Sample or population?
Use population when your numbers are every member of the group (all students in a class). Use sample when they are a part used to estimate a larger group (50 customers out of thousands); the sample formula divides by n β 1 instead of n, which gives a slightly larger, unbiased estimate.
Ο = β( Ξ£(x β ΞΌ)Β² Γ· N ) and s = β( Ξ£(x β xΜ)Β² Γ· (n β 1) )
The 95% interval uses the normal approximation (mean Β± 1.96 Γ standard error), which suits samples of about 30 or more.