In the statistic course, Hypothesis testing section, I do not understand what it means to reject H0 at 5% and accept it at 1% significance level. There is this magnesium pill example where the H0: Mean before >= Mean after, and H1: Mean before < Mean after. This basically means that H0: the magnesium pill is not working, H1: the magnesium pill is working. Then after the calculation we get that P-value is 0.024 so at 5% significance level we reject H0 (so the magnesium pill is working) and at 1% level we accept H0 (so the magnesium pill is not working)
Does this mean that it is 2,4% that we believe the pill is working but in reality it does not?
Thank you in advance :).
The p-value shows the highest level of significance at which we can reject the null hypothesis.
If 2.4% is a level of significance which is good enough for you, then you can conclude that the pill is working.
Most often, we pick a significance level of 5% and compare the p-value with it.
The 365 Team