CRIM 320 Lecture Notes - Lecture 5: Situation Two, Chi-Squared Test, Chi-Squared Distribution
Document Summary
Goodman"s and kruskal"s gamma: how to find critical value, alpha level, p. Phi coefficient: nominal level of variables, conduct after doing chi square, measure of the strength of the association for 2x2 table ith nominal data, add on to the chi square test, variable from 0 to 1, formula a. 6: close to 1 = stronger relationship, close to 0 = weaker the relationship. Interpretation i. does not tell the direction of relationship. 50 = strong association ignore the -ve sign. If conduct chi square test, result is insignificant, phi will be weak b. c. d. e. 7: no point of testing phi if we accept h0 for the chi square. Key limitation: only appropriate for 2x2 table: 2 variables with 2 categories each, tables larger the 2x2 cramer"s v. 5 step interpretation: chi square test, h0, x square, df, p