Question 1

A natural experiment occurs when:

Accepted Answer

A)  The values of an independent variable have been determined by an endogenous process. 
B)  The values of an independent variable have been determined by an exogenous process. 
C)  The values of the dependent variable have been determined by an endogenous process. 
D)  The values of the dependent variable have been determined by an exogenous process. 
A)  The values of an independent variable have been determined by an endogenous process. 
B)  The values of an independent variable have been determined by an exogenous process. 
C)  The values of the dependent variable have been determined by an endogenous process. 
D)  The values of the dependent variable have been determined by an exogenous process. B

Question 2

Suppose we run an experiment and measure several possible independent variables after treatment. Which of the following best characterizes why this is a problem?

Accepted Answer

A)  Independent variables measured after treatment may be affected by the treatment and may therefore soak up some of the treatment effect. 
B)  Including post-treatment variables will induce multicollinearity. 
C)  Post-treatment variables will not satisfy the inclusion condition. 
D)  Including post-treatment variables will cause imbalance in the treated and control groups. 
A)  Independent variables measured after treatment may be affected by the treatment and may therefore soak up some of the treatment effect. 
B)  Including post-treatment variables will induce multicollinearity. 
C)  Post-treatment variables will not satisfy the inclusion condition. 
D)  Including post-treatment variables will cause imbalance in the treated and control groups. A

Question 3

Which of the following is a reason on as to why non-compliance can be a problem?

Accepted Answer

A)  Non-compliance can lead to a correlation between treatment delivered and the error term. 
B)  Non-compliance can lead to a correlation between treatment and other independent variables. 
C)  Even completely random causes biased estimates of the treatment effects. 
D)  Non-compliance is not a problem if compliance is not random. 
A)  Non-compliance can lead to a correlation between treatment delivered and the error term. 
B)  Non-compliance can lead to a correlation between treatment and other independent variables. 
C)  Even completely random causes biased estimates of the treatment effects. 
D)  Non-compliance is not a problem if compliance is not random. A

Question 4

In order to check for balance we assess whether the dependent variable differs for the treatment and control groups.

Accepted Answer

A) True 
 B)False

Question 5

Which of the following explains why a 2SLS model can be used to analyze randomized experiments with imperfect compliance?

Accepted Answer

A)  The random assignment (Z) is uncorrelated with the independent variable. 
B)  The random assignment variable (Z) is correlated with the error term in the outcome equation. 
C)  The random assignment (Z) is correlated with the independent variable but not with the error term in the outcome equation. 
D)  The random assignment (Z) causes multicollinearity. 
A)  The random assignment (Z) is uncorrelated with the independent variable. 
B)  The random assignment variable (Z) is correlated with the error term in the outcome equation. 
C)  The random assignment (Z) is correlated with the independent variable but not with the error term in the outcome equation. 
D)  The random assignment (Z) causes multicollinearity.

Question 6

Attrition generally will not lead to bias because OLS simply deletes these observations from the sample.

Accepted Answer

A) True 
 B)False

Question 7

When faced with issues of non-compliance, the coefficient on an intention-to-treat treatment variable will be

Accepted Answer

A)  Larger than if there was perfect compliance 
B)  Smaller than if there was perfect compliance 
C)  The same as if there was no non-compliance 
D)  Potentially larger in magnitude than the actual treatment effect. 
A)  Larger than if there was perfect compliance 
B)  Smaller than if there was perfect compliance 
C)  The same as if there was no non-compliance 
D)  Potentially larger in magnitude than the actual treatment effect.

Question 8

Which of the following is a method for dealing with attrition in a randomized experiment?

Accepted Answer

A)  Control for the variables associated with attrition 
B)  A Heckmann selection model. 
C)  Use a trimmed data set 
D)  All of the above 
A)  Control for the variables associated with attrition 
B)  A Heckmann selection model. 
C)  Use a trimmed data set 
D)  All of the above

Question 9

Which of the following is not one of the problems that leads to endogeneity in randomized experiments.

Accepted Answer

A)  Random attrition 
B)  Lack of balance for covariates 
C)  Non-compliance 
D)  Natural experiments 
A)  Random attrition 
B)  Lack of balance for covariates 
C)  Non-compliance 
D)  Natural experiments

Question 10

Which of the following must be undertaken when the treatment and control groups differ with regard to some independent variable, X1.

Accepted Answer

A)  Run an intention-to-treat analysis 
B)  Run a 2SLS model where the instrument is the random assignment 
C)  Use multivariate OLS and control for X1. 
D)  Restart the experiment and pick a new randomized sample. 
A)  Run an intention-to-treat analysis 
B)  Run a 2SLS model where the instrument is the random assignment 
C)  Use multivariate OLS and control for X1. 
D)  Restart the experiment and pick a new randomized sample.

Question 11

A compliance problem can lead to endogeneity because non-compliance is generally not-random.

Accepted Answer

A) True 
 B)False

Question 12

The difference in means of the treated and control groups is not a good measure of the treatment effect when randomization fails to create balanced treatment and control groups.

Accepted Answer

A) True 
 B)False

Question 13

When using the intention-to-treat method to deal with non-compliance:

Accepted Answer

A)  The higher the compliance rate, the more likely the ITT estimate overstates the true coefficient. 
B)  The higher the compliance rate, the further away (lower) the coefficient estimate is from the true coefficient. 
C)  The lower the compliance rate, the closer the coefficient estimate is to the true coefficient. 
D)  The lower the compliance rate, the further away (lower) the coefficient estimate is from to the true coefficient. 
A)  The higher the compliance rate, the more likely the ITT estimate overstates the true coefficient. 
B)  The higher the compliance rate, the further away (lower) the coefficient estimate is from the true coefficient. 
C)  The lower the compliance rate, the closer the coefficient estimate is to the true coefficient. 
D)  The lower the compliance rate, the further away (lower) the coefficient estimate is from to the true coefficient.

Question 14

One way to check for balance between the treatment and control groups is to look for a statistically significant coefficient on the treatment dummy variable in a model in which variable X1 is the dependent variable.

Accepted Answer

A) True 
 B)False

A natural experiment occurs when:

Suppose we run an experiment and measure several possible independent variables after treatment. Which of the following best characterizes why this is a problem?

Which of the following is a reason on as to why non-compliance can be a problem?

In order to check for balance we assess whether the dependent variable differs for the treatment and control groups.

Which of the following explains why a 2SLS model can be used to analyze randomized experiments with imperfect compliance?

Attrition generally will not lead to bias because OLS simply deletes these observations from the sample.

When faced with issues of non-compliance, the coefficient on an intention-to-treat treatment variable will be

Which of the following is a method for dealing with attrition in a randomized experiment?

Which of the following is not one of the problems that leads to endogeneity in randomized experiments.

Which of the following must be undertaken when the treatment and control groups differ with regard to some independent variable, X1.

A compliance problem can lead to endogeneity because non-compliance is generally not-random.

The difference in means of the treated and control groups is not a good measure of the treatment effect when randomization fails to create balanced treatment and control groups.

When using the intention-to-treat method to deal with non-compliance:

One way to check for balance between the treatment and control groups is to look for a statistically significant coefficient on the treatment dummy variable in a model in which variable X1 is the dependent variable.

The Quest for Causality

Stats in the Wild: Good Data Practices

Bivariate Ols: the Foundation of Econometric Analysis

Hypothesis Testing and Interval Estimation: Answering Research

Multivariate Ols: Where the Action Is

Dummy Variables: Smarter Than You Think

Specifying Models

Using Fixed Effects Models to Fight Endogeneity in Panel Data a

Instrumental Variables: Using Exogenous Variation to Fight Endogen

Regression Discontinuity: Looking for Jumps in Data

Dummy Dependent Variables

Time Series: Dealing With Stickiness Over Time

Advanced Ols

Advanced Panel Data

Filters

Exam 10: Experiments: Dealing With Real-World Challenges

A natural experiment occurs when:

Suppose we run an experiment and measure several possible independent variables after treatment. Which of the following best characterizes why this is a problem?

Which of the following is a reason on as to why non-compliance can be a problem?

In order to check for balance we assess whether the dependent variable differs for the treatment and control groups.

Which of the following explains why a 2SLS model can be used to analyze randomized experiments with imperfect compliance?

Attrition generally will not lead to bias because OLS simply deletes these observations from the sample.

When faced with issues of non-compliance, the coefficient on an intention-to-treat treatment variable will be

Which of the following is a method for dealing with attrition in a randomized experiment?

Which of the following is not one of the problems that leads to endogeneity in randomized experiments.

Which of the following must be undertaken when the treatment and control groups differ with regard to some independent variable, X1.

A compliance problem can lead to endogeneity because non-compliance is generally not-random.

The difference in means of the treated and control groups is not a good measure of the treatment effect when randomization fails to create balanced treatment and control groups.

When using the intention-to-treat method to deal with non-compliance:

One way to check for balance between the treatment and control groups is to look for a statistically significant coefficient on the treatment dummy variable in a model in which variable X1 is the dependent variable.

The Quest for Causality

Stats in the Wild: Good Data Practices

Bivariate Ols: the Foundation of Econometric Analysis

Hypothesis Testing and Interval Estimation: Answering Research

Multivariate Ols: Where the Action Is

Dummy Variables: Smarter Than You Think

Specifying Models

Using Fixed Effects Models to Fight Endogeneity in Panel Data a

Instrumental Variables: Using Exogenous Variation to Fight Endogen

Regression Discontinuity: Looking for Jumps in Data

Dummy Dependent Variables

Time Series: Dealing With Stickiness Over Time

Advanced Ols

Advanced Panel Data

Filters