3.Significance level Ð setting the **a**

n**a** - our choice of probability (p) value Ð (.05 - .01 -
.001 - etc.)

n.05 = 5 (out of 100) chances of being different

n.p < **a** - reject H0, accept H1 (this is
statistically significant)

n.p > **a** - accept H0, reject H1 (this is __NOT__
statistically significant)

4.Rejection Region

nreject H0 = sample x
is significantly different than **m**

nCritical values Ð
calculated test values

nz-scores (must know **m**** **and **σ)**

nt-scores (when **m**** **and **σ** are not
known Ð more common method)

nIf sample z-scores
in tail Ð reject the H0 (*you
found sig. diff)*

nif sample z-scores
in body Ð accept the H0 (*you
found NO sig.)*