00:01
So there are two general cases where we use the t distribution rather than the z or z distribution.
00:08
Those are for small sample sizes and unknown population standard deviation.
00:22
The reason why it matters is that for a small sample size, we have that for a small number of degrees of freedom, which correlates exactly with a small sample size, we have that for a small degrees of freedom, the t distribution has heavier tails, heavier tails than the normal distribution.
00:54
So if we have that the normal distribution would be something like this, the t distribution, something a little bit more like that.
01:05
So when we have these smaller sample sizes, we have that the t distribution has heavier tails, which matters because we have that the standard error for sample means, well we know that the standard error for sample means would be equal to the population standard deviation divided by the square root of n...