SOC 3142 Lecture Notes - Lecture 4: Univariate, Frequency Distribution, Univariate Analysis
Document Summary
Recode: the point of recoding is to make a complex variable simple, for example recoding. The purpose of all quantitative research, in all disciplines is to establish relationships between variables. Only experimental research allows the research to establish causal relationships between variables, by manipulating independent variables. Most social research is correlation and probabilistic, not causal and deterministic. Researchers are only able to describe correlations and to point to possible causal relations by eliminating other plausible causal variables. But before establishing and describing relationships between variables, researchers must understand the basic properties of the available data. A number of methods are available to this effect. Although researchers usually know what questions they seek to answer by analyzing the data, descriptive methods may help refine these questions. Need to recode or otherwise manipulate certain variables (e. g. if too many categories) Need to drop or add variables (if possible).