Section B Question 7 Flashcards

1
Q

What conclusions can you reach re the relationship between salary & number of years of graduation

A

A positive linear relationship exists between salary & number of years of graduation. The relationship is significant as P-value is less than1.00

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the predicted salary for a person in survey in 2012, who graduated from the program in 2008? Comment on how safe the prediction is.

A

The model or regression equation is
salaryi = 40.73 + 6.42 yearsi
With Years = 4 we have;
salaryi = 40.73 +6.42 * 4 = 66.41K

The prediction is good because of high R square (0.86) an low p value or the significant level of co efficient years.
However the prediction might not be reliable because there could be factors captured by E that explain salary but at the same time be correlated with years, eg experience which explains salary and at the same time is correlated with the number of years after graduation.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Find & Interpret the value of standard error of the estimate?

A

SYX=9.45 It is the standard deviation of the actual salary around the predicted salary.

Alternatively it is the standard deviation of the new disturbance terms.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What assumptions have been employed in order to obtain these regression results?

A

Linearity-The relationship between Y (salary) and X (years) is linear.
Independence of errors-E associated with a particular value of Y (salary) ins independent of E associated with any other value of Y (salary).
Normality of error-The errors at a particular value of x (years) follow a normal distribution.
Equal variance (homoscedasticity): The variance of the errors is constant for all values of x (years)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly