MATH 203 Lecture Notes - Lecture 20: Test Statistic, Complement Factor B

• School: McGill University
• Department: Mathematics & Statistics (Sci)
• Course: MATH 203
• Professor: Michael Wallace

Document Summary

Course evaluaion is open talk about the hot room environment: h0: p1 = p2 = p3 = 1/3 or h0: p1 = , p2=p3 = , ha: h0 is not true. Test staisic: k i=1 ( ni np i (0 ))2 (0) npi. X 2= if each np i (0) 5, large sample approximation satisfied. Npi (0): expected count in cell i if h0|t. Squaring the test staisics make it posiive; stabilize the results. Then we have evidence against ho at the signiicance level, where chisq (k-1) is the (1- ) right hand tail criical value. Ei = npi (0) = 90 * 1/3 = 30. 30 calculating deviation only not its direction: chisq0. 05(2) = 5. 99; 6. 67 > 5. 99. We reject h0 that all three sports are equally referred. For the two-way table, we can test the null. H0: factor a and factor b levels are assigned independently. Classiicaion by factor a is independent of classiicaion by factor b.