Provide your answers below to each question. Do not delete the questions. Please do not use red font. Blue or green is OK.
Part I (35 POINTS)
The General Social Survey research department was tasked to determine the number of hours per day government employees in the USA spent on emailing. The department reported the data for a sample of 1765 government employees from GSS 2018 data on age (number of years), sex of respondent (1=males, 2=females), total family income (in constant dollars), and hours worked last week. Refer to the SPSS output below. Conduct the full regression analysis.
Regression
Model Summary
Model
R
R Square
Adjusted R Square
Std. E
or of the Estimate
1
.312a
.097
.092
12.165
a. Predictors: (Constant), Respondents sex, Age of respondent, Number of hours worked last week, Respondent income in constant dollars
ANOVAa
Model
Sum of Squares
df
Mean Square
F
Sig.
1
Regression
XXXXXXXXXX
4
XXXXXXXXXX
16.495
.000
Residual
XXXXXXXXXX
611
147.994
Total
XXXXXXXXXX
615
a. Dependent Variable: Email hours per week
b. Predictors: (Constant), Respondents sex, Age of respondent, Number of hours worked last week, Respondent income in constant dollars
Coefficientsa
Model
Unstandardized Coefficients
Standardized Coefficients
t
Sig.
B
Std. E
o
Beta
1
(Constant)
-2.633
3.044
-.865
.387
Age of respondent
-.017
.036
-.019
-.483
.629
Number of hours worked last week
.090
.036
.102
2.464
.014
Respondent income in constant dollars
8.863E-5
.000
.278
6.659
.000
Respondents sex
3.523
1.025
.138
3.438
.001
a. Dependent Variable: Email hours per week
Hint: For the coefficient that ends in E-5, move the point 5 decimal places to the left
1. Which variables are independent variables?
2. Which variable is the dependent variable?
3. What does coefficient of determination (R2) tell you?
4. What is the intercept value? What can it indicate here?
5. Is the model overall significant?
6. Determine the multiple regression equation:
7. Discuss significant regression coefficients
8. What variable has the strongest effect on the number of hours US government employees spent emailing?
9. What is the estimated number of hours emailing per day for a woman who works 40 hours per week and who earns $50.000 per year?
Part II (35 POINTS)
Give the full logistic regression analysis of the SPSS 2014 output provided below. The coding for the dependent variable is presented in the SPSS output. Interpret and include in your regression equation only significant coefficients (use alpha= 0.05).
The question that the respondents were asked is: “Would you favor or
oppose the teaching of sex education in public schools?” (Favor= 1; Oppose=0)
Coding for variables:
polviews:
1. Extremely liberal
2. Liberal
3. Slightly Liberal
4. Moderate
5. Slightly Conservative
6. Conservative
7. Extremely Conservative
pillok: Birth Control to Teenagers 14-16
XXXXXXXXXXQuestion asked: Do you strongly agree, agree, disagree, or strongly disagree that it is acceptable to make birth control devices available to teenagers, age 14-16?
1. Strongly Agree
2. Agree
3. Disagree
4. Strongly disagree
educ: XXXXXXXXXXHighest year of school completed
Logistic Regression
Case Processing Summary
Unweighted Casesa
N
Percent
Selected Cases
Included in Analysis
1579
62.2
Missing Cases
959
37.8
Total
2538
100.0
Unselected Cases
0
.0
Total
2538
100.0
a. If weight is in effect, see classification table for the total number of cases.
Dependent Variable Encoding
Original Value
Internal Value
Oppose
0
Favo
1
Categorical Variables Codings
Frequency
Parameter coding
(1)
(2)
(3)
BIRTH CONTROL TO TEENAGERS 14-16
STRONGLY AGREE
403
1.000
.000
.000
AGREE
518
.000
1.000
.000
DISAGREE
389
.000
.000
1.000
STRONGLY DISAGREE
269
.000
.000
.000
Block 0: Beginning Block
Classification Tablea,
Observed
Predicted
SEX EDUCATION IN PUBLIC SCHOOLS RECODED
Percentage Co
ect
Oppose
Favo
Step 0
SEX EDUCATION IN PUBLIC SCHOOLS RECODED
Oppose
0
141
.0
Favo
0
1438
100.0
Overall Percentage
91.1
a. Constant is included in the model.
b. The cut value is .500
Block 1: Method = Ente
Omnibus Tests of Model Coefficients
Chi-square
df
Sig.
Step 1
Step
161.537
6
.000
Block
161.537
6
.000
Model
161.537
6
.000
Model Summary
Step
-2 Log likelihood
Cox & Snell R Square
Nagelkerke R Square
1
788.732a
.097
.215
a. Estimation terminated at iteration number 7 because parameter estimates changed by less than .001.
Classification Tablea
Observed
Predicted
SEX EDUCATION IN PUBLIC SCHOOLS RECODED
Percentage Co
ect
Oppose
Favo
Step 1
SEX EDUCATION IN PUBLIC SCHOOLS RECODED
Oppose
7
134
5.0
Favo
9
1429
99.4
Overall Percentage
90.9
a. The cut value is .500
Variables in the Equation
B
S.E.
Wald
df
Sig.
Exp(B)
Step 1a
AGE OF RESPONDENT
-.020
.006
13.774
1
.000
.980
HIGHEST YEAR OF SCHOOL COMPLETED
.100
.031
10.271
1
.001
1.105
THINK OF SELF AS LIBERAL OR CONSERVATIVE
-.447
.074
36.855
1
.000
.639
BIRTH CONTROL TO TEENAGERS 14-16
45.412
3
.000
BIRTH CONTROL TO TEENAGERS XXXXXXXXXX)
1.940
.364
28.347
1
.000
6.957
BIRTH CONTROL TO TEENAGERS XXXXXXXXXX)
1.402
.261
28.755
1
.000
4.064
BIRTH CONTROL TO TEENAGERS XXXXXXXXXX)
.800
.227
12.437
1
.000
2.226
Constant
3.177
.646
24.168
1
.000
23.966
a. Variable(s) entered on step 1: AGE OF RESPONDENT, HIGHEST YEAR OF SCHOOL COMPLETED, THINK OF SELF AS LIBERAL OR CONSERVATIVE, BIRTH CONTROL TO TEENAGERS 14-16.
1. What variables are independent variables?
2. What variable is the dependent variable?
3. Determine the logistic regression equation
4. Determine which independent variables are statistically significant and interpret the regression (B) coefficients
Variable Name
Level of significance
5. Interpret the Exp (B) coefficients
Variable Name
Level of significance
6. Provide overall conclusion
Part III (30 POINTS)
Review the article: Anna Ya Ni and Stuart Bretschneider. “The Decision to Contract Out: A study of Contracting for E-Government Services in State Governments.” Public Administration Review 64, 3: XXXXXXXXXXIt is available on Module 9 in the Readings folder. Write you review keeping the following questions in mind:
1. What is the policy analysis question?
2. What is the unit of analysis (identify a case in this study)?
3. What method of analysis is used in this research?
4. Draw the model using boxes and a
ows diagram.
5. Indicate the dependent and independent variables used in this model with their level of measurement and unit of measurement; fill in the table provided below. You can add or delete rows from the table.
Unit of Analysis:
Variable Name
Level of Measurement
Unit of measurement/coding
Independent/Dependent Variable
6. Explain the logistic regression results. Use the table format and the information you have inputted in question 5.
Hypothesis Give a number to hypothesis. (H1 etc.) It is conventional way to indicate hypotheses.
Logistic Regression results
7. How type of service influences the outsourcing decision?
Extra credit (10 points)
Using your regression equation from part II, calculate the probability of favoring teaching of sex education in public schools for a 40-year-old person with some college education (15 years), liberal in political views. That person strongly agrees that birth control pills should be available to teenagers.
6
Th e Decision to Contract Out: A Study of Contracting for E-Government Services in State Governments
The Decision to Contract Out 531
Government contracting, especially for information
technology products and services, has accelerated in recent
years in the United States. Drawing on the insights of
privatization studies, the authors examine the economic
and political rationales underpinning government
decisions to contract out e-government services. Th is
article tests the extent to which economic and political
ationality infl uence governments’ contracting decisions
using data from multiple sources: a survey conducted by
National Association of State Chief Information Offi cers,
a survey by the National Association of State Procure-
ment Offi cers, the Council of State Legislatures, and
macro-level state data from the U.S. Census Bureau.
Important factors aff ecting the state-level contracting
decision are population size, market size, the competitive-
ness of the bidding process, the professional management
of contracts, the partisan composition of legislatures, and
political competition. Political rationales appear to play a
major role in state contracting decisions. Some arguments
associated with markets and economic rationality are
clearly politically motivated.
During the last two decades, as privatization has gained political cu
ency, contracts with private sector organizations have increased
dramatically in bipartisan governments. Privatization
advocates believe that contracting out is an eff ective
tool for government to reduce costs, increase effi -
ciency, improve services, and encourage innovations
( Gore 1993; Kettl 1993; Osborne and Gaebler 1992;
Salamon 1989; Savas 1987 ).
Along with the growth of contracting out, interest in
electronic government (e-government) has also grown
for many of the same reasons. A substantial number
of e-government initiatives have made heavy use of
contracting out ( Gant, Gant, and Johnson 2002 ).
Th e popularity of contracting out services such as data
processing, Web site hosting, training, and project
management has spread across all levels of government.
It is believed that contracting helps governments over-
come fi nancial diffi culties in accessing the esoteric
expertise and professional management skills of private
fi rms to develop e-government applications ( Brown
and Brudney 1998; Chen and Pe
y 2002 ).
Growing interest in privatization has fueled many