Can anyone help me understand in simple terms what a one tail and two tailed test is and what is suitable for what?
Re: one tail/two tail p value.
one tailed test... when you are testing if one statistic is higher or lower than the other value and you already know that you are testing for which direction. For example, if you want to test if pollution in NY is "higher" this year compared to last year.
two-tailed test... when you do not assume any direction. The things you are comparing could be higher OR lower, that is, you are testing if they are DIFFERENT. For example, if you want to test whether pollution in NY is "different" this year compared to last year.
Re: one tail/two tail p value.
thank you. Makes sense.