# Using z statistics instead of t statistics

Hi!

Why do we use z statistics in "Rejection region and significance level" video?

The population variance is unknown and according to previous videos we should use Student's t-distribution instead of Normal distribution?

Significance Level is what you choose before performing the hypothesis test.

Rejection region depends on the Significance level.

when to use normal distribution and when to use student's t distribution?

* use Normal Distribution when your sample size is larger (say greater than 30)

* if the sample size is large according to Central limit theorem, the underlying data distribution is normal

* you have to use z- statistic

* for unknown population variance, substitute sample variance in place of population variance in the formula

* student's distribution is used for smaller samples, for sample sizes less than 30

* you calculate t-statistic in this case

* if you know the population variance the use the population variance in t-statistic formula, if not substitute sample variance.