SOC210 Lecture 18: Multivariate Measures of Association
Document Summary
Soc 210 lecture 18 multivariate measures of association. Partial correlation: allows us to examine how a bivariate relationship changes when control variables are introduced. Multiple regressions: investigates the separate effects of two independent variables on a dependent variables. Multiple correlations: examines the combined effects of more than one independent variable on a dependent variable. Interested in the relationship between x1 and y1, but also consider the role x2. A multivariate relationship between where the control variable, has no effect on the bivariate relationship between x1 and y. A multivariate relationship where the control variable, x2 is the cause of both the independent variable x1, and the dependent variable y. A multivariate relationship where x1 and x2 are liked through the control variable, Ry1. 2 = (ry1 ry2r12) / (1 ry2. Ry1. 2 = the partial correlation coefficient between x1 and y, controlling for x2. Ry1 = the zero-order correlation coefficient between x1 and y.