In a previous blog we looked at how Type I error means we wrongly reject our null hypothesis
TYPE II ERROR COMES ABOUT WHEN WE WRONGLY ACCEPT OUR NULL HYPOTHESIS.
Say you’re developed a new drug. You give it to one patient and they don’t get better. One of two conclusions can be made at this point. Either the drug genuinely doesn’t work and so this is true negative. Or the drug does work but unfortunately not in this patient’s case and so this would be a false negative.
Type II Error is about too many false negatives in our results and not finding a relationship when there is one. This will mean that we will find our new drug isn’t better than the standard treatment (or placebo) when it actually is.
TYPE II ERROR IS ALSO CALLED BETA
In the above example you can see with one patient you can’t tell the difference between a true negative and a false positive.
This means we need to design our study with enough patients to ensure we can tell true and false negatives apart.
This brings us on to the next blog and Power…