QTM 100 Lecture Notes - Lecture 19: Contingency Table, Randomized Experiment, Normal Distribution

• School: Emory University
• Department: Quantitative Theory & Methods
• Course: QTM 100
• Professor: Prof.Gong

Document Summary

Lec 19: 2 sample z-test & chi squared test. Hypothesis testing (2 sample z-test: ho: p1, ho: p1. 0: z-test statistic, standard error (for hypothesis test *note is different than ci, p = pooled estimate: pooling the total # of successes and observations from the 2 samples. The 2 test of association/independence assesses: ho: 2 variables are independent (not associated, ha: 2 variables are dependent (associated) 2 test: randomization (random sampling or randomized experiment, observations are independent, expected cell count 5 in all cells. Chi-squared test statistic summarizes the difference between the observed and expected counts o. Characteristics of the chi-squared test: chi-squared distribution is always positive (the test statistic cannot be negative, chi-squared distribution is defined by degrees of freedom: df = (r 1) (c 1, associations: : the two variables are independent (not associated) is true, the . 2 test statistic is relatively small (close to zero) Large chi-squared test statistics are evidence against h.