Statistical Sciences 1024A/B Chapter Notes - Chapter 22: Null Hypothesis, Test Statistic, Statistical Parameter
Document Summary
Categorical variables: category of interest success, all other categories failure. Theoretical sampling distribution of p the distribution of all p "s from all possible sample of the same size from the same population. Properties: mean p, standard deviation (cid:4666)(cid:2869) (cid:4667, normal distribution of n(p, (cid:4666)(cid:2869) (cid:4667, np 10, n(1 p) 10. Population is at least 10x larger than sample (n 10. Assuming p = 0. 5 is true, what is the probability of getting 80 or more boys in 100 births? (what is the probability that p 0. 8?) Answer: z = 6 is off the chart 0% Imagine taking all possible sample to create the sampling distribution of p for every inference impossible. Instead, we predict the sampling distribution of p using facts about center, spread and shape (using clt) Use p because p is unknown z* is the standard normal density curve with area between z* and +z* Use z* not t* because we are not estimating an additional parameter.