Type 1 or error, Reject H α0 when it is true. Type 2 or error, not rejecting H β0 when it is false. •For a population mean, known , Test: σ n x z /0 σ μ-= Ex) Given: n=49, H0 = <26, H μ a = >26, =5, =.05 and new =28.04. Get: RR= z>1.645(from ) μ σ α μ α Calculate: Z= (28.04-26)/(5/rad(49))= 2.886. Decide to reject H0 because 2.886>1.645 Ex) Given: n=16, H0 = =95, H μ a = 95, =1.2, =.01 and new =94.32. Get: RR= z<-2.575 or z>2.575 μ≠ σ α μ Calculate: Z=(94.32-95)/(1.2/rad(16))= -2.267, so we keep Ho and reject Ha •FOR BETA ERRORS d= '0-Look at table A.17. The curve represents n. Calculate B that way. Do not need the below table (I think?) Ha: > μ μ0 -+ Φ n z / '0 α ( found on Z table) φ Ha: < μ μ0 1-(the above) Ha: μ≠μ0 -+ Φ n z / '0 2 /- -+-Φ n z / '0 2 / Ex) n=10, H0 : =2.5 verses H μ a : >2.5 with =.05 and =.1, how likely that H μ α σ0 is not rejected when
This is the end of the preview. Sign up
access the rest of the document.