POL 3371 Lecture Notes - Lecture 14: Linear Regression, Interval Ratio, Statistical Significance

• School: University of Ottawa
• Department: Political Science
• Course: POL 3371
• Professor: Ivan Katchanovski

Document Summary

Widely used to test research hypotheses: can test more than one research hypothesis simultaneously keeping all other. Variables at interval-ratio level: in practice, can also be used for ordinal variables if they are treated as interval- ratio, can be used for nominal variables only when they are transformed into dummy variables. Bivariate linear regression: examines the relationship between two variables. Regression: can determine how much of the variation in the dependent variable is explained (predicted) by the independent variable, if the independent variable is a statistically significant predictor of the dependent variable, direction of relationship. Positive sign of regression coefficient: positive effect of the independent variable on the dependent variable. Negative sign of regression coefficient: negative effect of the independent variable on the dependent variable. Statistical significance of regression coefficient: statistically significant if in spss p(obtained) = < 0. 05 or 0. 01 or 0. 001, for small n: can be statistically significant if spss p(obtained) < 0. 1, statistically non-significant otherwise.