PSYCH 100A Lecture Notes - Lecture 15: Null Hypothesis, Level Of Measurement
Document Summary
Parametric tests: tests hypotheses about specific population parameters, require data from an interval or ratio scale of measurement. Nonparametric tests: measurement of individuals involve classification, involves measurement on the nominal or ordinal scale of measurement, data are usually frequencies of individuals in a particular category. Most commonly used test of association for nominal/ordinal data. Null hypothesis (no symbols): no preference. The population is divided equally among the categories: no difference from a known population. Observed cell frequency f0: the number of individuals from the sample who are classified into a particular category. Expected cell frequency fe: the frequency that is predicted from the null hypothesis and sample size fe = pn: p: proportion stated in the null hypothesis, n: sample size. Preferences for zombie walks are equally divided among each potential zombie walk venue. Reflects degree to which the observed differs from the expected. Degrees of freedom: goodness of fit. Df = (c-1) = 4-1 = 3, cv = 7. 81.
