STAT-S 300 Lecture Notes - Lecture 27: Test Statistic, Random Assignment, Categorical Variable
Document Summary
Comparing observed and expected counts: the chi-square statistic: chi-square goodness-of-fit (gof) test- significance test using chi-square ( kye - ) test statistic. Compares observed counts from sample with expected counts when h0 is true. The more observed differ from expected, the more evidence against h0: observed counts = numbers from sample, expected counts = claimed (null) proportion sample size, chi-square statistic- measure of how far observed counts are from expected counts. The chi-square distributions and p-values: chi-square distribution. Density curve skewed right and always positive- value can never be negative due to squaring in calculations. No confidence intervals for: degrees of freedom (df) More df = more normal df = number of categories (k) 1 df = center of distribution. For df > 2, mode/peak = df 2. Chi-square gof test: state- h0: list all proportions as equal to number (or) the distribution of context is as claimed by context .
