MATH 1050Y Lecture Notes - Lecture 2: Null Hypothesis, Statistical Parameter, Type I And Type Ii Errors
22 Jan 2018
School
Department
Course
Professor
Document Summary
Lecture #2 inference for means of large samples (continued) Hypothesis testing: allows us to use between 2 hypotheses, hypothesis is never about statistics, but always about parameters. In test for single means, the hypotheses will always be about the population mean u. In test for two means, the hypotheses will always be about the two population means. The null hypotheses (h0): status-quo are about population parameter (no change, no effect, no difference, always contain an = sign, assume is true, and you look for rare event. The alternative hypotheses (h1, ha: statement about population parameters, that would be true if the null hypotheses were false, always contain , or sign. Decision and potential mistakes in hypothesis testing: there are two types of decisions, and two potential mistakes that can occur. Reject h0: evidence is rare and usual, so the null hypothesises rejected (not supported as true) Test statistic: hypothesis test of one population mean (large samples)
