MGEB12H3 Study Guide - Midterm Guide: Type I And Type Ii Errors, Test Statistic, Confidence Interval
14 Apr 2014
School
Department
Course
Professor
Document Summary
Rules: with =/ hypothesis must be critical values (c) h0 or symmetric. The farther away c1 & c2 are, the harder to conclude ha or h0. If you want to control a to be smaller, then will become larger, and vice versa. Step 1: set up null & alt hypothesis. Step 2: decision/rejectuin rule eg. (x>c, reject h, x c accept h) Can write in terms of alpha, x or z (3 ways) When you change the decision rule, the conclusion changes. Step 3: look @ sample results (eg. x) A is given a p value % and is associated with. When doing a u test, a must be divided by 2 as it is a 2 sided test. Assume std dev is known & n large ( 30) or normal. When std dev is unknown, z value test statistic: a test. Use t chart & degree of freedom chart.
