NEUR 2002 Lecture Notes - Lecture 9: Null Hypothesis, Otaku, Statistical Hypothesis Testing
Document Summary
Set the criteria for decision-making, define alpha the alpha sets the tails of the normal distribution. This is the part where the sample mean is unlikely to be if h0 is true. Alpha is usually 0. 05, which is 5% (z score is -1. 96 and 1. 96) Make a decision; reject or fail to reject the h0. Types of error: hypothesis testing is an inferential process. We retain the null hypothesis when in reality it is true (good) We reject the null hypothesis when in reality it is true (type i error, alpha) We retain the null hypothesis when in reality it is false (type ii error, beta) We reject the null hypothesis when in reality it is false (good, power) Research random sampling - this is representative of the population from which it was derived. Independent observations - score from one individual is not influenced by the score form another individual. The value of is not altered by the treatment.