SOC SCI 10B Chapter Notes - Chapter 15: Null Hypothesis, Dependent And Independent Variables, Contingency Table
Document Summary
15. 1 x^2 hypothesis tests (one variable or goodness of fir tests) A nominal scale variable divides a population into separate categories and a dichotomous variable is one that divides the population into just two categories: many nominal scale variables have more that two categories. Hypotheses about these variables cannot be tested as if the variables were dichotomous. This kind of test is referred to as a goodness of t test, because ti checks if the proportions obtained from the sample data seem to t well with a set of proportions speci ed in advance. The number of degrees of freedom for a one variable x^2 test is (the number of different categories) -1. If the tables are arranged as in the format above this is equal to (the number of rows in each table) -1. The frequencies represent count data, which no case appearing in more than one cell in the table. The cases scores are independent of each other.