Professional Documents
Culture Documents
1. An agency analyst seeks to utilize multiple regression analysis for a study on health
care. She wants to specifically determine if the outcome of a policy is explained by other
factors (such as staffing, budgets, technical resources). In this research design, policy
outcome is the ______variable, while the other factors represent the ________variables.
*A. Dependent, independent
B. Independent, dependent
C. Cross-tab, functional
D. Intermediate, Advanced
3. Professor Jones wants to adjust for the number of variables in her multiple regression
equation so as to increase the predictive power of her newly added variable. To do so,
she will use the:
A. Multiplied R-squared
B. Reactive R-squared
*C. Adjusted R-squared
D. Enhanced R-squared
4. In measuring the outputs of her program, an analyst found that one of the independent
variables she is using is a perfectly correlated to another, forcing her to use only one of
the two in her multiple regression analysis. This is an example of:
*A. Perfect multicollinearity
B. Perfect magnitude
C. Extreme correlation
D. Flawless relatedness
7. In order to distinguish between a set of dummy variables that have multiple categories
so that the regression coefficients produce contrasts, the research may use what is
referred to as:
A. Seamless coding
B. Linear coding
*C. Effect coding
D. Relative coding
9. An ____ variable is the product of two other variables, such as the product of the
variables, hypertension and diabetes.
*A. Interaction
B. Interrelated
C. Interface
D. Encompassing
10. The _______ is the most easily interpretable method for doing a multiple regression
with a dummy dependent variable.
A. Curvilinear regression fit
*B. Linear probability model
C. Sequential fit model
D. Allocated effects model
11. Researchers use ___ equations where the log odds of an event are predicted, and
_______ equations where the probability is predicted.
A. Surface, interface
B. Decimal, cyclical
C. Focal, prospect
*D. Logit, probit
12. ____ estimates the pattern of relationships between variables in a presumed causal
structure.
a. Exploratory factor analysis
b. Principal components analysis
*c. Path analysis
d. Confirmatory factor analysis
13. In his research on college students, Professor Gupta is interested in grouping several
of his variables on study habits into a smaller set of underlying factors which measure
student success in college. He is using the method of:
*A. Factor analysis
B. Reduction analysis
C. Diminishment analysis
D. None of the above
14. ____ is a method that uses data from time series in the past to predict future values of
the dependent variable.
15. ________ combines both factor and path analysis into one estimation method.
A. Effective attribute modeling
*B. Structural equation modeling
C. Systematic formula modeling
D. Core equation modeling
16.____ often consist of repeated measures of the same variable from the same
individuals over time.
a. Standard deviations
b. Standard errors
c. Coefficients
*d. Standardized coefficients
20. A researcher wanted to know whether contracting out government services improves
municipal efficiency. The political party in charge is unrelated to its government’s
efficiency but strongly correlated with contracting out. If political party were added as a
control variable in this model, the standard error would most likely . . .
*A. Increase
B. Decrease
C. Remain the same
D. Impossible to say
The table on the next page is taken from “Maternal Employment and Teenage
Childbearing” by Leonard Lopoo in the Journal of Policy Analysis and Management
24(1) 2004. We will only consider the OLS regression results. The unit of analysis is the
mother. The variable “daughter has a birth at age 17 or 18” is a dummy variable with the
meaning implied. “Mother’s education” refers to the years of education that a mother has.
“Mother’s annual work hours” is the hours worked during the year. Use the table to
answer the following questions:
a. The percent change in income predicted for a one percent increase in education.
*b. The percent change in income predicted for a one-year increase in education.
c. The dollar change in income predicted for a one percent increase in education.
d. The dollar change in income predicted for a one-year increase in education.
e. None of the above
24. In a survey, an individual’s health insurance status is reported as being in one of the
following categories: Medicare, Medicaid, Employer provided insurance, Individual
private insurance, Other insurance or Uninsured. If health insurance status is to be used as
an independent variable in a regression, the following are acceptable combinations of
dummy variables:
Please read the descriptions of the following hypothetical research findings, look at the
regression tables, and answer the questions.
A study focused on factors influencing state spending on social services per capita (in
dollars per capita)
b Std. error
Constant 500 3.71
Per capita 200 .256
income (tens of
thousands)
Democratic 500 .024
governor
R-squared .35
25. Per capita income and governor’s political party explain what percent of the variation
in per capita spending on social services?
a. 35%
27. Interpret in words understandable to a policy audience the coefficient of per capita
income.
A. For each $10,000 more in per capita income that a state has, its per capita spending on
social services is predicted to rise by $200, if the party of the Governor stays constant.
28. What would per capita spending be in a state with a governor who is not a Democrat
with a per capita income of $40,000?
a. $1300
A survey asked people to rate their feelings toward paying taxes on a scale of 0 to 100,
with 0 being cold or unfavorable and 100 being very warm or favorable. In this model,
income and age are quantitative variables, and homeowner is a dummy variable with 1
representing owning a home.
B
Constant 90
Income (tens of thousands of -0.5
dollars)
Homeowner -4
Income x homeowner -2
Age -0.5
29. For someone who does not own a home, what value does the interaction term take
on?
a. 0
30. What would the expected rating be for someone who is 40 years old who owns a
home and earns $60,000 per year?
a. 51
31. How are feelings towards paying taxes predicted to change when age increases by
one year, with income and homeownership status staying constant?
a. They are predicted to decrease by 0.5.
33. Women who had Medicaid insurance were also included in the study. Why was this
category not included as a dummy variable?
a. Answers will vary.
35. With 95% confidence, what is the true odds ratio for getting a mammogram among
women with Medicare insurance?
a. Answers will vary.
36. What are the odds of getting a mammogram among women with a family history of
breast cancer?
a. Answers will vary.