SOC 3142 Lecture Notes - Lecture 5: Multimodal Distribution, Central Tendency
Document Summary
Recoding variables: save backup copy of dataset (recode trasnforms original data, recode into another variable, not over the existing variable, recode using syntax, to better keep track of changes. Why recode at all: to simplify a variable (see example of alcohol consumption in sweet and grace-martin p. 68, to exlcude outliers (e. g. example of texas pg. 70: to modify the ranking of values of a variable. Data reduction: the process of using graphs, figures, or numbers to represent data. Proportions (p=f/n) and percentages (%= (f/n) x 100) are the most common single numbers used to represent data. Both are usually easier to read and compare than frequencies. When dealing with small frequencies (fewer than 20), proportions and percentages are not recommended. Ratios are used to express the relative frequency of one category per unit of another category (cid:523)the ratio of women to men in this class is (cid:524)