 # Stat 301 Statistics 1 Day 3 Statistical significance

• Slides: 11 Stat 301 – Statistics 1 Day 3: Statistical significance Logistics n n n Added practice quizzes (under Quizzes) Added solutions to investigations Added solutions to practice problems Last Time - Random Babies Last Time - Probability n The probability of a random event is the longrun proportion (or relative frequency) of times that the event would occur if the random process were repeated over and over under identical conditions. q Can simulate the random process to estimate the probability n q More trials give more accurate estimates Can use probability rules to calculate the exact value n Total number of equally likely outcomes: 24 Investigation 1. 1: Friend or Foe Bar Graph n 14 of 16 chose the helper toy 14/16 = 0. 875 Investigation 1. 1 What are different possible explanations for why more infants chose the helper toy over the hinderer toy? 1. There really is a genuine preference for the helper toy 2. It was just a fluke and infants don’t have a genuine preference n How might you convince someone that it’s NOT #2? n Investigation 1. 1 n n To convince someone it’s number #2, we need to demonstrate that 14 out of 16 is an unlikely outcome if each infant was just guessing? Can we simulate that? Investigation 1. 1 n n n Question (g): Flip a coin 16 times and record the number of heads = picking the helper toy Question (i): Combine results with the rest of the class Question (l)-(r) Investigation 1. 1 n n Does it appear to be surprising to get 14 heads in 16 coin tosses? Does it appear to be surprising for 14 infants to pick the helper toy if they don’t have a genuine preference? Do we have evidence that infants have a genuine preference? So we say the result of 14 helper choices is statistically significant – we don’t think it happened just by random chance alone To Do for Thursday n Finish Investigation 1. 1 through the Technology Detour and Study Conclusions q q n n n Use online solutions to fill in holes Read Summary (p. 26) Take Quiz 3 Finish HW 1 Preview binomial distribution? q q Module in Canvas Bring JMP/R to class?