Question

Explain how, given any attempt to use linear regression to explain economic phenomenon, you can be sure that your model is indeed measuring what you set out to measure and not simply reflecting spurious correlation between variables.

Answer 1

Linear Regression develops a model to predict a dependent variable by manipulating independent variables. It is of the form Y = A + b1X1 + b2 x2 + b3x3 + ..... where Y is the dependent variable and the Xs are the independent variable.

When the Regression is run the R²   will indicate the extent of variation in the dependent variable that is being explained by the Regression model.

Each of the coefficients of the Independent variables can be tested for significance ( significantly different from zero). When a independent variable is significant, it indicates a causality.

Linear Regression tackles the problem of spurious correlation by looking at auto correlation between the variables. The presence of positive or negative auto correlation indicates that the data is not correct in the sense that there is correlation between the predictor variables. This is measured by the Durbin Watson coefficient.