Compare two independent means when both population standard deviations are known (or the samples are very large). You can also test against a non-zero hypothesized difference.
z = ((x̄₁ − x̄₂) − Δ₀) ÷ √(σ₁²/n₁ + σ₂²/n₂)
Example
Two machines fill bottles with known SDs of 9 ml and 10 ml. Samples of 60 and 55 bottles average 78.4 ml and 74.9 ml. The standard error is √(81/60 + 100/55) = 1.78, so z = 3.5 ÷ 1.78 = 1.97 and the two-tailed p = 0.049, a just-significant difference at α = 0.05.
Frequently asked questions
What if the population SDs are unknown?
Use the independent t-test, which estimates them from the samples. With 30+ per group the results are almost identical.