Answer to Question #210623 in Economics for Subham

Question

Below I have presented results for a regression between expenditure and education:

hhexp1=​Annual household expenditure per capita for household i (in USD)

hheduc=​Years of education for adults in household i

Here is the output:

3.1. These estimators are derived by applying the OLS Principle to the two-variable linear regression model. What is the OLS Principle? What is the underlying intuition? (2)

3.2. State in plain but precise language (involving numbers and units of measurement) what the estimate of the slope coefficient tells us about what happens on average to household expenditures when the adults in the household get an additional year of education. (Note that you need the precise definitions and units of measurement of each variable)

Answer 1

Ordinary least squares (OLS) is a statistical procedure for fairly accurate prediction of the behavior of dependent variables.

For example, you can understand how the turnover ("y" value) of a chain of stores will change with a change in the size of the sales area ("x" value).

The essence of OLS is to find the best approximation to reality from all linear functions. This can be done by searching for the function with the smallest deviation (more precisely, by the OLS process: finding the minimum sum of squares of deviations of y values from the obtained regression equation).