BU academic adviser wants to predict the average annual sala

BU academic adviser wants to predict the average annual salaries of graduating students. She estimates the following regression model: Salary= Experience + epsilon where M1, M2, and M3 are dummy variables representing three majors (Engineering, History, and Statistics respectively), while International Affairs is the reference major The variable Experience represents years of experience (since graduation). You may assume the model was a good fit to the data, and all assumptions of multiple linear regression were satisfied. Below is the statistical software output. Use the level of significance alpha = 0.05. 1. What was the sample size? 2. Test the global null hypothesis. Write out all steps of hypothesis testing. 3. If the global null hypothesis were rejected, test the individual parameters. Write out all steps of hypothesis testing. If the global null hypothesis were not rejected, take a break and then move to Problem 2. 4. Interpret the significant parameters of the estimated model.

Solution

Significant parameters of the estimated model:

F statistic is very high, higher than F_4,167 as seen from the tables. Hence the regression is explained by the explanator variables in a stasticially significant manner.

R² statistics shows that the explanatory variables explain approx 98.9% of the variation in salary.

P-Values of individual variables is less than significance level of 5% and hence, each independent variable is statistically significantly different from zero.

Experience has a positive impact on salary.