lesson dates from 2019; it does not establish introduction or retirement dates, exact build coverage, or current Mac, Linux, web and mobile behavior. Examples are independently calculated expectations, not observed WPS outputs. Native execution and genuine current-application screenshots remain unverified.

STDEV, STDEVPA, STDEV.S, AVERAGEA. Browse the WPS function library.