STAT 200 Chapter Notes - Chapter 25: Location Test, Null Hypothesis, Statistical Hypothesis Testing

• School: UBC
• Department: Statistics
• Course: STAT 200
• Professor: Eugenia Yu

Document Summary

Ch. 25: comparing means between two paired groups: dependent populations. Looking at within-pair differences di"s: estimate true mean difference b/w the two paired groups d by. D: sampling distribution of d : d . ~n( d, d/rt(n): use sd (sample sd of the differences of the pairs) to estimate d. Finding ci for the mean difference between two paired groups: Clt applied: when underlying distribution of di"s is non-normal or unknown, need a. Identify type of test: compute test statistic: work under the null h0: d = 0, case 1: is unknown use t model: paired t-test, check assumptions and conditions to work under the null model, random sample. 68-95-99. 7% rule here: look at degrees of freedom: n-1, pick the closest t-score in the table to the found test statistic (at the degrees of freedom, look at top: 1 tailed probability or 2 tailed probability iv.