BIOL361 Lecture Notes - Lecture 15: Continuity Correction, Categorical Variable, Statistical Population
24 Apr 2016
School
Department
Course
Professor
Document Summary
Categorical data = qualitative data that can be separated into diff categories that are distinguished by some non-numeric characteristic. = data at nominal or ordinal level of measurement. Counts of items or events eg; frequency of males and females in a sample. 3 approaches: chi-square (x2) for, log-likelihood ratio (g-test) in spss called likelihood ratio for multinomial, exact tests (fisher) when sample size is small. Gof tests = a general method to test a claim that the observed frequencies (=sample) in the different categories fit ( or h1 = deviation) a particular (hypothesized) distribution (in the population) Ho = observed freq distribution from sample fits a claimed distribution ( for statistical population) We know there is variability in data with repeated samples, we need a test to assess: Test calc a test statistic that measures the discrepancy between observed and expected frequencies. X2 = (o- e)2/e for the k categories d. f.
