Question

Explain why we might want to assess the strength of evidence against the null hypothesis, rather than delivering a simple “reject H0 or don’t reject H0” conclusion? Give an example of when it would benefit to describe the strength of evidence one may have to reject the null instead of simply stating “reject H0”

Answer 1

@jim_thompson5910

Answer 2

I'd like to discuss this some more.

Answer 3

life isn't in black and white. There is no simple "yes" or "no" answer

for example, let's say that we found that the p value is 0.0001 which is very small. We have a lot of evidence to reject the null. We consider this statistically significant and there's a lot of significance here. Ie we have strong evidence to reject the null. Keyword: strong

now let's say we do another test and the p value is 0.048
at alpha = 0.05 this means we do reject the null but notice how close the pvalue is to the alpha value. Sure it's smaller but not significantly smaller. We reject Ho but it's not a strong rejection. There is possibly some room for cases where we fail to reject Ho

Answer 4

so while it's useful to know if we reject or fail to reject, it's also good to know how strong of a rejection we have

if the strength isn't that high, then maybe more tests are needed. If the strength is overwhelming, then that makes it more certain the conclusion is correct

of course, nothing is ever perfect and there's going to be error

Answer 5

Okay I will digest that and do a little more reading. thanks

Answer 6

np