Question 1

You can control for sample-selection bias by performing

Accepted Answer

Sample-selection bias occurs when the sample used for analysis is not random and is based on some criteria, which can lead to biased estimates. Heckman selection correction is used to control for sample-selection bias by modeling the process of sample selection and estimating the probability of selection. It then transforms the original equation to correct for the bias caused by sample selection. Two-stage least squares and difference-in-difference estimation are used to control for endogeneity, while weighted least squares is used to control for heteroscedasticity. Hence, C is the best choice to control for sample-selection bias.

Question 2

Sample-selection bias occurs when

Accepted Answer

Sample-selection bias occurs when individuals non-randomly select themselves into a given outcome of the dependent variable. This can lead to skewed results and incorrect conclusions about the relationship between variables.

Question 3

Sample-selection bias presents a problem because it

Accepted Answer

Sample-selection bias occurs when a subset of observations is systematically excluded from the sample, leading to biased and inconsistent OLS coefficient estimates. This bias occurs when the omitted observations are correlated with both the dependent variable and the independent variable.

Question 4

Non-negative count data occur when the dependent variable takes on

Accepted Answer

Non-negative count data refers to data that are integers that are greater than or equal to 0 and arise from counting. This means that the data cannot be negative or continuous. Therefore, option C is the best choice as it accurately describes the nature of non-negative count data.

Question 5

When performing difference-in-difference estimation,the control group is the group

Accepted Answer

The control group is the group for which the policy shock did not occur. This serves as a comparison group to estimate the effect of the policy shock on the treatment group.

Question 6

Quantile regression

Accepted Answer

Quantile regression estimates the conditional median or other percentiles of the dependent variable, and is not limited to estimating marginal effects at the mean like ordinary least squares regression. Skewed distributions do not necessarily produce biased estimates in quantile regression, unlike some other methods. Quantile regression can be conducted in Excel using various add-ins or programming in VBA.

Question 7

Quantile regression is different than OLS in that it

Accepted Answer

Quantile regression estimates the effects of independent variables on different percentiles of the dependent variable, not just at the mean value. Thus, it does not estimate marginal effects at the mean values of the dependent and independent variables, which is the case for OLS.

Question 8

The first-stage in the Heckman selection correction is estimating

Accepted Answer

The first stage in the Heckman selection correction is estimating the individual's self-selection decision and using those estimates to calculate inverse Mill's ratios. This involves estimating a probit model to predict the probability of being selected into the sample, and then calculating the inverse Mill's ratio for each observation.

Question 9

Suppose you wish to determine factors affecting the number of surfers observed surfing at a given surf spot,an appropriate model to estimate the model would be

Accepted Answer

Since we are modeling the count of surfers, the appropriate model to use would be the Poisson model. The Poisson model is designed specifically for count data, and it models the number of occurrences of a certain event in a fixed time interval or in a certain area. Therefore, the Poisson model is the best choice for estimating the number of surfers observed at a given surf spot.

Question 10

In which of the following cases would you want to estimate a Negative Binomial model?

Accepted Answer

Negative Binomial models are commonly used when dealing with count data or discrete data where the variance is greater than the mean. Negative Binomial models account for overdispersion and are more appropriate when dealing with non-negative count data than Poisson models, which assume the mean is equal to the variance.

Question 11

Quasi-experimental methods attempt to

Accepted Answer

The answer of Quasi-experimental methods attempt to
...

Question 12

You can choose between the Poisson and the Negative Binomial models by performing a

Accepted Answer

Testing for over-dispersion is the best way to determine whether the data are better described by a Poisson or a Negative Binomial model. The Poisson model assumes that the mean and variance of the data are equal, which may not always be the case. The negative binomial model allows for over-dispersion and is often a better choice when the variance of the data is larger than the mean.

Question 13

In which of the following cases would you want to estimate a Poisson model?

Accepted Answer

A Poisson model is appropriate for non-negative count data, where the mean and variance of the dependent variable are equal. Examples of such data could include the number of accidents per day or the number of defects in a manufacturing process. The other options listed do not necessarily require a Poisson model.

Question 14

Non-negative count data presents a challenge because OLS estimates

Accepted Answer

Non-negative count data violates the assumption of normality required while estimating with OLS. The OLS estimates in the case of non-negative count data would be biased.

Question 15

The second-stage in the Heckman selection correction is including the _____ in the second-stage regression to control for the potential sample-selection bias.

Accepted Answer

In the Heckman selection correction, the inverse Mills ratio is included in the second-stage regression to account for the potential sample-selection bias. The inverse Mills ratio is derived from the first-stage regression, which models the self-selection variable as a function of the observed variables. The inverse Mills ratio is then used as a control variable in the second-stage regression to correct for sample selection bias.

Question 16

Difference-in-difference estimators attempt to

Accepted Answer

Difference-in-difference estimators compare changes in outcomes before and after a treatment is introduced to a group that receives treatment and a control group that doesn't. This method attempts to replicate randomized clinical trials by estimating the impact of a policy intervention. The other options are not accurate descriptions of difference-in-difference estimators.

Question 17

Suppose you wish to explain the number of nights per week that individuals eat dinner at a restaurant,an appropriate model to estimate would be

Accepted Answer

The negative binomial model is appropriate for count data such as the number of nights per week that individuals eat dinner at a restaurant. It allows for over-dispersion (variation in the data greater than what would be expected under a Poisson model) and accounts for the fact that the variance is not equal to the mean.

Question 18

In which of the following cases would you want to use a Heckman selection correction model?

Accepted Answer

The Heckman selection correction model is used when individuals non-randomly select different outcomes of the dependent variable. The model corrects for sample selection bias and allows estimation of the parameters of interest while accounting for the selection process.

Question 19

When performing difference-in-difference estimation,the treatment group is the group

Accepted Answer

The treatment group is the group for which the policy shock occurred. The difference-in-differences approach compares the change in the outcome variable over time for the treatment group with that of a control group. The treatment group is the group that experienced the policy shock, while the control group did not.

Question 20

In which of the following cases would you want to use quantile regression?

Accepted Answer

Quantile regression is useful when you suspect that the marginal effects are different for different values of the dependent variable. It allows you to estimate the conditional quantiles of the dependent variable as a function of the independent variables, providing a more nuanced understanding of the relationship between the variables than traditional linear regression.

Quiz 15: Quantile Regression, Count Data, Sample Selection Bias, and Quasi-Experimental Methods