Explore BrainMass

Regression Line for salary data

See attached data files.

Engineering Salaries Analysis for an MS

A project requires us to develop a regression model to assist us in our salary/bonus prediction based on variables represented by analyzing various factors about each engineering school. For example:

(1) if a student wants to go to a specific engineering school and finances his/her education through student loans, is it worth doing it? or

(2) looking at the employment rate within 6 month after his/her graduation is it worth doing it?

The file salary of new engineering students [Salary Data XLS file] contains a recent survey about new engineering graduates from the top 40 engineering schools [School Data] in the U.S.

(a) State statistical objective(s) for the project.

(b) Perform EDA including numerical descriptive measures.

(c) Construct scatter diagrams for pairs of variables. Do any of these appear to have some association?

(d) From (b) and (c), does any simple linear model appear to hold? Run some testing to substantiate your findings.

(e) Does any multiple regression model appear to hold? Run some testing including LINE analysis to substantiate why or why not. If so, is there more than one variable that may be used as a dependent variable?

(f) Is the regression significant? Report the results of the appropriate test, and interpret its meaning.

(g) Suppose now that you want to develop a regression model based on your choice of dependent variables against various independent variables (of your choice), do the data on region play any role in your model? Did you have to modify the region data in such a way that the location of each school makes a significant contribution to your model? What happens when you include the data on the type of schools (public vs. private)?

(h) Do you find any interaction term in the model that makes a significant contribution to the model?

(i) Summarize and comment on your results.


Solution Summary

The present study tries to model the salary / bonus of various engineering school based on some input variables like Employment Rate within 6 months, Annual cost, Average Indebtedness, ranking of the school and Average Indebtedness. Scatter diagram, simple linear regression and multiple regressions are used for this purpose. Based on the regression coefficients , appropriate decision can be taken to check whether the independent variables have significant influence in determining the salary / bonus.