Question 1

The value of the $\alpha$i parameter can potentially take on a different value for each unit i.

Accepted Answer

A) True 
 B)False  True

Question 2

Pooled OLS models are biased when:

Accepted Answer

A)  When the fixed effects are correlated with the independent variable. 
B)  When the fixed effects are correlated with the dependent variable. 
C)  When the fixed effects model is used when in reality there are no fixed effects in the data. 
D)  When we don't know the direction of the fixed effects. 
A)  When the fixed effects are correlated with the independent variable. 
B)  When the fixed effects are correlated with the dependent variable. 
C)  When the fixed effects model is used when in reality there are no fixed effects in the data. 
D)  When we don't know the direction of the fixed effects. A

Question 3

In order to employ the LSDV approach in a one-way fixed effects model, we have to:

Accepted Answer

A)  Create a dummy variable for every unit and include it in the model. 
B)  Create a dummy variable for every unit and include all but one in the model. 
C)  Subtract the unit specific averages from both the dependent and independent variables. 
D)  Create dummy variables for every unit and subtract the unit specific average from the dependent variable. 
A)  Create a dummy variable for every unit and include it in the model. 
B)  Create a dummy variable for every unit and include all but one in the model. 
C)  Subtract the unit specific averages from both the dependent and independent variables. 
D)  Create dummy variables for every unit and subtract the unit specific average from the dependent variable. B

Question 4

Which of the following factors could lead us to believe that a fixed effect model is necessary to control for endogeneity when examining the relationship between corruption (dependent) and regime type (independent) across countries over time?

Accepted Answer

A)  Religion 
B)  Culture 
C)  External threat 
D)  All of the above. 
A)  Religion 
B)  Culture 
C)  External threat 
D)  All of the above.

Question 5

List and provide a brief explanation of three potential fixed effects that could bias a model in the analysis of the relationship between attendance at a private school and test scores.

Accepted Answer

parents' income, par

Question 6

Given Yit = B0 + B1Treatedi + B2Aftert + B3(Treatedi*Aftert) + eit, the expected value of the dependent variable for those who got the treatment before they received it is:

Accepted Answer

A)  B0+B1+B2+B3 
B)  B0+B1+B3 
C)  B0+B1 
D)  B0+B1+B2 
A)  B0+B1+B2+B3 
B)  B0+B1+B3 
C)  B0+B1 
D)  B0+B1+B2

Question 7

When dealing with fixed effects, a de-meaned model approach is superior to the LSDV approach because the de-meaned model gives us more accurate coefficient estimates.

Accepted Answer

A) True 
 B)False

Question 8

Given Yit = B0 + B1Treatedi + B2Aftert + B3(Treatedi*Aftert) + eit, the expected value of the dependent variable for those who in the control group after the treatment was administered is:

Accepted Answer

A)  B0+B1+B2+B3 
B)  B0+B1+B3 
C)  B0+B1 
D)  B0+B2 
A)  B0+B1+B2+B3 
B)  B0+B1+B3 
C)  B0+B1 
D)  B0+B2

Question 9

Explain why a fixed effects model is not necessary in a case where fixed effects exist but the fixed effects are not correlated with the independent variable.

Accepted Answer

In this case, the conditions f

Question 10

Which of the following is a reason why fixed effect models can't estimate coefficients on variables that do not vary within unit?

Accepted Answer

A)  Because no fixed effects exists in such a case. 
B)  The de-meaned variables will not vary. 
C)  We can estimate such a variable with fixed effects. 
D)  The errors will be homoscedastic. 
A)  Because no fixed effects exists in such a case. 
B)  The de-meaned variables will not vary. 
C)  We can estimate such a variable with fixed effects. 
D)  The errors will be homoscedastic.

Question 11

The fixed effect ($\alpha$) is the part of the unobserved error and it is different for every observation for unit i.

Accepted Answer

A) True 
 B)False

Question 12

A pooled model will produce biased coefficient estimates when factors in the error term are not correlated with the independent variable.

Accepted Answer

A) True 
 B)False

Question 13

Having a model that controls for fixed effects even when the fixed effects are 0 ($\alpha$i=0 for all units) will not cause bias.
:

Accepted Answer

A) True 
 B)False

Question 14

Explain why coefficients on variables that do not vary within the unit cannot be estimated in a fixed effects model.

Accepted Answer

The de-meaned variable will be

Question 15

The benefit of the de-meaned approach over the LSDV approach is:

Accepted Answer

A)  It gives more accurate coefficient estimates. 
B)  The coefficient estimates have a lower standard error. 
C)  Easier to implement if there are a lot of different units. 
D)  We will be able to estimate coefficients on variables that do not vary within unit. 
A)  It gives more accurate coefficient estimates. 
B)  The coefficient estimates have a lower standard error. 
C)  Easier to implement if there are a lot of different units. 
D)  We will be able to estimate coefficients on variables that do not vary within unit.

Question 16

Which of the following is true about fixed effects models?

Accepted Answer

A)  We cannot use fixed effects models when we suspect there are fixed effects but no such effect exists ($\alpha$i=0 for all units) 
B)  Fixed effects models cannot generate a coefficient on a variable that is constant within all units. 
C)  Fixed effects will generate a coefficient on a variable as long as that variable differs across units. 
D)  A fixed effects model is not possible when there are a large number of units in the model. 
A)  We cannot use fixed effects models when we suspect there are fixed effects but no such effect exists ($\alpha$i=0 for all units) 
B)  Fixed effects models cannot generate a coefficient on a variable that is constant within all units. 
C)  Fixed effects will generate a coefficient on a variable as long as that variable differs across units. 
D)  A fixed effects model is not possible when there are a large number of units in the model.

Question 17

Write down and explain the basic difference in difference equation. Point out which of the coefficients we are most interested in.

Accepted Answer

We are most interested in B3, the coeffi

Question 18

Describe the issues involved with using a pooled model and given an example of why a pooled model could lead to biased results.

Accepted Answer

A pooled model ignores the fact that obs

Question 19

Which of the following is most likely not a factor that would cause endogeneity (fixed effect is zero) when examining the relationship of bilateral trade (dependent) and trade alliances (independent).

Accepted Answer

A)  Shared language 
B)  Distance between trading partners 
C)  Regime type 
D)  Inflation rate 
A)  Shared language 
B)  Distance between trading partners 
C)  Regime type 
D)  Inflation rate

Question 20

In order to run a two-way fixed effect model, we can use:

Accepted Answer

A)  A de-meaned approach 
B)  A LSDV approach 
C)  Both a and b 
D)  None of the above. 
A)  A de-meaned approach 
B)  A LSDV approach 
C)  Both a and b 
D)  None of the above.

The value of the $\alpha$ _iparameter can potentially take on a different value for each unit i.

Pooled OLS models are biased when:

In order to employ the LSDV approach in a one-way fixed effects model, we have to:

Which of the following factors could lead us to believe that a fixed effect model is necessary to control for endogeneity when examining the relationship between corruption (dependent) and regime type (independent) across countries over time?

List and provide a brief explanation of three potential fixed effects that could bias a model in the analysis of the relationship between attendance at a private school and test scores.

Given Y_it= B₀+ B₁Treated_i+ B2After_t+ B3(Treated_i*After_t) + e_it, the expected value of the dependent variable for those who got the treatment before they received it is:

When dealing with fixed effects, a de-meaned model approach is superior to the LSDV approach because the de-meaned model gives us more accurate coefficient estimates.

Given Y_it= B₀+ B₁Treated_i+ B2After_t+ B3(Treated_i*After_t) + e_it, the expected value of the dependent variable for those who in the control group after the treatment was administered is:

Explain why a fixed effects model is not necessary in a case where fixed effects exist but the fixed effects are not correlated with the independent variable.

Which of the following is a reason why fixed effect models can't estimate coefficients on variables that do not vary within unit?

The fixed effect ( $\alpha$ ) is the part of the unobserved error and it is different for every observation for unit i.

A pooled model will produce biased coefficient estimates when factors in the error term are not correlated with the independent variable.

Having a model that controls for fixed effects even when the fixed effects are 0 ( $\alpha$ _i=0 for all units) will not cause bias. :

Explain why coefficients on variables that do not vary within the unit cannot be estimated in a fixed effects model.

The benefit of the de-meaned approach over the LSDV approach is:

Which of the following is true about fixed effects models?

Write down and explain the basic difference in difference equation. Point out which of the coefficients we are most interested in.

Describe the issues involved with using a pooled model and given an example of why a pooled model could lead to biased results.

Which of the following is most likely not a factor that would cause endogeneity (fixed effect is zero) when examining the relationship of bilateral trade (dependent) and trade alliances (independent).

In order to run a two-way fixed effect model, we can use:

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

Instrumental Variables: Using Exogenous Variation to Fight Endogen

Experiments: Dealing With Real-World Challenges

Regression Discontinuity: Looking for Jumps in Data

Dummy Dependent Variables

Time Series: Dealing With Stickiness Over Time

Advanced Ols

Advanced Panel Data

Filters

Exam 8: Using Fixed Effects Models to Fight Endogeneity in Panel Data a

The value of the α\alphaα i parameter can potentially take on a different value for each unit i.

Pooled OLS models are biased when:

In order to employ the LSDV approach in a one-way fixed effects model, we have to:

Which of the following factors could lead us to believe that a fixed effect model is necessary to control for endogeneity when examining the relationship between corruption (dependent) and regime type (independent) across countries over time?

List and provide a brief explanation of three potential fixed effects that could bias a model in the analysis of the relationship between attendance at a private school and test scores.

Given Yit = B0 + B1Treatedi + B2Aftert + B3(Treatedi*Aftert) + eit, the expected value of the dependent variable for those who got the treatment before they received it is:

When dealing with fixed effects, a de-meaned model approach is superior to the LSDV approach because the de-meaned model gives us more accurate coefficient estimates.

Given Yit = B0 + B1Treatedi + B2Aftert + B3(Treatedi*Aftert) + eit, the expected value of the dependent variable for those who in the control group after the treatment was administered is:

Explain why a fixed effects model is not necessary in a case where fixed effects exist but the fixed effects are not correlated with the independent variable.

Which of the following is a reason why fixed effect models can't estimate coefficients on variables that do not vary within unit?

The fixed effect ( α\alphaα ) is the part of the unobserved error and it is different for every observation for unit i.

A pooled model will produce biased coefficient estimates when factors in the error term are not correlated with the independent variable.

Having a model that controls for fixed effects even when the fixed effects are 0 ( α\alphaα i=0 for all units) will not cause bias. :

Explain why coefficients on variables that do not vary within the unit cannot be estimated in a fixed effects model.

The benefit of the de-meaned approach over the LSDV approach is:

Which of the following is true about fixed effects models?

Write down and explain the basic difference in difference equation. Point out which of the coefficients we are most interested in.

Describe the issues involved with using a pooled model and given an example of why a pooled model could lead to biased results.

Which of the following is most likely not a factor that would cause endogeneity (fixed effect is zero) when examining the relationship of bilateral trade (dependent) and trade alliances (independent).

In order to run a two-way fixed effect model, we can use:

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

Instrumental Variables: Using Exogenous Variation to Fight Endogen

Experiments: Dealing With Real-World Challenges

Regression Discontinuity: Looking for Jumps in Data

Dummy Dependent Variables

Time Series: Dealing With Stickiness Over Time

Advanced Ols

Advanced Panel Data

Filters

The value of the $\alpha$ _iparameter can potentially take on a different value for each unit i.

Given Y_it= B₀+ B₁Treated_i+ B2After_t+ B3(Treated_i*After_t) + e_it, the expected value of the dependent variable for those who got the treatment before they received it is:

Given Y_it= B₀+ B₁Treated_i+ B2After_t+ B3(Treated_i*After_t) + e_it, the expected value of the dependent variable for those who in the control group after the treatment was administered is:

The fixed effect ( $\alpha$ ) is the part of the unobserved error and it is different for every observation for unit i.

Having a model that controls for fixed effects even when the fixed effects are 0 ( $\alpha$ _i=0 for all units) will not cause bias. :