AAEC 4302 Ag advanced research
What values of the VIF are said to be problematic? (i.e., they indicate a multicollinearity problem)
10 and higher
For the problem related to the effect of math aptitude test score on a Statistics class grade: What are the number of observations used in the model?
5
Good luck!
:)
Suppose that x is the high school GPA and y is the college GPA, and we happen to know that E[collegeGAP]= 1.5+ 0.5highschoolGPA. This means that:
A 1 point increase in highschoolGPA will increase the average collegeGPA of all students by 0.5 points
For the estimated model: log(y)hat = 10 + 0.56log(x), what is the correct interpretation of the coefficient related to log(x)?
A 1% change in x changes y by 0.56%
Econometric models can be used to:
All of the above Answers: To estimate economic relationships To test economic theories To evaluate policies All of the above
What does BLUE stands for?
Best, Linear, Unbiased Estimators
An estimate is the formula used to estimate a coefficient in the regression model
False
For hypothesis testing, the Ha: Bj = (with line in =) 0 is the same as saying that xj does NOT have an effect on y.
False
For tests of hypotheses about individual Bj's, a p-value is the probability related to the t-critical value from the t Table.
False
For the model: log(y)hat = 10 + 0.56log(x), calculation of the predicted value of y at x=10 equals e^(10+0.56log(10)).[Note that e^() is the constant e to the value inside the parenthesis]
False
In a quadratic model, the marginal effects are constant for any value of x.
False
In multiple linear regression, we cannot include as explanatory variables (x's) those variables denoting qualitative information
False
In the MLR model, a VIF=0 for a variable xj indicates that there is no correlation between this variable and all other explanatory variables.
False
Large t-calculated values are associated with large p-values.
False
Log-log models are very popular in Economics because the intercept coefficient has a very meaningful and easy to understand interpretation
False
Most of the data used in Economics in general is "experimental" data (i.e., data from experiments)
False
Multiple linear regression assumption 6 (MLR6) is related to the normality of the explanatory variables
False
SSE stands for error sum squares and corresponds to a measure of the variability in y that is not explained by the model
False
Simple Linear Assumption 1 (Linear in Parameters) indicates that no matter what is the real relatiomship between E[y] and x, the linear regression model is always the correct model.
False
Tests of hypotheses are about the estimated Bj's
False
The 95% confidence interval for Bj in a MLR model is [10, 20]. Since the confidence interval does not contain zero, we CAN'T reject the Ho: Bj=0.
False
The R2 measures the percentage of total variability in the independent variable (x) that is explained by the model
False
The expression E[y] means the estimated value of y
False
The formulas used to estimate B0 hat and B1 hat ensure that the sum of squared residuals (SSR= N all sum of squares U Hat Ui2 ) is a large as possible:
False
The log function used in Econometric models in this class refers to logarithms with base = 10
False
The main objective of the econometric model is to analyze the effect of an outcome (or dependent variable) on a set of factors (or explanatory variables)
False
The total variability of y that is not explained by the model = total explained variability of y in the model + total observed variability in y.
False
What is NOT an advantage of the multiple linear regression model relative to the simple linear regression model?
It is easier to estimate as the formulas are simpler
What are the minimum set of assumptions required for the least square estimators in the MLR model to be unbiased estimators of the true population coefficients?
MLR1, MLR2, MLR3, MLR4
What is correct regarding marginal effects in linear and nonlinear models?
Marginal effects in linear models are constant whereas marginal effects in nonlinear models are non-constant
What are the maximum and minimum values for R2?
Minimum = 0; maximum=1
The least squares estimators and are unbiased estimators for and since:
On average, they are equal to the true population values (i.e., E[]= and E[]= ).
Two-sided confidence intervals about the Bj's:
Provide a range of potential values of the true (population) coefficients
What assumptions are needed for the Gauss Markov Theorem to work?
SLR1 to SLR 5
One star (*) next to a coefficient in the Table form used to report regression results denote:
Statistical significance at the 10% level
Suppose that x is the high school GPA and y is the college GPA, and we happen to know that E[collegeGAP]= 1.5+ 0.5highschoolGPA. This means that:
The average college GPA for students whose high school GPA was 0 is 1.5.
What does endogeneity refers to?
The correlation between x and u
Which of the following is NOT an example of characteristics denoting qualitative information
The price of gasoline
The stars (*) used to report regression results denote information about:
The statistical significance of the individual explanatory variables
A dummy variable is a variable that takes only 2 values: zero and one
True
A log-level type of model indicates that the dependent variable is in log form, but the explanatory variable is not in log form
True
For hypothesis testing, the Ho: Bj=0 is the same as saying that xj does NOT have an effect on y
True
For tests of hypotheses about individual Bj's, a p-value is the probability related to the calculated t-statistic
True
For tests of hypotheses about individual Bj's, if the calculated t-statistic falls outside the rejection region we fail to reject the Ho.
True
In a MLR model, confidence intervals for can be used to 1) obtain a range of estimates of Bj, and 2) to conduct tests of hypothesis regarding Bj.
True
In the MLR model multicollinearity is a problem since it increases the variances of the estimated coefficients.
True
In the MLR model, the maximum value of the VIF is infinity ()
True
In the simple linear regression model , use of the least squares estimators for and , and , result in different estimated values when different samples are used:
True
In the simple linear regression model y = Bo + B1x + U , use of the least squares estimators for B0 and B1 , B0 hat and B1 hat , result in different estimated values when different samples are use
True
Log-log models are very popular in Economics because the coefficients can be interpreted as elasticities
True
Marginal effects and elasticities measure the effect of an explanatory variable on the dependent variable.
True
Most of the data used in Economics in general is "observational" data
True
Multiple linear regression assumption 6 (MLR6) is related to the normality of the error term
True
Random sampling means that each individual is the population has the same probability of being selected in the sample
True
SST stands for "Total Sum of Squares" and corresponds to a measure of the total variability of the dependent variable.
True
The "Best" property in the Gauss Markov theorem refers to the variance of the estimators (i.e., they have small relative variance).
True
The Rj^2 term to the R-square of a regression of xj as the dependent variable and all other x's as explanatory variables.
True
The expression E[y] means the expected value of y:
True
The formulas for the variances of B0 Hat and B1 Hat provided in class are only valid if the homoskedasticity assumption is satisfied. If it is not satisfied, the formulas are no longer valid.
True
The log-log type of model indicates that both the dependent variable and the explanatory variable are in log form
True
The main objective of the econometric model is to analyze the effect of a set of factors (or explanatory variables) on an outcome (or dependent variable)
True
The marginal effect is the same as the same effect of an additional unit of x on the average value of y.
True
The t distribution is an asymmetric distribution
True
Theoretically, the variance of the errors equals the variance of the dependent variable
True
What is the correct formula for the residuals or predicted errors?
Ui hat = yi - yi hat
What is the correct formula for the residuals or predicted errors?
Ui(hat) = Yi - Yi(hat)
What does VIF stands for?
Variance Inflation Factor
For a 2 tailed test of hypothesis about an individual coefficient, if the p-value is 0.09.
We reject the Ho at the 10%
For a 2 tailed test of hypothesis about an individual coefficient, if the p-value is 0.05
We reject the Ho at the 5% and 10% level
For the linear regression model y=bo +dx + u, where x is a dummy variable:
bo represents the average value of y when x=0
For a linear regression model y=bo +dx + u, where x is a dummy variable:
d represents the difference between the average value of y when x=1 and the average value of y when x=0
For the estimated quadratic model: Y hat = 10 + 12X - 3x^2 , what is the estimated effect of an additional unit of x when x=0
x = 12
The simple linear regression model: y = Bo + B1x +u implies a linear relationship between
x and the average value of y (which is E[y])
In the econometric model: y = Bo + B1x + u
y is the dependent variable, x is an explanatory variable or factor