# Null Hypothesis

Hi,

I am suddenly not sure about the null hypothesis when computing BFs for t tests and correlations in JASP:

When I specify that my alternative hypothesis (H1) do be one-sided, what is the BF computed against?

is H0 that t/r=0?

or is it complementary of my H1 (such as if H1:r>0, H0:r<=0)?

Thanks,

M

## Comments

Hi,

If my understanding is correct, you only specify H1 here. H0 is always a real null hypothesis, i.e. that there is no difference/ correlation. There is no need to for H0 and H1 to jointly cover all conceivable effects.

Cheers!

Sebastiaan