|
Administrator
|
Sure. The criterion is just one thing:
whether we know population standard deviation, "sigma", or not.
When sigma is known, then everything is happy by using z-score;
that is, you can choose just one number among 1.645, 1.96, or 2.58.
When sigma is unknown, then you have to replace it to sample standard deviation, "s".
By replacing it, we have a penalty. The penalty is the use of t-score;
that is, you need to find out the number from the t-table.
|
Locked
