PO218 Lecture 8: BUILDING MODELS AND MULTIVARIATE APPROACHES
Document Summary
Building models: allow us to quickly identify important features of a relationship. Independent (x) variable is arrayed along the horizontal axis; dependent (y) variable is arrayed along the vertical axis. Each dot represents a case of a sample: the summarizing line is called a regression line. Spss: use correlation to see if certain variables stick together. Ces (canadian election study: lots of different questions/ideas because the concept we are trying to measure is not easily measured by one question, we use a variety of measures to measure a not easily observable concept. Conceptual consistency: recoding specific items so that low scores all point to one polarity of a concept and high scores all point to another polarity of a concept. X y (bivariate: anything more is multivariate. Elaborate: replication < 1/3 of the original, explanation (spurious, relationship between x and y is explained by z, all partials not significant, 2 types, antecedent z proceeds x and y temporal order .
