POLSCI 391 Chapter Notes - Chapter 2: Environmental Noise, Linear Regression, Data Mining
Document Summary
Models: posit functional relationships between variables of interest. Linear regression: most common model with data. Y = mx + b slope describes the size of the effect of the independent variable on the dependent variable. Fitting a linear mode consists of finding the line that"s closest to the data. Differences in the world that have been left out of the model. Difference between what the model predicts and the actual value of the dependent variable. Error can result from measurement error or environmental noise. Error based interpretation of the data can assume underlying process or mechanism, a true cause that generates the data. When we gather data, the reason that it does not fit the law is because of errors in measurement laws are true. Heterogeneity interpretation, the mistake captures variation within the population, which is very important because corresponds to deviation between entity"s value and the model"s prediction. By analyzing the variation, social scientists can improve their models.