Please use your own data set assigned to you for the regression analysis, and attach

computer output. The main purpose of this problem set is to study the value of education

with your own regression model based on your data set. Here's a quote from one labor

economist's report,

"Higher education is the key to higher earnings, the data say: More than $60$ per cent of

full$-$time workers who made less than $20,000,000$ won had only a high$-$school diploma, while

more than $60$ per cent of those earning over $40,000,000$ had university degrees. For men in

all age groups taken together, university graduates were the only group to earn significantly

more than in $1990.$ "You can't say to everyone we should all have PhDs$,"$ said this economist.

"But for those with less education these are clearly depressing results." What's more, says

this economists, the numbers give value to the bachelor of arts degree, demonstrating that its

earning power, while less than that of other university degrees, still exceeds the benefits of a

technical diploma from a non$-$college, university institution."

To study these descriptive statistical results using a regression model, you need to set up

one econometric model. You can choose the dependent and independent variables for your

own, but they should be carefully chosen. For example, one may consider whether university

graduates males are earning more than females with the same education, or whether personal

income is usually proportional to years of educations, etc.

First, describe your data set briefly in a few sentences based on the results in the

problem #$5.$

Set up your regression model, and explain why you choose the model and what the

interesting points will be in your regression model.

Run the regression model, and give the results and interpretations. You may need to

carry out tests based on your model or to support your conjecture. Do not run the

tests which are not pertinent to your regression model, for example, you do not need to

test multicollinearity unless you suspect a collinearity problem among your independent

variables.