ZTEST

What it does

ZTEST(array, x, [sigma])

ZTEST(array,x,[sigma]) returns a one-tailed normal-test probability for a hypothesised population mean x. Supply a known population standard deviation sigma, or let the function use sample standard deviation. This is the compatibility name.

Try it

=ZTEST({1,2,3},2,1) returns 0.5: the sample mean equals the hypothesis. If the sample mean exceeds x, this one-tail result is below one half.

Quirks

The public page documents #N/A for an empty array. The current helper has its own empty/nonfinite validation, so retain that error distinction in comparisons. A supplied scale must be positive, and an omitted scale needs enough sample data. A two-tailed test uses 2*MIN(p,1-p) under the stated worksheet composition, not simply 2*p in every case.

How to implement

Compute the numeric count and mean. If sigma is omitted, obtain sample standard deviation from the recovered variance family. The recorded current combination is z=(x-mean)*sqrt(n)/sigma, then the standard-normal left CDF at z. Do not rewrite it as a right-tail subtraction on (mean-x)/(sigma/sqrt(n)): mathematically equivalent forms have different stored results. The equal-mean example gives z=0 and probability one half. Inner normal tails still depend on the open ERFC factor.

Rust code

107 ztest "ZTEST" (2, 3) [R "array", V "x", V "sigma"] |s, a, sv| { z_test_of(s, a, sv) }

No implementation has been published with a Handbook test suite for this function.

Version differences

No same-input comparison between named Excel builds is recorded for this function here. The documented behaviour below is not a measurement of every Excel release.

Alternatives

Z.TEST is the modern spelling; T.TEST supplies a t-based comparison; NORM.S.DIST evaluates a normal CDF directly.

Named variants

No named variants are recorded yet.