Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

The provost at a major university would like to develop a model to examine the r

ID: 3315020 • Letter: T

Question

The provost at a major university would like to develop a model to examine the relationship between thesalaries of full-time associate professors at the institution (the dependent variable) and the following independent variables (explanatory variables): an associate professor’s performance rating on a scale of 1 to 20, his or her gender, student approval rating on a scale of 0 to 100%, the associate professors age, years of teaching experience, and the college (arts, sciences, or business) the associate professor teaches. The provost has collected these data from a random sample of associate professors, which can be found in the Excel fileFaculty Salaries (the data may be found in the content area).

Simple Linear Regression

Run a simple linear regression of faculty salaries as a function of years of experience.

Is the regression coefficient on years of experience significant at the 2.5% level? Explain.

Provide a forecast of the mean faculty salary given 25 years of experience.

Construct a 95% confidence prediction interval for the predicted salary in part c.

Report and interpret the R2 for this regression equation.

Multiple Linear Regression

Construct and run a regression model using all the independent/explanatory variables.

Check for the presence of multi-collinearity between the independent/explanatory variables. If multi-collinearity is suspected explain why and then eliminate that variable(s) which you determine to be multi-collinear.

Run a regression of this new linear model after deleting the collinear variables.

Using the results from this regression model interpret each regression coefficient.

Test the significance of each of the estimated coefficients using a level of significance, , of 10% (ie. .10).

Interpret the coefficient of determination, R2.

Test the significance of the R2 using a level of significance, , of 5% (ie. .05).

Remove those explanatory variables from this model that were not significant at the 10% level.

Run an additional regression for this linear regression model after removing the insignificant coefficients.

Using the p-value approach again test the significance of the estimated coefficients using a level of significance, , at the .05 level.

Perform a Regression with Faculty Salary as the dependent variable and Years of Teaching Experience, Performance Score, Student Approval Rating, the Gender Dummy Variable and the College Dummy Variable as the explanatory variables. Then answer p and q.

Predict the salary of a female associate professor in the business college who is 41 years old with 11 years of teaching experience, a performance score of 17, and a student approval rating of 88.3%.

Predict this salary of a male associate professor in the business college who is 41 years old with 11 years of teaching experience, a performance score of 17, and a student approval rating of 88.3%.

Data:

(Salary ($)) (Performance) (College) (Age) (Student Approval) (Experience Gender)

110714 14 Arts & Science 45 89.2 19 Male

103756 13 Business 53 81.8 22 Female

105764 16 Business 50 85.7 19 Female

106118 18 Business 48 82.8 22 Male

100266 16 Business 36 69.7 11 Female

60705 14 Arts & Science 55 89.2 29 Female

104843 16 Business 45 77.7 23 Female

100705 13 Business 54 85.6 23 Male

78195 15 Arts & Science 44 79.8 14 Female

92867 14 Business 38 79.6 10 Female

76675 17 Arts & Science 48 85.6 20 Male

73309 20 Business 31 78.9 6 Male

114377 14 Business 49 86.6 23 Male

73952 14 Business 33 97.2 6 Female

70750 14 Business 33 93.5 8 Female

101661 19 Business 48 75.3 20 Female

113276 18 Business 49 94.5 23 Male

76831 14 Business 40 78.4 14 Female

97251 17 Business 41 83.6 12 Female

105763 18 Business 34 99.4 6 Female

88489 17 Business 42 78.1 18 Female

95185 16 Business 42 84 15 Male

88430 15 Business 34 83.2 8 Male

66569 15 Arts & Science 55 76.6 29 Female

68104 16 Arts & Science 34 84.7 6 Female

60023 15 Business 39 82.5 10 Female

82395 15 Arts & Science 59 82.3 26 Female

107532 20 Business 56 98.9 24 Female

90208 15 Business 42 88.2 18 Female

71750 12 Arts & Science 46 89.5 22 Female

80801 17 Business 45 71.3 15 Male

73038 14 Arts & Science 42 96.1 13 Male

58753 14 Arts & Science 38 73.4 14 Male

83084 14 Business 49 88.6 20 Male

58163 14 Arts & Science 36 76 11 Female

82375 16 Business 36 83.5 12 Male

78510 18 Business 34 84.1 9 Male

103295 13 Business 48 90.9 24 Male

102911 18 Arts & Science 44 91.7 20 Male

74561 19 Arts & Science 49 95.1 18 Male

75975 13 Arts & Science 44 75.3 20 Male

89213 19 Arts & Science 40 89.6 13 Male

78566 13 Business 38 97 10 Female

70206 14 Arts & Science 35 88.5 7 Female

63864 16 Arts & Science 59 75.2 26 Male

74535 13 Arts & Science 43 77.7 14 Male

99667 17 Business 40 97.9 11 Female

65824 15 Arts & Science 50 83.8 19 Male

89298 15 Business 43 90.5 16 Female

72244 13 Business 32 86 5 Male

81216 15 Business 48 82.2 20 Female

65317 17 Arts & Science 61 81.4 29 Male

55357 15 Arts & Science 35 87.5 11 Female

65681 12 Arts & Science 50 79.1 19 Male

87191 15 Business 32 91.4 5 Male

108409 20 Business 41 88.9 14 Female

63459 17 Arts & Science 55 82 25 Female

62931 12 Arts & Science 53 77.8 22 Male

111807 15 Business 45 90.8 17 Male

95754 15 Arts & Science 48 86 20 Male

74162 16 Arts & Science 40 87.8 14 Female

68311 14 Arts & Science 36 80.9 10 Female

111610 17 Business 48 78.9 20 Female

72434 14 Arts & Science 41 86.9 17 Female

75522 13 Arts & Science 37 93.8 9 Female

125407 17 Business 58 88.6 30 Male

58986 13 Arts & Science 40 81.3 15 Male

69668 15 Arts & Science 41 75.8 14 Female

81318 16 Business 45 81.1 17 Male

94079 17 Business 38 95.7 12 Female

80252 19 Arts & Science 56 97.5 24 Male

120263 18 Business 48 78.8 18 Male

74511 19 Arts & Science 40 74.2 17 Female

68715 17 Arts & Science 36 85.8 10 Female

66271 15 Arts & Science 48 82.2 20 Male

87741 20 Arts & Science 47 71.3 23 Male

94076 18 Business 58 78.7 26 Female

58908 13 Arts & Science 33 83.5 8 Female

79548 13 Business 34 91.9 6 Male

94613 12 Business 44 82.7 16 Male

56345 12 Arts & Science 30 88 5 Female

74608 16 Arts & Science 43 80.3 17 Female

60040 13 Arts & Science 36 87.3 10 Female

70647 15 Business 38 92.9 14 Female

76272 12 Business 30 85.3 4 Male

Explanation / Answer

LINEAR REGRESSION

> model=lm(Salary~Experience)
> summary(model)

Call:
lm(formula = Salary ~ Experience)

Residuals:
Min 1Q Median 3Q Max
-31677 -10308 -2373 14011 35849

Coefficients:
Estimate Std. Error t value Pr(>|t|)   
(Intercept) 71374.9 4778.9 14.935 <2e-16 ***
Experience 724.4 276.2 2.623 0.0104 *  
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 16710 on 83 degrees of freedom
Multiple R-squared: 0.07655, Adjusted R-squared: 0.06542
F-statistic: 6.88 on 1 and 83 DF, p-value: 0.01037

Yes, the slope is significant at 2.5% level since the p-value < 0.025.

Predicted salary = 71374.9 + 724.4*25 = $89,484.88

> newdata=data.frame(Experience = 25)

> predict(model,newdata,interval="predict")
fit lwr upr
1 89484.88 55692.33 123277.4 #95%Prediction interval

R2 = 0.07655 means that 7.655% of the total variation is explained by the fitted regression model.