BUSS1020 Lecture Notes - Lecture 9: Statistical Parameter, Data Cleansing, Statistical Hypothesis Testing
Document Summary
Hypothesis testing at the heart of all scientific enquiry. Allows us to effectively use a sample to determine the population value of a mean/proportion: customers) The alternative hypothesis, h1: h1 opposes h0 in some way, generally the hypothesis researchers are trying to find evidence for; the interesting one. The critical value of a test statistic creates a limit for decision making to convey how. )f the sample mean is (cid:494)close(cid:495) to the (0 null hypo not rejected. )f sample mean is (cid:494)far(cid:495) from (0 h0 is rejected far is (cid:494)far enough(cid:495) Note: type ) errors are the (cid:494)serious(cid:495) errors that we must minimise true. The confidence coefficient (1-(cid:2009)) is the probability of not rejecting h0 when. The confidence level of a hypothesis test is (1-(cid:2009)) x 100% The power of a statistical test (1-(cid:2010)) is the probability of rejecting h0 when false. Type i only occurs given h0 true. Type ii only occurs given h0 false.