Lecture 12 Outline of Last Lecture I Hypothesis Testing II Inferential Errors Psych 311 1nd Edition Outline of Current Lecture I Terminology of Hypothesis Testing Current Lecture I Terminology of Hypothesis Testing Null Hypothesis Ho no change no difference no relationship any difference we see between sample and population is due to sampling error and NOT IV Alternative Hypothsis H1 there is a change difference relationship any difference between out data and population is due to IV and NOT sampling error HT effect of IV effect of SE must interpret quantitative value outcome in regard to Ho and H1 formally testing Ho not H1 either reject Ho or fail to reject Ho never accept H1 Reject Ho IV is the cause of difference overall test stat will be large reject if value falls within extreme score Fail to Reject Ho Sampling error is cause of difference fail to reject if value doesn t fall within extreme scores alpha p 0 05 determines cutoff for what we consider extreme scores 5 These notes represent a detailed interpretation of the professor s lecture GradeBuddy is best used as a supplement to your own notes not as a substitute reference distribution used to dertermine extreme scores for Z test ref distr sampling distribution use UNT number value tells us whether test is true or not Type I Error false positive we say our result is due to IV when it isn t control using Type II Error false negative we say our result is due to sampling error when its not controled by Type I Error Rate rate at which you make a false positive is equal to if 0 05 then 5 of the time you ll make a Type I Error decrease level to decrsease Type I Error Rate makes it harder to reject Ho Type II Error Rate rate at which you make a false negative increase in order to increase statistical power power probability of correctly rejecting Ho when it s false power influenced by n increase n to increase power Type I and Type II Errors are inversly related by increasing type I error you decrease type II error and vice versa Effect Size magnitude of the effect since HT only tells us IF there is an affect not the size of the affect Cohen s d d lM l tells us what the effect size is d 0 00 the larger the value of d the larger the value of the effect of IV d 0 2 is samll effect d 0 5 is moderate effect d 0 8 is large effect Cohen s d accompanies result of our HT Effect Size Fail to Reject Ho Reject Ho Zero No error Type I Error Small Type II Error No error Large Type II Error No error
View Full Document
Unlocking...