Deck 12: Observational Studies With Control Variables

Question 1

Because government corruption is difficult to measure directly, a student used citizen perceptions of corruption as reported in a survey as a variable, even though it is not an ideal measure .Perceptions of corruption in this case is an example of

A)A proxy measure
B)An omitted variable
C)A control variable
D)An unmeasured variable

Answer 1

A

Question 2

Which is NOT an example of an observational study?

A)Researchers assign participants to teach either older or younger people how to use a computer to test whether participants appear more frustrated teaching older people.
B)A manager uses sales records to compare the performance of different employees.
C)A professor studies the relationship between democracy and per capita income using country-level data from the World Bank.
D)A political scientist uses data from a random sample survey of 2000 US adults to examine the effect of media exposure on people's trust of government.

Answer 2

A

Question 3

An agency analyst is studying factors which motivate employees to work, but fails to consider job perks in her equation .In this case, job perks are a ___ variable since they are in the background, and may potentially bias the estimated causal relationship between her independent and dependent variables.

A)Lurking
B)Covert
C)Concealed
D)Clandestine

Answer 3

A

Question 4

A researcher is interested in conditioning the data, where she will break it into groups based upon student GPA's .In this situation, the researcher is going to _____ the datA.

A)Allocate
B)Intensify
C)Partition
D)Stratify

Question 5

To investigate the impact of population density on crime, a researcher compared average crime rates in low and high density neighborhoods .To control for poverty, which is related to both density and crime, he further stratified the data into low and high poverty neighborhoods .Instead of using this method of stratification, multiple regression analysis is often preferable because

A)It can accommodate more than one or two control variables
B)It is not limited to categorical variables
C)A fuller range of values of both the independent and control variables can be taken into account
D)All of the above

Question 6

High school students from low-income families sometimes work after school to bring in extra income .Because of the hours they spend working, they have less time to do homework, which can lead to poorer grades .A researcher explores the effect of family income on grades .What type of variable is working part time in this study?

A)Common cause
B)Intervening variable
C)Control variable
D)None of the above

Question 7

Program evaluators seek to estimate the causal effect of a program (X) on an outcome (Y) using observational datA .The evaluators attempt to adjust their estimate for common causes (Z) that influence both X and Y .In this evaluation, the common causes (Z) are best described as

A)Intervening variables
B)Independent variables
C)Control variables
D)All of the above

Question 8

To collect data as part of an observational study, a researcher may use:

A)Telephone surveys
B)Administrative record data
C)Secondary data from large government surveys
D)All of the above

Question 9

A study found higher rates of crime in cities with an international airport, compared to cities without an international airport .But the study failed to take account of the size of the city (population), and bigger cities tend both to have more crime and to have international airports .The findings of this study suffer from

A)Measurement bias
B)Omitted variable bias
C)Sample selection bias
D)All of the above

Question 10

A(n) __ is a variable that causally occurs before both X and Y, and is viewed as a potential common cause.

A)Cofounder
B)Replacement
C)Distortion
D)Antecedent

Question 11

A study found that eating junk food is related to getting cancer .Which of these statements is true?

A)Adding control variables could increase the relationship between junk food and cancer.
B)Adding control variables could decrease the relationship between junk food and cancer.
C)Adding control variables could eliminate the relationship between junk food and cancer.
D)All of the above