Introdoctury Econometrcis Midterm
437 Questions
7 Views

Choose a study mode

Play Quiz
Study Flashcards
Spaced Repetition
Chat to Lesson

Podcast

Play an AI-generated podcast conversation about this lesson

Questions and Answers

Which aspect of the econometric methods will students learn in practical exercises?

  • Application of methods to actual data. (correct)
  • Only simulation techniques.
  • Theoretical aspects only.
  • Primarily historical perspectives.

What is the significance of consistency in econometric estimators?

  • It indicates the estimator becomes more accurate as sample size increases. (correct)
  • It ensures the estimator is unbiased.
  • It allows the estimator to be dependent on the observer.
  • It is not considered a minimal requirement for estimators.

Which of the following is NOT one of the three central data requirements?

  • Validity
  • Reliability
  • Objectivity
  • Scalability (correct)

What is the primary focus of economic theory?

<p>Causality (D)</p> Signup and view all the answers

What does mutual consistency in empirical strategy refer to?

<p>Ensuring repeated measurements yield similar results. (B)</p> Signup and view all the answers

What is a common misconception regarding correlation and causality?

<p>Correlation implies causation. (C)</p> Signup and view all the answers

Which of the following best illustrates reverse causality?

<p>Higher income leads to higher education. (C)</p> Signup and view all the answers

Which unit of observation would be most appropriate for a study on national economic performance?

<p>Countries (C)</p> Signup and view all the answers

According to the content, where can researchers obtain reliable data?

<p>National statistical agencies (C)</p> Signup and view all the answers

In a linear regression model, what does the term β1 represent?

<p>The slope of the regression line. (D)</p> Signup and view all the answers

What is a potential reason for eliminating data points in empirical research?

<p>To remove impossible realizations or errors. (D)</p> Signup and view all the answers

What does a confounding factor in a causal relationship imply?

<p>It influences both the independent and dependent variables. (D)</p> Signup and view all the answers

What type of data organization is the ICPSR?

<p>A database for political and social research (C)</p> Signup and view all the answers

How does econometrics contribute to economic theory?

<p>By establishing the size of effects reliably. (A)</p> Signup and view all the answers

Which notable econometrician emphasized the importance of consistency in estimators?

<p>Clive W.J. Granger (D)</p> Signup and view all the answers

What is a potential problem when proposing subsidies based on correlation?

<p>The correlation may suggest a false causation. (A)</p> Signup and view all the answers

Which of the following is NOT a characteristic of a good economic theory?

<p>Irrelevance to policy-making. (D)</p> Signup and view all the answers

What is the purpose of scientific research in econometrics?

<p>To formulate a research question and answer it using various models (D)</p> Signup and view all the answers

Which of the following is NOT a basic econometric tool listed in the course plan?

<p>Maximum likelihood estimation (B)</p> Signup and view all the answers

Which of the following statements best describes the concept of endogeneity in econometrics?

<p>It occurs when the regressor and error term are correlated (D)</p> Signup and view all the answers

What is the significance of 'as simple as possible, as complex as necessary' in research design?

<p>It emphasizes finding a balance between simplified models and realistic complexity (D)</p> Signup and view all the answers

Which method listed is considered an advanced econometric tool?

<p>Panel data methods (B)</p> Signup and view all the answers

In econometrics, what is the primary focus of discrete choice models?

<p>To analyze choices made by individuals among discrete alternatives (D)</p> Signup and view all the answers

Which of the following is a potential problem encountered in regression analysis?

<p>Multicollinearity (D)</p> Signup and view all the answers

What do quasi-experiments involve in econometric analysis?

<p>Natural variations in data without random assignment (B)</p> Signup and view all the answers

How does an increase in the error variance, σ 2, affect the variance of the slope estimate?

<p>It increases the variance of the slope estimate. (B)</p> Signup and view all the answers

What is the purpose of using residuals in the estimation of error variance?

<p>To estimate the unknown error variance. (D)</p> Signup and view all the answers

In the formula for estimating variance, what does the term (N - K - 1) represent?

<p>The degrees of freedom adjustment for the OLS estimation. (D)</p> Signup and view all the answers

What consequence does a larger sample size have on the variance of the slope estimate?

<p>It decreases the variance of the slope estimate. (C)</p> Signup and view all the answers

Which of the following expressions represents an unbiased estimator for the error variance σ²?

<p>$\frac{1}{N - K - 1} \sum_{i=1}^{N} u_i^2$ (C)</p> Signup and view all the answers

What does R-squared represent in regression analysis?

<p>The fraction of the total sum of squares explained by the model (B)</p> Signup and view all the answers

Which statement is true regarding R-squared when adding more independent variables to a model?

<p>R-squared never decreases when another independent variable is added (C)</p> Signup and view all the answers

Why is R-squared not a good metric for comparing different models?

<p>It typically increases with more independent variables (D)</p> Signup and view all the answers

How is R-squared related to the correlation coefficient?

<p>R-squared is equal to the squared correlation coefficient between actual and predicted values (D)</p> Signup and view all the answers

What range does the R-squared value fall within?

<p>Between 0 and 1 (B)</p> Signup and view all the answers

What is the primary goal of a randomized controlled experiment in the context of causal effect estimation?

<p>To ensure that all variables other than x remain constant (D)</p> Signup and view all the answers

What does the notation $ rac{ ext{∂E}(y|x,u)}{ ext{∂x}}$ represent in econometric analysis?

<p>The causal effect of a unit change in x on y (C)</p> Signup and view all the answers

Which identification assumption must hold in simple linear regression (SLR)?

<p>There is a linear relationship between X and Y (A)</p> Signup and view all the answers

What is a characteristic of the random sample used in econometric analysis?

<p>Entities are selected randomly from the same population (C)</p> Signup and view all the answers

What is a potential issue with relying on non-i.i.d samples in econometric studies?

<p>It may lead to biased estimates due to dependence between observations (A)</p> Signup and view all the answers

Which of the following is not a characteristic of how ideal randomized controlled experiments are designed?

<p>Participants can opt-out of the study (D)</p> Signup and view all the answers

How does the SLR identification assumption relate to unobserved observation pairs?

<p>It asserts that the relationship must also hold for them (C)</p> Signup and view all the answers

What is one requirement for a valid treatment effect measure in econometric research?

<p>Random assignment must be used to minimize selection bias (D)</p> Signup and view all the answers

What does a variance of zero in the independent variable indicate?

<p>There is no variation in the independent variable. (A)</p> Signup and view all the answers

What assumption must be satisfied for the zero conditional mean assumption to hold?

<p>Knowledge of x provides no information about u. (D)</p> Signup and view all the answers

Why is the zero mean assumption for the error term not considered restrictive?

<p>It can be adjusted using the intercept term β0. (B)</p> Signup and view all the answers

What impact can large outliers have on regression coefficients such as β̂1?

<p>They may lead to skewed and unreliable estimates. (A)</p> Signup and view all the answers

What is implied by the orthogonality condition in econometrics?

<p>E(xu) must equal zero for correct estimations. (C)</p> Signup and view all the answers

What does E(u) = 0 signify in a regression model?

<p>The average error in the model is zero. (C)</p> Signup and view all the answers

When evaluating outliers, which of the following questions is most crucial?

<p>Why is it an outlier in context? (D)</p> Signup and view all the answers

Which of the following relationships best exemplifies the need for a zero conditional mean assumption?

<p>No association between errors and the input variable. (B)</p> Signup and view all the answers

What does the multiple R-squared value of 0.4975 indicate in the regression output?

<p>The model explains 49.75% of the variance in life expectancy. (C)</p> Signup and view all the answers

Which of the following correctly describes the significance of the p-value for public expenditure in the regression output?

<p>The variable is statistically significant at the 0.001 level. (C)</p> Signup and view all the answers

What does the term 'Intercept' in the regression output represent?

<p>The expected life expectancy when public expenditure is zero. (C)</p> Signup and view all the answers

What is the residual standard error in this regression analysis indicative of?

<p>The standard deviation of the residuals from the fitted model. (D)</p> Signup and view all the answers

In the context of the OLS estimator, what does the assumption SLR.1 state?

<p>The relationship between the dependent and independent variables is linear. (D)</p> Signup and view all the answers

Which statement reflects the importance of the F-statistic in the regression output?

<p>It measures the overall significance of the regression model. (D)</p> Signup and view all the answers

How many degrees of freedom are indicated in the F-statistic section of the regression output?

<p>1 for the regression and 28 for the residuals. (D)</p> Signup and view all the answers

What does the Adjusted R-squared value of 0.4796 suggest about the model?

<p>The model accounts for approximately 48% of the variance in life expectancy, adjusting for the number of predictors. (B)</p> Signup and view all the answers

In the context of the data presented, which country had the highest life expectancy?

<p>JPN (A)</p> Signup and view all the answers

How is the fitted value related to the residual in the regression context?

<p>Residual represents the error in prediction of the fitted value. (B)</p> Signup and view all the answers

What does the assumption SLR.4 state about the error term in a regression model?

<p>The expected value of the error term conditional on the independent variable is zero. (C)</p> Signup and view all the answers

Which assumption is violated if the OLS estimators are biased?

<p>Assumption SLR.4 - Zero conditional mean. (C)</p> Signup and view all the answers

What is the significance of E(β̂1 |x) equaling $β_1$ in the context of OLS?

<p>The estimator is unbiased under the specified assumptions. (B)</p> Signup and view all the answers

What does the term $S_x^2$ represent in the context of OLS estimators?

<p>The sum of the squared deviations of the independent variable from its mean. (A)</p> Signup and view all the answers

Under which condition can the OLS estimator β̂1 be regarded as consistent?

<p>Sample size must approach infinity and assumptions SLR.1 to SLR.4 hold. (B)</p> Signup and view all the answers

What does the unbiasedness of OLS estimators imply about their expected values?

<p>The expected values are equal to the population parameters. (D)</p> Signup and view all the answers

In proving the unbiasedness of OLS, what component is considered when rewriting the estimator?

<p>The population parameter associated with the independent variable. (D)</p> Signup and view all the answers

When either SLR.1 through SLR.4 is not satisfied, what is the implication for OLS estimators?

<p>There is no guarantee that OLS estimators are unbiased. (A)</p> Signup and view all the answers

What condition must hold true for the summation of $(x_i - ar{x})$ across all observations?

<p>It must approach zero as sample size increases. (C)</p> Signup and view all the answers

What implication does the assumption of finite fourth moments (E x^4 < ∞ and E y^4 < ∞) have on OLS estimators?

<p>It ensures the variability of the estimators remains manageable. (D)</p> Signup and view all the answers

What condition must be met for the estimator β̂ to be defined?

<p>det(X0 X) must not equal 0 (D)</p> Signup and view all the answers

In the context of OLS estimation, what does the matrix $M_X$ represent?

<p>The projection of y onto the regressor space (C)</p> Signup and view all the answers

What does the first-order condition with respect to β̂ indicate when set to zero?

<p>The residual sum of squares is minimized (B)</p> Signup and view all the answers

Which property of the matrix $P_X$ is correctly stated?

<p>$P_X^2 = P_X$ (A)</p> Signup and view all the answers

What does the expression $û = y - Xβ̂$ represent?

<p>The difference between observed and predicted values (C)</p> Signup and view all the answers

What does the notation $\hat{\beta} = (X'X)^{-1}X'y$ represent?

<p>The formula for the ordinary least squares (OLS) estimator. (B)</p> Signup and view all the answers

Which assumption is primarily concerned with the behavior of the error term in linear regression?

<p>Homoscedasticity of the error term. (C)</p> Signup and view all the answers

What implication does the assumption of homoskedasticity have on estimating variances?

<p>It allows for simplifying the calculations of the variance-covariance matrix. (D)</p> Signup and view all the answers

Which factor primarily affects the unbiasedness of the OLS estimator?

<p>The correct specification of the regression model. (C)</p> Signup and view all the answers

What does the variance-covariance matrix of the error term indicate?

<p>The distribution and relationship among the errors in the regression. (C)</p> Signup and view all the answers

If an error term has a variance $\sigma^2$, what does this imply about the error across all observations?

<p>All observations have the same level of error variability. (B)</p> Signup and view all the answers

What role does the assumption of the zero conditional mean play in OLS estimators?

<p>It guarantees the error term does not influence the independent variables. (C)</p> Signup and view all the answers

What does the expression $E(\beta) = \beta + (X'X)^{-1}X'E(u)$ imply about OLS estimators?

<p>The bias of the OLS estimator is dependent on the expected value of the error term. (D)</p> Signup and view all the answers

Under what condition is the alternative estimator β̃ considered unbiased?

<p>When DX = 0. (A)</p> Signup and view all the answers

What does the term DD0 represent in the context of the variance of β̃?

<p>A positive semidefinite matrix. (B)</p> Signup and view all the answers

What is the formula for the coefficient of determination R2?

<p>SSR/SST. (A)</p> Signup and view all the answers

How is the variance of β̃ derived from the variance of y?

<p>Using the matrix multiplication of C and the variance of y. (A)</p> Signup and view all the answers

What does the OLS estimator allow for when applied to the Frisch-Waugh theorem?

<p>Decomposing the effect of independent variables. (C)</p> Signup and view all the answers

What does R2 measure in the context of regression analysis?

<p>The strength of the relationship between dependent and independent variables. (B)</p> Signup and view all the answers

In the context of the Frisch-Waugh theorem, what does the variable $M1$ represent?

<p>A matrix that removes the effect of $X2$ on $y$. (A)</p> Signup and view all the answers

What is the significance of the residuals in the Frisch-Waugh theorem's interpretation?

<p>They show the uncaptured variation from another regression. (A)</p> Signup and view all the answers

Which of the following statements is true regarding the decomposition of total sum of squares (SST)?

<p>SST can be separated into SSR and SSE. (C)</p> Signup and view all the answers

When using the correction formulas for $etaˆ1$ and $etaˆ2$, what is the relationship between the variables being analyzed?

<p>Each variable relies on both variables for proper estimation. (B)</p> Signup and view all the answers

In the context of the Gauss-Markov theorem, which property is associated with the estimator β̃?

<p>β̃ is unbiased under certain conditions. (C)</p> Signup and view all the answers

What does the OLS estimator's second expression in the Frisch-Waugh theorem imply?

<p>The residual from regressing $y$ on $X1$ can be effectively used. (D)</p> Signup and view all the answers

What is implied by the concept of positive semidefiniteness in relation to variance matrices?

<p>They have non-negative eigenvalues. (A)</p> Signup and view all the answers

Which of the following represents the correct calculation method for $etaˆ1$ as per the theorem?

<p>$etaˆ1 = (X1^0 M2 X1)^{-1} (X1^0 M2 y)$ (C)</p> Signup and view all the answers

What does the decomposition y0 y = ŷ0 ŷ + û0 û reveal about the regression model?

<p>It separates explained variance from unexplained variance. (C)</p> Signup and view all the answers

How does the Frisch-Waugh theorem contribute to understanding linear regression models?

<p>It facilitates the understanding of partial effects while controlling for other regressors. (D)</p> Signup and view all the answers

What does the notation $(X1^0 X1)^{-1} X1^0 (y - X2 etaˆ2)$ ultimately indicate in econometric analysis?

<p>An estimation approach for assessing variable impacts while controlling for others. (A)</p> Signup and view all the answers

What does the variable y represent in the context of the statements regarding correlation and causality?

<p>Whether person i ends up dying (A)</p> Signup and view all the answers

What is the role of the independent variable x in the correlation versus causality discussion presented?

<p>Whether person i confuses correlation and causation (A)</p> Signup and view all the answers

What question can be derived to evaluate if x causes y in the context provided?

<p>Is y = 1 more likely for x = 0 than for x = 1? (B)</p> Signup and view all the answers

In the context of the statement about correlation and causality, what does the term 'observations' refer to?

<p>Individual data points of persons i (D)</p> Signup and view all the answers

What can be inferred about the relationship between correlation and causation based on the content provided?

<p>Causation can exist without correlation. (A)</p> Signup and view all the answers

Which estimation method is discussed as being complex due to challenges in finding appropriate instruments?

<p>Instrumental variable estimation (B)</p> Signup and view all the answers

What is one characteristic of the dependent variable yi as mentioned in the statements?

<p>It exists only in a binary form. (C)</p> Signup and view all the answers

In the context of the content, which approach is suggested for understanding complex market structures?

<p>Structural approach (A)</p> Signup and view all the answers

What does the representation of $Y_i$ in the CEF-decomposition property imply about the relationship between $Y_i$ and $u_i$?

<p>$u_i$ captures the random component of $Y_i$ (C)</p> Signup and view all the answers

Which statement accurately describes the mean-independence of $u_i$ from $X_i$?

<p>The expected value of $u_i$ remains constant regardless of $X_i$ (B)</p> Signup and view all the answers

How is the law of iterated expectations utilized in the context of the CEF-decomposition property?

<p>To establish that $E[h(X_i) · u_i] = 0$ (C)</p> Signup and view all the answers

What is the implication of the error term $u_i$ being defined as $Y_i - E[Y_i |X_i]$?

<p>It suggests that $u_i$ carries no predictive power for $Y_i$ (C)</p> Signup and view all the answers

In the regression model where $p_i = eta_0 + eta_1 N_i + eta_2 z_i + v_i$, what does adding $z_i$ likely introduce?

<p>The risk of omitted variable bias if $z_i$ is not included (B)</p> Signup and view all the answers

When considering the direction of the bias when $z_i$ represents cost factors, what is a likely outcome?

<p>The estimated $eta_1$ will be downward biased (C)</p> Signup and view all the answers

What is the expected result if $z_i$ captures demand factors in the pricing model?

<p>Estimation of $eta_1$ would likely be biased in the positive direction (C)</p> Signup and view all the answers

Which expression correctly summarizes the law of iterated expectations relevant to $u_i$?

<p>$E[u_i | X_i] = 0$ (A)</p> Signup and view all the answers

What is one method to approximate a nonlinear regression function?

<p>Applying a quadratic or cubic polynomial (B)</p> Signup and view all the answers

How can logarithmic transformations be beneficial in regression analysis?

<p>They provide a percentage interpretation of the coefficients (A)</p> Signup and view all the answers

In which type of relationship might a nonlinear regression function be more appropriate?

<p>When the relationship between variables is not linear in form (D)</p> Signup and view all the answers

What is the form of the polynomial representation in a multiple regression model?

<p>yi = β0 + β1 xi + β2 xi^2 + ... + βr xir + ui (A)</p> Signup and view all the answers

What characteristic does a nonlinear function exhibit in relation to independent variables?

<p>The impact changes at different values of the independent variable (C)</p> Signup and view all the answers

What percentage change is associated with a β1 value of -0.15?

<p>-13.9% (A)</p> Signup and view all the answers

In a log-log regression model, what does a 1% increase in income represent in terms of test score change?

<p>0.0554 (A)</p> Signup and view all the answers

What is the potential issue when including multiple dummy variables in a regression model?

<p>2008 (C)</p> Signup and view all the answers

If one dummy variable is omitted to avoid multicollinearity, how is the interpretation of the coefficients adjusted?

<p>1 (A)</p> Signup and view all the answers

What mathematical transformation is used to express the change in y relative to a change in log points?

<p>100 (A), 100 (B), 100 (C), 100 (D)</p> Signup and view all the answers

How does a dummy variable function in econometrics?

<p>1 (C)</p> Signup and view all the answers

How is the percentage change calculated when β's are larger than small values?

<p>100 (A), 100 (B), 100 (C), 100 (D)</p> Signup and view all the answers

What happens if you include all dummy variables and a constant in a model?

<p>You may create perfect multicollinearity. (A)</p> Signup and view all the answers

What does the negative coefficient difference for university education between women and men represent?

<p>The wage gap increases with higher education levels for both genders. (A)</p> Signup and view all the answers

What can be inferred from the noted professional experience coefficients for both genders?

<p>Both genders equally benefit from increased professional experience in wages. (B)</p> Signup and view all the answers

What implication does the result indicate about the 'ratio of women to men' in a firm?

<p>Increasing the ratio of women in a firm negatively affects wages. (D)</p> Signup and view all the answers

What is indicated by the adjusted R-square values for women and men?

<p>Men's wage predictors are more reliable than those of women. (D)</p> Signup and view all the answers

What does the constant value represent in this analysis for women and men?

<p>The baseline wage level for each gender without any adjustments. (D)</p> Signup and view all the answers

What does the significant difference in partnership status impact suggest about wage dynamics?

<p>Partnership leads to a wage increase only for men. (A)</p> Signup and view all the answers

What inference can be drawn from the squared terms associated with professional experience?

<p>They suggest diminishing returns on wages as experience increases. (C)</p> Signup and view all the answers

What does the coefficient for women in the wage of women/wage of men ratio suggest?

<p>Women earn less than men, equating to a detrimental ratio. (C)</p> Signup and view all the answers

What can be inferred from a linear-log model where a 1% increase in income leads to a 0.36 points increase in test score?

<p>Test scores are positively influenced by increases in income. (C)</p> Signup and view all the answers

In the log-log population regression function, what is the interpretation of the coefficient β1?

<p>It signifies the elasticity of y with respect to x. (A)</p> Signup and view all the answers

What does the model ln(y) = β0 + β1 ln(x) imply when considering small changes?

<p>A small change in x results in a proportionate change in y based on the elasticity. (D)</p> Signup and view all the answers

How does the log-linear model express the relationship between changes in x and y?

<p>∆y is directly proportional to ∆x. (B)</p> Signup and view all the answers

In the log-log model, how is the percentage change in y calculated when x changes by 1%?

<p>It is equal to β1 times the percentage change in x. (A)</p> Signup and view all the answers

What must hold true for the coefficient β1 in the context of small changes in the log-linear model?

<p>It is equivalent to the ratio of ∆y to ∆x. (A)</p> Signup and view all the answers

What does the expression ln(yi) = β0 + β1xi + ui indicate in regression analysis?

<p>The model assesses the impact of changes in x on the logarithmic scale of y. (B)</p> Signup and view all the answers

How can the interpretation of β1 in a linear-log model be summarized?

<p>A unit increase in x leads to a percentage change in y. (D)</p> Signup and view all the answers

What does the hypothesis test H0: population coefficients on Income2 = 0 and Income3 = 0 assess?

<p>The linearity of the population regression function (C)</p> Signup and view all the answers

In the context of polynomial regression, what does a quadratic specification imply?

<p>Both linear and squared terms of the independent variable are included (A)</p> Signup and view all the answers

What is a necessary step when interpreting the individual coefficients in polynomial regression?

<p>Assess the expected change in the outcome variable across all values of the independent variable (D)</p> Signup and view all the answers

What statistical tests can be employed to assess the degree of polynomial fit in regression analysis?

<p>t-tests and F-tests (A)</p> Signup and view all the answers

Which of the following is true regarding the interpretation of coefficients in polynomial regression?

<p>Coefficients indicate varying marginal effects depending on the level of the independent variable (B)</p> Signup and view all the answers

Why is it important to visually plot predicted values in polynomial regression analysis?

<p>To understand potential non-linear relationships (B)</p> Signup and view all the answers

What is the consequence of rejecting the null hypothesis of linearity in a regression model?

<p>The model should be adjusted to include higher-degree polynomial terms (B)</p> Signup and view all the answers

Which of the following represents a component of estimating polynomial regression using OLS?

<p>Creating new regressors based on powers of existing variables (B)</p> Signup and view all the answers

In the linear-log model, how is the change in dependent variable y related to changes in independent variable x?

<p>The change in y is the product of β1 and the percentage change in x. (D)</p> Signup and view all the answers

Which statement correctly describes the interpretation of the slope coefficient β1 in the log-log regression model?

<p>It shows the percentage change in y for each percentage change in x. (A)</p> Signup and view all the answers

What formula is used to approximate small changes in y in the context of the linear-log function?

<p>∆y = β1 ∆x/x (B)</p> Signup and view all the answers

When interpreting the log-linear regression model, what does a 1% increase in x indicate regarding y?

<p>y increases by 0.01 × β1. (D)</p> Signup and view all the answers

Which regression specification directly utilizes the natural logarithm of both the dependent and independent variables?

<p>log-log (A)</p> Signup and view all the answers

What concept is primarily used to derive the interpretation of the slope coefficient in econometric models?

<p>The general 'before and after' rule. (C)</p> Signup and view all the answers

In a log-linear model, how does a change in y relate to a change in the independent variable x?

<p>Percentage change in y equals β1 times the percentage change in x. (A)</p> Signup and view all the answers

What simplifies the expression ln(x + ∆x) - ln(x) for small changes in x?

<p>∆x/x (D)</p> Signup and view all the answers

Which characteristic is unique to the log-linear specification compared to other forms?

<p>It allows for percentage interpretations of changes in y. (D)</p> Signup and view all the answers

What happens to the F statistic in the case of a model with only an intercept?

<p>It is equal to 0. (D)</p> Signup and view all the answers

In restricted regression, how is the F statistic calculated?

<p>R-squared divided by the product of (1 - R-squared) and (N - K - 1). (B)</p> Signup and view all the answers

What is the relationship between F statistic and t statistic when testing only one exclusion?

<p>F is equal to the square of t. (A)</p> Signup and view all the answers

What is a potential challenge when imposing linear restrictions in regression analysis?

<p>The need to redefine variables. (D)</p> Signup and view all the answers

Which of the following correctly summarizes the conditions necessary for a restricted regression analysis?

<p>Both restricted and unrestricted models must be estimated first. (D)</p> Signup and view all the answers

What does the statistic $z$ represent in the hypothesis testing framework provided?

<p>The standardized test statistic (C)</p> Signup and view all the answers

Under the null hypothesis, what distribution does the statistic $z$ follow?

<p>Normal distribution (A)</p> Signup and view all the answers

In the F-test formula, what does the term $SSR_r$ represent?

<p>Sum of squared residuals for the restricted model (D)</p> Signup and view all the answers

What is the role of the term $(SSR_r - SSR_{ur})/q$ in the modified F-test statistic?

<p>It compares the explanatory power of two models (A)</p> Signup and view all the answers

What is the significance of the degree of freedom term $N - K - 1$ in regression analysis?

<p>It accounts for the number of predictors in the model (A)</p> Signup and view all the answers

What does the notation $H_0: eta_j = a_j$ imply?

<p>The coefficient equals a specified value (D)</p> Signup and view all the answers

What is the implications of the relationship $z|H_0 ext{ ~ } N(0, ext{Var}(z))$?

<p>It describes the distribution of the test statistic under the null hypothesis (A)</p> Signup and view all the answers

In the modified F-test for regression analysis, what does the term $SSR_{ur}$ represent?

<p>Sum of squared residuals for the unrestricted model (D)</p> Signup and view all the answers

What does the coefficient β2 represent in the proposed model?

<p>The impact of age on wage (A)</p> Signup and view all the answers

Which variable is not included in the interaction terms of the regression model?

<p>wage (B)</p> Signup and view all the answers

What is the significance of the term (years · tenure) in the model?

<p>It captures the interaction effect between years of education and tenure. (C)</p> Signup and view all the answers

The log-linear model presented primarily indicates that changes in which variables will most influence wage?

<p>Education, age, and gender interactions (D)</p> Signup and view all the answers

What does a negative coefficient in the estimates for age suggest about its relationship with wage?

<p>As age increases, wage decreases when controlling for other variables (C)</p> Signup and view all the answers

What does the model's term (female · years) capture in the context of wage determination?

<p>The differential impact of education on wages for female workers (B)</p> Signup and view all the answers

In the context of this model, what does SSR and SST represent?

<p>Sum of squared residuals and total sum of squares (D)</p> Signup and view all the answers

What do the standard errors in parentheses next to coefficients indicate?

<p>How much the coefficient estimates vary across observations (D)</p> Signup and view all the answers

What does the F statistic measure in the context of econometric models?

<p>The relative increase in SSR when moving from the restricted to the unrestricted model (A)</p> Signup and view all the answers

Why can the F statistic never be negative?

<p>SSR from the restricted model can not be less than SSR from the unrestricted model (A)</p> Signup and view all the answers

What are the degrees of freedom used in the F distribution for testing the significance of models?

<p>q and N - K - 1 (B)</p> Signup and view all the answers

When should the null hypothesis H0 be rejected concerning the F statistic?

<p>If F is greater than the critical value c (B)</p> Signup and view all the answers

How can the R² form of the F statistic be expressed?

<p>F = ((1 - R²r) / (1 - R²ur)) / (N - K - 1) (D)</p> Signup and view all the answers

What does the parameter 'q' represent in the context of the F statistic?

<p>The number of restrictions imposed by the restricted model (D)</p> Signup and view all the answers

What is the appropriate condition to test whether all coefficients in the model are jointly significant?

<p>H0: β1 = β2 = ... = βn = 0 (C)</p> Signup and view all the answers

What does SSR stand for in the context of regression analysis?

<p>Sum of Squared Residuals (A)</p> Signup and view all the answers

What does the notation SSRr represent in econometrics?

<p>Sum of square residuals for restricted regression (D)</p> Signup and view all the answers

Which of the following inequalities regarding SSR holds true?

<p>SSRr ≥ SSRur (D)</p> Signup and view all the answers

In the context of hypothesis testing, what do H0 and H1 typically represent?

<p>Null hypothesis and alternate hypothesis, respectively (C)</p> Signup and view all the answers

What is the meaning of the matrix R in hypothesis testing?

<p>It is a known matrix of constants defining linear restrictions (B)</p> Signup and view all the answers

Which of the following is an example of a joint hypothesis?

<p>β2 + β3 = 1 (C)</p> Signup and view all the answers

What does the term SSRr - SSRur denote in econometric analysis?

<p>The increase in variability explained by including additional parameters (A)</p> Signup and view all the answers

Which hypothesis is represented by H0: Rβ̃ = r?

<p>The coefficients meet specific linear restrictions (A)</p> Signup and view all the answers

What implication does the statement 'Rβ̃ = r' convey in hypothesis testing?

<p>The restrictions imposed on coefficients are met (B)</p> Signup and view all the answers

What is the marginal effect of age on ln(wage) for male workers?

<p>$β2 + 2β3 , age$ (C)</p> Signup and view all the answers

How is the expected percentage change in wage interpreted when considering age for male workers?

<p>It is an expected change of 100($β1 + 2β3 , age$)%. (A)</p> Signup and view all the answers

What is the impact of the term $2(β3 + β9) , age$ in the marginal effect equation for female workers?

<p>It modifies the effect of age on wage based on tenure. (B)</p> Signup and view all the answers

Which coefficients are involved in calculating the total effect of age on ln(wage) for females?

<p>$β2 + β8 + 2β3 + β9$ (B)</p> Signup and view all the answers

Which option accurately reflects the term β̂2 + β̂8 in the context of female workers' wages?

<p>It is the marginal effect of age on ln(wage). (A)</p> Signup and view all the answers

What does the coefficient β̂1 represent in the given equation for females?

<p>The constant term affecting ln(wage) for females. (A)</p> Signup and view all the answers

What is the theoretical interpretation of the marginal effect of age for female workers?

<p>It reflects the interaction of age and other variables on wages. (C)</p> Signup and view all the answers

Which term reflects the quadratic nature of age's impact on wage in the equations provided?

<p>$2(β3 + β9 , age)$ (B)</p> Signup and view all the answers

What occurs when the t-statistic is in the interval (-c, c) during hypothesis testing?

<p>The coefficient βj is not significantly different from zero. (B)</p> Signup and view all the answers

What indicates that a coefficient βj is statistically significant at the α% level?

<p>The t-statistic is greater than c or less than -c. (C)</p> Signup and view all the answers

In a one-sided hypothesis test where H1 states βj < 0, what criteria must be met to reject the null hypothesis?

<p>t-statistic is less than -c. (D)</p> Signup and view all the answers

What is typically assumed about the alternative hypothesis in hypothesis testing?

<p>It is a two-sided hypothesis. (A)</p> Signup and view all the answers

What does it mean when an estimator is said to be statistically insignificant?

<p>The null hypothesis fails to be rejected. (C)</p> Signup and view all the answers

What is the purpose of the Chow test in econometrics?

<p>To determine if there are structural breaks in regression models (A)</p> Signup and view all the answers

In the expression for the Chow test, what does the term $RSSR$ represent?

<p>Residual sum of squares for the restricted model (A)</p> Signup and view all the answers

What hypothesis does $H0: β11 = β21 ∧ β12 = β22$ imply in the context of the Chow test?

<p>The coefficients for both groups are equal (A)</p> Signup and view all the answers

Which of the following describes the variables $N1$ and $N2$ in the Chow test?

<p>They indicate the number of observations from each sample used in regression analysis. (A)</p> Signup and view all the answers

What does the term $K$ represent in the context of the Chow test formula?

<p>The number of independent variables in the regression (D)</p> Signup and view all the answers

What is the theoretical marginal effect of age on ln(wage) for male workers expressed in terms of the coefficients from the regression?

<p>0.0798 (A)</p> Signup and view all the answers

In a hypothesis test for the marginal effect of age on ln(wage) between male and female workers, which of the following represents the null hypothesis if the regression equations are assumed to be identical?

<p>The marginal effect of age for males is equal to the marginal effect of age for females. (D)</p> Signup and view all the answers

What would the significance level of 5% imply when testing the hypothesis about the equality of marginal effects of age between male and female workers?

<p>You should reject the null hypothesis if the p-value is less than 0.05. (A)</p> Signup and view all the answers

What is the sum of squared residuals for the restricted OLS estimation in the provided analysis?

<p>93.1805 (A)</p> Signup and view all the answers

What does a decrease in the sum of squared residuals (SSRR) indicate when comparing models in this context?

<p>The last model fits the data better than the previous model. (A)</p> Signup and view all the answers

What distribution does the test statistic z follow under the null hypothesis?

<p>N(0, σ² R(X′ X)⁻¹ R′) (A)</p> Signup and view all the answers

What does the term SSRr represent in the context of the F-test?

<p>Sum of Squared Residuals under the restricted model (C)</p> Signup and view all the answers

In the context of the univariate test statistic formula, what does 'n' represent?

<p>The value of the test statistic (B)</p> Signup and view all the answers

What is the modified statistic when the original F-test is infeasible?

<p>(SSRr - SSRur)/q (C)</p> Signup and view all the answers

What does the notation v/(N - K - 1) indicate in the context of the F-test?

<p>Degrees of freedom from the residuals (C)</p> Signup and view all the answers

What does the symbol β̂ represent in the equation for the test statistic z?

<p>The OLS estimator (B)</p> Signup and view all the answers

Under the null hypothesis, how does the variance of z behave?

<p>It is σ² R(X′ X)⁻¹ R′ (B)</p> Signup and view all the answers

What distribution does the modified statistic follow if n and v are independent?

<p>F-distribution (A)</p> Signup and view all the answers

What is the purpose of a restricted model in multiple linear restrictions testing?

<p>To impose constraints on certain parameters (C)</p> Signup and view all the answers

What is the null hypothesis for jointly testing multiple linear restrictions?

<p>H0: All specified parameters equal zero (A)</p> Signup and view all the answers

What approach is suggested for testing linear combinations of parameters?

<p>Rearranging the model and computing new standard errors (D)</p> Signup and view all the answers

Which statement about testing joint hypotheses is true?

<p>It is sufficient if one parameter is not equal to zero to reject the null (D)</p> Signup and view all the answers

How can the model regarding campaign expenditures and voting outcomes be rearranged?

<p>By substituting a linear combination of parameters into the equation (D)</p> Signup and view all the answers

What is necessary to conduct a test for multiple linear restrictions?

<p>Estimate the restricted model without some variables (B)</p> Signup and view all the answers

When hypothesizing about a single linear combination of parameters, which equation might be tested?

<p>H0: β1 + β2 = 0 (B)</p> Signup and view all the answers

In the context of testing exclusion restrictions, what is the proper formulation of the null hypothesis?

<p>H0: βk = βk−q+1 = ... = βk = 0 (D)</p> Signup and view all the answers

What does the F statistic measure in regression analysis?

<p>The relative increase in SSR when moving from the unrestricted to the restricted model (A)</p> Signup and view all the answers

What condition needs to be satisfied to reject the null hypothesis using the F statistic?

<p>F must be greater than the critical value (C)</p> Signup and view all the answers

What is represented by the term 'q' in the F statistic formula?

<p>Number of restrictions imposed on the model (C)</p> Signup and view all the answers

In the context of the F statistic, what do N and K represent?

<p>N is the total number of observations, K is the number of parameters estimated (A)</p> Signup and view all the answers

Which distribution does the F statistic follow?

<p>F-distribution with parameters q and N-K-1 (D)</p> Signup and view all the answers

How is the F statistic related to the R-squared values of the regression models?

<p>It calculates the ratio of the R-squared values, adjusted for degrees of freedom (A)</p> Signup and view all the answers

What is the primary purpose of using the F statistic in econometrics?

<p>To determine if the overall model has statistical significance (D)</p> Signup and view all the answers

What happens to the F statistic if the restricted model does not provide a significant increase in SSR?

<p>The null hypothesis cannot be rejected (D)</p> Signup and view all the answers

What is the purpose of the Lagrange Multiplier (LM) statistic in econometric analysis?

<p>To test multiple exclusion restrictions simultaneously (B)</p> Signup and view all the answers

How does the asymptotic standard error behave as the sample size (N) increases?

<p>It shrinks at a rate proportional to the inverse of N (A)</p> Signup and view all the answers

In the context of LM statistic calculation, what is meant by an 'auxiliary regression'?

<p>A regression run on the residuals to test a specific hypothesis (C)</p> Signup and view all the answers

What does the term se $\hat{β}_j$ represent in the formula provided for asymptotic standard errors?

<p>The standard error of the estimated coefficient (A)</p> Signup and view all the answers

In the context of regression models, what is the primary characteristic of the constant cj in the asymptotic standard error formula?

<p>It does not depend on the sample size (C)</p> Signup and view all the answers

What distribution does the LM statistic follow?

<p>$ ext{χ}^2_q$ distribution (A)</p> Signup and view all the answers

Under the Gauss-Markov assumptions, which of the following best describes the efficiency of OLS estimators?

<p>OLS estimators are asymptotically efficient (D)</p> Signup and view all the answers

What condition must hold for OLS estimators to be considered asymptotically efficient?

<p>The error term must be homoskedastic (C)</p> Signup and view all the answers

What happens to the results from the F test and the LM test in large samples?

<p>They should yield similar results (B)</p> Signup and view all the answers

What is a key distinction between the LM test and the F test regarding model exclusion?

<p>The LM test and F test are not identical (C)</p> Signup and view all the answers

What distribution does $etâ_j - eta_j$ approximately follow under large sample conditions?

<p>t-distribution (D)</p> Signup and view all the answers

What condition must still hold true when using t-tests for large samples?

<p>Homoskedasticity (A)</p> Signup and view all the answers

In the context of the estimators, what does $Q^{-1}$ represent?

<p>The limit of the covariance matrix (D)</p> Signup and view all the answers

What is the significance of the Central Limit Theorem in econometric analysis?

<p>It underpins the asymptotic normality of estimators. (B)</p> Signup and view all the answers

Which aspect is NOT necessary for the asymptotic properties discussed?

<p>Requirement to derive asymptotic properties (A)</p> Signup and view all the answers

What does it mean for an estimator to be consistent?

<p>Its mean approaches the parameter value as the sample size increases. (C)</p> Signup and view all the answers

Under which assumptions is the OLS estimator consistent for all parameters βj?

<p>Under the Gauss-Markov assumptions MLR.1-MLR.5. (D)</p> Signup and view all the answers

What condition is necessary for the probability limit to establish consistency?

<p>The second moments of the explanatory variables must be finite. (C)</p> Signup and view all the answers

What does the 'plim' notation signify in econometric analysis?

<p>It denotes the probability limit of an estimator as the sample size approaches infinity. (B)</p> Signup and view all the answers

Which of the following is NOT a characteristic of a consistent estimator?

<p>The variance does not depend on sample size. (C)</p> Signup and view all the answers

What is implied by saying that an estimator is 'BLUE' under the Gauss-Markov theorem?

<p>It is the best linear unbiased estimator among all linear estimators. (C)</p> Signup and view all the answers

Which mathematical approach can be used to prove the consistency of the OLS estimator?

<p>The law of large numbers. (A)</p> Signup and view all the answers

What happens to the variance of an estimator as the sample size approaches infinity?

<p>It tends toward zero. (D)</p> Signup and view all the answers

What does asymptotic normality imply about OLS estimators as the sample size approaches infinity?

<p>They are distributed according to a normal distribution. (B)</p> Signup and view all the answers

Which equation represents the standardized sample mean under the Central Limit Theorem?

<p>$Z = \frac{Y - \mu}{\sigma \sqrt{n}}$ (C)</p> Signup and view all the answers

In the context of OLS estimators and asymptotic normality, what is denoted by $se(\hat{\beta_j})$?

<p>The standard error of the estimated coefficient. (C)</p> Signup and view all the answers

What is the implication of the asymptotic distribution of the OLS estimator according to the Gauss-Markov assumptions?

<p>It implies the estimator converges to the true parameter value in distribution. (C)</p> Signup and view all the answers

What does the notation $\hat{\beta_j} - \beta_j \sim N(0, 1)$ signify in terms of asymptotic normality?

<p>The difference between the estimator and the true parameter is asymptotically normal. (B)</p> Signup and view all the answers

What does the symbol $\text{plim}$ represent in the context of econometrics?

<p>Probability limit (C)</p> Signup and view all the answers

What condition must hold for $\text{Cov}(x_1, u)$ to equal zero?

<p>The error term must be independent of the independent variable (B)</p> Signup and view all the answers

In the context of the OLS estimator, what happens to the variance of $w$ if $E(u_i | x_i) = 0$?

<p>The variance is zero (D)</p> Signup and view all the answers

What is represented by the term $\text{Var}(w) = E(\text{Var}(w|X)) + \text{Var}[E(w|X)]$?

<p>Total variance decomposition (A)</p> Signup and view all the answers

In the formula $\hat{\beta} = (X'X)^{-1}X'y$, what do the symbols $X$ and $y$ represent?

<p>Independent and dependent variables respectively (A)</p> Signup and view all the answers

Which element is required to ensure the consistency of the OLS estimator?

<p>Large sample size approaching infinity (D)</p> Signup and view all the answers

What does the term $\text{Var}[E(w|X)]$ indicate in the variance formula?

<p>Explained variance (B)</p> Signup and view all the answers

In proving consistency of the OLS estimator, what effect does adding more data points have?

<p>Decreases the bias (D)</p> Signup and view all the answers

What does the notation $E[x_i' u] = 0$ imply?

<p>The errors are orthogonal to the regressors (D)</p> Signup and view all the answers

In the equation $plim \hat{\beta} = \hat{\beta} + plim(\frac{1}{N}(X'X)^{-1})X'u$, what does $plim(\frac{1}{N}(X'X)^{-1})X'u$ signify?

<p>Convergence of the error term to zero (C)</p> Signup and view all the answers

What does E(y | x = 1) represent in the context of a randomized controlled experiment?

<p>The expected outcome for individuals receiving the vaccine (C)</p> Signup and view all the answers

Which of the following best describes a potential issue with internal validity in experiments?

<p>Random assignment is not truly random (A)</p> Signup and view all the answers

How can quasi-experiments be distinguished from true experiments?

<p>Quasi-experiments utilize a source of randomization that is not explicitly assigned (D)</p> Signup and view all the answers

What does the average treatment effect (ATE) represent in econometrics?

<p>The population mean of the individual treatment effect (A)</p> Signup and view all the answers

How does including school fixed effects affect the OLS estimator of the coefficient on years of experience?

<p>It makes the estimator unbiased (B)</p> Signup and view all the answers

What is the primary advantage of conducting randomized controlled experiments over observational studies?

<p>They provide a clearer estimation of causal relationships (B)</p> Signup and view all the answers

What does the term 'treatment effect' refer to in experimental research?

<p>The resultant change in the dependent variable due to intervention (D)</p> Signup and view all the answers

In what way do actual experiments pose threats to internal validity?

<p>They often face issues like incomplete treatment (C)</p> Signup and view all the answers

Why are actual randomized controlled experiments considered rare, despite their importance?

<p>They require complex and time-consuming designs (B)</p> Signup and view all the answers

What is the main finding regarding the effects measured in the Tennessee Class Size Experiment?

<p>The effects were small and similar to gender differences (C)</p> Signup and view all the answers

In the context of a randomized controlled trial, what does a placebo serve to achieve?

<p>To control for psychological effects on participants (B)</p> Signup and view all the answers

Which method can be used to address threats to internal validity in experiments?

<p>TSLS with initial assignment as an instrument (D)</p> Signup and view all the answers

What role does random assignment play in a randomized controlled experiment?

<p>It eliminates the potential for selection bias (D)</p> Signup and view all the answers

What is a common characteristic of an ideal randomized controlled experiment according to the summarization?

<p>They provide unbiased estimates of treatment effects (C)</p> Signup and view all the answers

What are the effects of adding control variables in multiple regression analysis?

<p>They can help mitigate the effects of confounding variables (C)</p> Signup and view all the answers

How is the treatment effect estimated in an ideal randomized controlled experiment?

<p>By assessing the difference between treated and untreated groups (C)</p> Signup and view all the answers

What does vaccine effectiveness imply in terms of the expected health measure between vaccinated and unvaccinated individuals?

<p>E(y | x = 1) &gt; E(y | x = 0) (D)</p> Signup and view all the answers

Which Gauss-Markov assumption is violated if everyone in a study receives the vaccine?

<p>SLR.3 Sample variation (B)</p> Signup and view all the answers

If individuals who are very sick are more likely to choose to take the vaccine, what assumption is compromised?

<p>SLR.2 Random sampling (A)</p> Signup and view all the answers

What can be done to ensure valid causal inference in vaccine studies?

<p>Randomize assignment in an experiment (C)</p> Signup and view all the answers

What is the result of measuring $E(y | x = 1)$ and $E(y | x = 0)$ in scenario 2, where only sick individuals choose to take the vaccine?

<p>$E(y | x = 1)$ is lower than $E(y | x = 0)$ (A)</p> Signup and view all the answers

In a randomized controlled experiment, what is the primary benefit of random assignment?

<p>Ensures equal distribution of covariates across treatment groups (C)</p> Signup and view all the answers

Which of these represents a methodological flaw in estimating the effect of vaccines if all participants are vaccinated?

<p>Neglecting the control group (C)</p> Signup and view all the answers

If a study indicates $E(y | x = 1) o -0.85$ and $E(y | x = 0) o -0.04$, what does this suggest about vaccine treatment?

<p>The vaccine appears harmful (C)</p> Signup and view all the answers

What does the difference-in-difference (DID) estimator primarily estimate?

<p>The causal effect of a treatment by comparing pre- and post-treatment outcomes between groups (C)</p> Signup and view all the answers

In the context of interactions between independent variables, how might a class size reduction be more effective?

<p>When classes have a higher percentage of English learners (D)</p> Signup and view all the answers

Which factors might influence how age or potential experience affects wages?

<p>The level of education and gender of the individuals (C)</p> Signup and view all the answers

What is a key consideration when modeling interactions between two continuous variables?

<p>Their relationship can change based on external factors (B)</p> Signup and view all the answers

Which of the following constitutes a classical example of difference-in-difference methodology?

<p>Card and Krueger's 1994 study on minimum wage policies (B)</p> Signup and view all the answers

What does the term β3 represent in the binary-continuous interaction model?

<p>The change in the effect of X when D = 1 (C)</p> Signup and view all the answers

In the regression model yi = β0 + β1 Di + β2 xi + β3 (Di × xi ) + ui, what does Di signify?

<p>A binary dummy variable indicating group membership (B)</p> Signup and view all the answers

What distinguishes the equations for the 'D = 0' and 'D = 1' groups in the binary-continuous interaction model?

<p>There are different slopes and intercepts for both groups (B)</p> Signup and view all the answers

How does the change in X influence the dependent variable Y according to the binary-continuous interaction model?

<p>The impact of X on Y varies based on the value of D (A)</p> Signup and view all the answers

What is the significance of the term (Di × xi) in the regression equation?

<p>It represents the interaction between a binary and a continuous variable (D)</p> Signup and view all the answers

In the linear regression equation, what is the result of subtracting the D = 0 group equation from the D = 1 group equation?

<p>It results in a differential equation representing ∆y (D)</p> Signup and view all the answers

What does the notation $y + ∆y = β0 + β1 D + β2 (x + ∆x) + β3 [D × (x + ∆x)]$ illustrate?

<p>The effect of both existing and new values of x on y (D)</p> Signup and view all the answers

Which statement best describes the general rule for comparing various cases in the model?

<p>The impact of the independent variable varies based on group membership (D)</p> Signup and view all the answers

What is the purpose of including the interaction term D1i × D2i in the regression model?

<p>To allow the effect of D1 to depend on the value of D2. (B)</p> Signup and view all the answers

How is the total effect of D1i on the expected value E(yi) influenced by D2i according to the regression specification?

<p>The effect of D1i varies based on the interaction between D1i and D2i. (B)</p> Signup and view all the answers

In the equation E(yi | D1i = 1, D2i = d2) - E(yi | D1i = 0, D2i = d2), what does the term β3 represent?

<p>The increment to the effect of D1 when D2 = 1. (B)</p> Signup and view all the answers

What does the coefficient β1 represent in the context of the regression model?

<p>The effect of changing D1 from 0 to 1 when D2 is held constant. (B)</p> Signup and view all the answers

What happens to the expected value of yi when D1i = 1 and D2i = 0?

<p>It reduces to β0 + β1. (B)</p> Signup and view all the answers

What is the significance of R-squared in the context of regression analysis presented?

<p>It measures the proportion of variance in the dependent variable explained by the independent variables. (D)</p> Signup and view all the answers

In which situation does the coefficient β3 equal zero?

<p>When D2 takes a constant value of 0. (D)</p> Signup and view all the answers

In the provided model, what does the residual term (ui) account for?

<p>The random error not explained by the model. (B)</p> Signup and view all the answers

What is the purpose of the 'no anticipation' assumption in difference-in-difference estimators?

<p>To suggest that groups do not change behavior before treatment (B)</p> Signup and view all the answers

In the regression output provided, what does the coefficient for 'bachelor' signify?

<p>An increase in average hourly earnings (C)</p> Signup and view all the answers

What does the term 'parallel trends' assume in the context of difference-in-differences analysis?

<p>Control and treatment groups experience similar trends over time before treatment (A)</p> Signup and view all the answers

What does an R-squared value of 0.1911 indicate about the regression model's fit?

<p>The model explains 19.11% of the variance in the dependent variable (A)</p> Signup and view all the answers

What does the coefficient for 'age_female' reflect regarding the relationship between age and earnings?

<p>An increase in age is associated with a slight decrease in earnings for females (C)</p> Signup and view all the answers

What is the significance of the Residual sum of squares in the regression output?

<p>It quantifies unexplained variation in the dependent variable (D)</p> Signup and view all the answers

What does the 't' statistic represent in the regression output for 'bachelor_female'?

<p>The statistical significance of the coefficient (C)</p> Signup and view all the answers

In the context of the study by Card and Krueger, what does the methodology aim to achieve?

<p>To differentiate treatment effects from seasonal employment effects (C)</p> Signup and view all the answers

What does the DID estimate in Row 3 Column (iii) represent in the context of the data?

<p>The average treatment effect on FTE employment between NJ and PA (A)</p> Signup and view all the answers

In the context of the data provided, what were the FTE employment figures for PA and NJ before any treatment?

<p>23.33 for PA and 20.44 for NJ (C)</p> Signup and view all the answers

What is indicated by the negative difference of -2.89 for FTE employment before in Column (iii)?

<p>PA had 2.89 more FTE employed than NJ (C)</p> Signup and view all the answers

What does the change in mean FTE employment recorded as 2.76 in Column (iii) suggest?

<p>Employment increased by 2.76 for NJ relative to PA post-treatment (D)</p> Signup and view all the answers

What statistical significance do the parentheses represent next to the FTE employment figures?

<p>Standard errors associated with the FTE employment estimates (B)</p> Signup and view all the answers

Given the binary treatment where Di = 0 for PA and Di = 1 for NJ, what does Di signify?

<p>The presence or absence of a treatment effect for each state (D)</p> Signup and view all the answers

Assessing the FTE employment after, what does the value of -0.14 indicate in the context of this analysis?

<p>A negligible change in employment for NJ compared to PA post-treatment (C)</p> Signup and view all the answers

How is the FTE employment of 21.03 for NJ characterized in the context of the analysis?

<p>The FTE employment level for NJ after implementing the treatment (C)</p> Signup and view all the answers

What does the DID estimator specifically provide an estimate of?

<p>The average treatment effect on the treated (ATT) (A)</p> Signup and view all the answers

Which assumption states that outcomes are not affected by treatment prior to its implementation?

<p>No anticipation assumption (D)</p> Signup and view all the answers

What does the parallel trends assumption imply about the trends of treated and untreated groups?

<p>They must be the same before treatment (A)</p> Signup and view all the answers

Under what condition is the DID estimator unbiased for ATT?

<p>When both no anticipation and parallel trends assumptions hold (B)</p> Signup and view all the answers

What is a common violation of the no anticipation assumption?

<p>Businesses preparing for layoffs before wage increases (B)</p> Signup and view all the answers

What does the notation $ȳt=0,D=1$ represent in the context of the DID estimator?

<p>The average outcome for the treated group before treatment (C)</p> Signup and view all the answers

Which of the following scenarios illustrates a potential violation of the parallel trends assumption?

<p>Sustained economic improvement in NJ compared to PA (B)</p> Signup and view all the answers

Which regression model is commonly used to implement the DID approach?

<p>Two-way fixed effects (TWFE) model (B)</p> Signup and view all the answers

What might imply non-zero selection bias even when parallel trends hold?

<p>Differences in average outcomes at baseline (A)</p> Signup and view all the answers

What does the variable $z_{it}$ represent in the regression formula for DID?

<p>The treatment group indicator (A)</p> Signup and view all the answers

Why must assumptions be imposed when estimating the ATT?

<p>To account for unobserved potential outcomes (D)</p> Signup and view all the answers

Which of the following statements accurately describes the relationship between the OLS estimate and the DID estimator?

<p>The OLS estimate is numerically equivalent to the DID estimator (D)</p> Signup and view all the answers

In the context of employing the DID method, what might collapse to group-level data provide?

<p>Simplified analysis with clarified trends (B)</p> Signup and view all the answers

What does a high value of R-squared indicate about a regression model?

<p>The model explains a large portion of the variability in the dependent variable. (A)</p> Signup and view all the answers

In the formula for estimating the variance of the estimator $eta_j$, which factor contributes negatively to the variance?

<p>Linear relationship among the independent variables, $R_j^2$. (B)</p> Signup and view all the answers

What common symptom may indicate the presence of multicollinearity in a regression analysis?

<p>High R-squared value with high standard errors. (B)</p> Signup and view all the answers

What impact does increasing the sample size have on the variance of the slope estimate in regression analysis?

<p>It decreases the variance of the slope estimate. (B)</p> Signup and view all the answers

What is a characteristic of the term $SST_j$ in relation to the variance of the estimator $eta_j$?

<p>It indicates the total sample variation of the independent variable xj. (A)</p> Signup and view all the answers

What does heteroskedasticity indicate in a regression model?

<p>Variable dispersion of error terms at different levels of an independent variable (D)</p> Signup and view all the answers

Which of the following is true about the variance-covariance matrix of the error term in the presence of heteroskedasticity?

<p>It is a diagonal matrix with varying variances on the diagonal (A)</p> Signup and view all the answers

How can researchers detect heteroskedasticity in their data?

<p>By plotting the residuals against predicted values (D)</p> Signup and view all the answers

What implication does increasing variance in the error term have on regression estimates?

<p>It makes the estimates less reliable (B)</p> Signup and view all the answers

Why is understanding heteroskedasticity particularly important in econometrics?

<p>It helps improve the efficiency of estimators (B)</p> Signup and view all the answers

What does a high R12 value indicate in the context of regressors x1 and x2?

<p>x1 and x2 are highly correlated (A)</p> Signup and view all the answers

What is the consequence of $R_j$ approaching 1 regarding the variance of the estimator $etâ_1$?

<p>The variance approaches infinity (A)</p> Signup and view all the answers

What is a common threshold for the Variance Inflation Factor (VIF) to indicate serious multicollinearity?

<p>VIF &gt; 10 (A)</p> Signup and view all the answers

How does omitting a relevant variable from a regression model affect the estimates?

<p>It causes omitted variable bias (C)</p> Signup and view all the answers

What does heteroskedasticity imply about the variance of the error term in a regression model?

<p>The variance changes with different values of the independent variables (D)</p> Signup and view all the answers

What factor is suggested for addressing multicollinearity issues effectively?

<p>Including all relevant variables and increasing observations (D)</p> Signup and view all the answers

What assumption does homoskedasticity in MLR.5 refer to regarding the errors in regression analysis?

<p>Errors have a constant variance for all observations (A)</p> Signup and view all the answers

Which of the following characteristics is NOT associated with multicollinearity issues?

<p>Improved model predictions (C)</p> Signup and view all the answers

What is the consequence of using OLS standard errors without adjusting for clustering in a dataset?

<p>The standard errors will be too low, resulting in invalid inference. (C)</p> Signup and view all the answers

What is the primary reason the OLS standard errors become too low when assuming independence within clusters?

<p>The model fails to account for correlated errors within clusters. (C)</p> Signup and view all the answers

What adjustment is necessary to improve the accuracy of standard errors in the presence of clustering?

<p>Inflate the standard errors to reflect within-cluster correlations. (D)</p> Signup and view all the answers

In which of the following scenarios would it be most inappropriate to use OLS standard errors without adjustments?

<p>Data are collected from multiple schools with students correlated within each school. (A)</p> Signup and view all the answers

What command in STATA is used to calculate standard errors that account for clustering?

<p>reg y x, vce(cluster clusterid) (A)</p> Signup and view all the answers

What is the primary purpose of using a robust estimator for the variance covariance matrix?

<p>To ensure unbiased estimates under heteroskedastic conditions (A)</p> Signup and view all the answers

What is represented by the notation $û_i = y_i - x_i β̂$?

<p>The OLS residual for the i-th observation (A)</p> Signup and view all the answers

What characterizes the matrix $Ω̂$ in the context of heteroskedasticity?

<p>It is a diagonal matrix containing the squared residuals (B)</p> Signup and view all the answers

What is used as a standard error for inference when estimating variance under heteroskedasticity?

<p>The square root of a consistent variance estimator (B)</p> Signup and view all the answers

Which of the following components is critical in the computation of the variance of the OLS estimator?

<p>$X'X$ matrix (D)</p> Signup and view all the answers

What does the term $E[uu']$ signify in the context of variance estimation?

<p>The expected value of the covariance of residuals (B)</p> Signup and view all the answers

In the context of the OLS estimator, what does the notation $β̂ − β$ represent?

<p>The difference between estimated and true parameters (C)</p> Signup and view all the answers

Which method is often utilized to adjust for heteroskedasticity in regression analysis?

<p>Using weighted least squares (D)</p> Signup and view all the answers

What is the primary advantage of using weighted least squares (WLS) over transforming an equation for ordinary least squares (OLS)?

<p>WLS can minimize the weighted sum of squares directly. (B)</p> Signup and view all the answers

In the context of feasible generalized least squares (GLS), what is typically assumed about the model for variance?

<p>It can be estimated through a flexible model structure. (D)</p> Signup and view all the answers

What transformation is applied to the squared residuals in WLS compared to the residuals in OLS?

<p>Squared residuals in WLS are weighted by the inverse of their variances. (A)</p> Signup and view all the answers

Which scenario is suitable for using weighted least squares?

<p>When observations are aggregated while the model is based on individual data. (D)</p> Signup and view all the answers

What is Feasible GLS primarily concerned with in cases of unknown heteroskedasticity?

<p>Estimating the structure of variance from data. (A)</p> Signup and view all the answers

What does the term $Var(u|x)$ signify in the context of feasible GLS?

<p>The error variance conditional on the independent variables. (A)</p> Signup and view all the answers

Why might it be challenging to perform the transformation required for OLS when using heteroskedastic data?

<p>The transformation can be excessively complex and non-intuitive. (D)</p> Signup and view all the answers

In the equation for variance $Var (u|x) = σ^2 exp(δ_0 + δ_1 x_1 + ... + δ_K x_K)$, what does the parameter $σ^2$ represent?

<p>The baseline variance of the error term. (A)</p> Signup and view all the answers

What does weakly stationarity require in a time series?

<p>Constant mean and variance over time (D)</p> Signup and view all the answers

How is a stationary time series described if the correlation between terms approaches zero as the time gap increases?

<p>Weakly dependent time series (B)</p> Signup and view all the answers

Which property does white noise exhibit?

<p>Zero covariance for all lags except lag zero (C)</p> Signup and view all the answers

What distinguishes independent and identically distributed (i.i.d.) noise from white noise?

<p>Each sample in i.i.d. has the same distribution (C)</p> Signup and view all the answers

What condition must hold for a covariance stationary process to be classified as weakly dependent?

<p>Correlation diminishes as time gap increases (C)</p> Signup and view all the answers

What does the null hypothesis H0 : β1 = β2 = 0 imply in an AR(2) model?

<p>Past returns do not influence future returns. (A)</p> Signup and view all the answers

In the context of the AR(2) model, which assumption is made about the error term?

<p>The error term has a constant variance given past returns. (B)</p> Signup and view all the answers

What does the R̄² value in the AR(2) model indicate?

<p>It measures the model's overall explanatory power. (C)</p> Signup and view all the answers

Which statement is true regarding the use of an AR(2) model compared to an AR(1) model?

<p>AR(2) can capture correlations from two previous periods. (D)</p> Signup and view all the answers

What is the primary function of the F statistic in the context of the AR(2) model estimation?

<p>To test the significance of the overall model fit. (A)</p> Signup and view all the answers

What is the primary purpose of the Newey-West estimator in econometrics?

<p>To ensure the consistency of covariance estimates in the presence of heteroscedasticity and autocorrelation. (C)</p> Signup and view all the answers

In a trending time series analysis, what does controlling for the trend typically imply?

<p>Directly incorporating the trend within the model to isolate other effects. (D)</p> Signup and view all the answers

What does the truncation lag parameter 'q' represent in the Newey-West estimator?

<p>The number of autocorrelations used to evaluate the dynamics of the OLS residuals. (A)</p> Signup and view all the answers

Which of the following is NOT a model for capturing trends in economic time series?

<p>Random trend model (A)</p> Signup and view all the answers

What does the term 'RSS' refer to in the context of estimation methods?

<p>Residual sum of squares. (D)</p> Signup and view all the answers

What is a common outcome of using the Newey-West HAC estimation method compared to using White's heteroscedasticity consistent estimator?

<p>It remains consistent even with unobserved residual patterns. (A)</p> Signup and view all the answers

What effect does the term '$T - (K + 1)$' have in the Newey-West covariance matrix formula?

<p>It represents the degrees of freedom used in the estimator. (D)</p> Signup and view all the answers

When estimating parameters iteratively, what is the criterion for stopping the iterations?

<p>When the change between estimates is less than a predefined threshold, δ. (D)</p> Signup and view all the answers

What is the null hypothesis (H0) for testing the absence of autocorrelation?

<p>All ρj values are zero. (C)</p> Signup and view all the answers

What condition must be satisfied for a weakly stationary process to be considered ergodic for the expectation?

<p>The average of the series must converge to a finite value. (A)</p> Signup and view all the answers

What distribution does the Box-Pierce statistic QBP(k) follow if the time series yt is i.i.d.?

<p>Chi-squared distribution. (A)</p> Signup and view all the answers

What is a characteristic of a moving average process of order one (MA(1))?

<p>It contains variables that are correlated only one period apart. (A)</p> Signup and view all the answers

Which statistic is proposed by Ljung and Box for testing autocorrelation?

<p>QLB(k). (A)</p> Signup and view all the answers

Which condition is required for the autoregressive process of order one (AR(1)) to be weakly dependent?

<p>|ρ| must be less than 1. (D)</p> Signup and view all the answers

What do the values ρ̂j represent in the context of autocorrelation testing?

<p>The correlations between time series values at different lags. (A)</p> Signup and view all the answers

What does the condition $ ext{Sum}_{h=0}^{ ext{∞}} |γ_h| < ∞$ signify in terms of ergodicity?

<p>The process has a stable correlation structure over infinite periods. (C)</p> Signup and view all the answers

Which statistic is known to have a higher power in small samples when testing for autocorrelation?

<p>Ljung-Box statistic QLB(k). (D)</p> Signup and view all the answers

In the context of an AR(1) process, what happens to the correlation between $y_t$ and $y_{t+h}$ as h increases?

<p>The correlation approaches zero. (B)</p> Signup and view all the answers

When yt is stationary, how is it expressed in terms of εt-i?

<p>yt = µ + ψ0 + Σ ψi εt-i. (B)</p> Signup and view all the answers

Under what conditions will ρ̂j converge to N(0, T^{-1})?

<p>If yt is i.i.d. with finite variance. (D)</p> Signup and view all the answers

What is a key assumption underlying the testing for autocorrelation using the various statistics mentioned?

<p>The error terms are serially uncorrelated. (A)</p> Signup and view all the answers

What does the term 'AR(1) model' imply about the relationship between current and past values?

<p>Only the immediately preceding value affects the current value. (D)</p> Signup and view all the answers

What is required for the weak dependence condition to hold in AR models?

<p>|β1| must be less than 1. (A)</p> Signup and view all the answers

Which statement about the error term 'ut' in the AR(1) model is correct?

<p>It has a zero expected value given all past values. (B)</p> Signup and view all the answers

What does the correlation between yt and ut in the AR(1) model imply?

<p>It indicates violation of strict exogeneity. (C)</p> Signup and view all the answers

What happens to the OLS estimator of β1 if the sample size is small and β1 is near 1?

<p>The estimator experiences severe downward bias. (A)</p> Signup and view all the answers

What do we understand by the term 'strict exogeneity' in the context of AR models?

<p>The error term is uncorrelated with all past values of y. (A)</p> Signup and view all the answers

In the AR(1) model, which assumption leads to a conflict with unbiasedness?

<p>The set of explanatory variables includes all past yt values. (A)</p> Signup and view all the answers

What implication does the equation 'E(yt | yt−1 , yt−2 ,...) = E(yt | yt−1)' have?

<p>It indicates yt is independent of previous values beyond yt−1. (C)</p> Signup and view all the answers

Flashcards

Econometrics Methods

Standard techniques for understanding and analyzing economic data.

Least Squares Estimation

A common econometric method for estimating relationships in data.

Instrumental Variables Estimation

Method used when direct estimation is biased.

Maximum Likelihood

A statistical method for finding the parameters of a model.

Signup and view all the flashcards

Cross-Sectional Data

Data collected at a single point in time for multiple subjects or entities.

Signup and view all the flashcards

Time Series Data

Data collected over several time periods for a single subject or entity.

Signup and view all the flashcards

Panel Data

Data collected over several time periods for multiple subjects, enabling analysis of both time and individual effects.

Signup and view all the flashcards

Field Experiments

Research method using treatment and control groups with random assignment, used to study treatment effects.

Signup and view all the flashcards

Econometrics

A branch of economics that uses statistical methods to analyze economic data and test theories.

Signup and view all the flashcards

Linear Regression Model

A statistical model that estimates a relationship between a dependent variable and one or more independent variables using a linear equation.

Signup and view all the flashcards

Multivariate Regression

A statistical model that estimates a relationship between a dependent variable and multiple independent variables using a linear equation.

Signup and view all the flashcards

OLS Estimator

A method of estimating the parameters of a linear regression model by minimizing the sum of squared residuals.

Signup and view all the flashcards

Difference-in-Differences

An econometric method used to estimate causal effects by comparing the changes in an outcome variable across different groups over time.

Signup and view all the flashcards

Multicollinearity

A phenomenon where independent variables in a regression model are highly correlated.

Signup and view all the flashcards

Panel Data Methods

Econometric methods used to analyze data collected from multiple individuals or entities over time (longitudinal).

Signup and view all the flashcards

Instrumental Variables (IV) Estimation

A technique to estimate causal effects when the independent variable is correlated with the error term (endogeneity).

Signup and view all the flashcards

Correlation vs. Causation

Correlation means two things occur together, but causation means one thing causes another.

Signup and view all the flashcards

Reverse Causality

When variable Y might actually cause variable X, even if X appears before Y.

Signup and view all the flashcards

Econometrics

Using statistical methods to test economic theories and determine the size of effects.

Signup and view all the flashcards

Causality

A crucial concept in economics (and policy); understanding what causes what matters.

Signup and view all the flashcards

Population Regression Line

A theoretical line showing the average relationship between variables

Signup and view all the flashcards

Linear Regression

A statistical method to estimate the relationship between variables.

Signup and view all the flashcards

Confounding Factor

An unmeasured variable that influences both the independent and dependent variables, creating spurious correlations.

Signup and view all the flashcards

Causal Relation

A direct relationship where one variable directly influences another.

Signup and view all the flashcards

Consistency (estimator)

An estimator is consistent if, as the sample size (N) increases, it gets closer and closer to the true value of the population parameter it estimates.

Signup and view all the flashcards

Bias

Systematic deviation of an estimator from the true value.

Signup and view all the flashcards

Data Objectivity

A measurement's result isn't affected by who's observing it, meaning the measurement is independent of the observer.

Signup and view all the flashcards

Data Reliability

Repeated measurements yield very similar results, meaning the measurement process is consistent and stable.

Signup and view all the flashcards

Data Validity

Measured variables accurately represent the theoretical quantities being investigated.

Signup and view all the flashcards

Data Units (Observation)

Categories of entities used to collect data, like individuals, firms, or countries.

Signup and view all the flashcards

Data Collection Sources

Places where data sets can be obtained, including national statistical agencies, international organizations, and databases.

Signup and view all the flashcards

Empirical Strategy

The process used to organize data, including checking for consistency, eliminating outliers, scrutinizing minimum and maximum values, and assessing data quality.

Signup and view all the flashcards

Confidence Interval

An interval estimated from a sample that has a specific probability of containing the true population value.

Signup and view all the flashcards

Point Estimator

A single value (estimate) that approximates a population parameter based on a sample.

Signup and view all the flashcards

Critical Value

A specific value from a known probability distribution used to define boundaries of a confidence interval.

Signup and view all the flashcards

Standard Normal Distribution

A specific bell-shaped probability distribution with a mean of 0 and a standard deviation of 1.

Signup and view all the flashcards

Sample Mean

The average of the values in a sample from a population.

Signup and view all the flashcards

Causal Effect (SLR)

The expected difference in the dependent variable (y) when the independent variable (x) changes, holding other relevant factors constant in an ideal randomized controlled experiment.

Signup and view all the flashcards

Identification Assumptions (SLR)

Assumptions needed to reliably estimate the causal effect between an independent and dependent variable in a simple linear regression, including a linear relationship, unidirectional influence, and control for other factors.

Signup and view all the flashcards

Random Sample (SLR)

A sample where each observation pair (x, y) has an identical and independent distribution, crucial for unbiased estimations in simple linear regression.

Signup and view all the flashcards

Counterfactual Question (SLR)

An unobservable question about the effect that a change in one variable would have on another. It's what would have happened to an observation if it were exposed to different conditions, but since every observation is limited to only one condition, it can't be known.

Signup and view all the flashcards

Simple Linear Regression (SLR)

A statistical model that estimates the relationship between one independent (x) and one dependent variable (y) using a straight line.

Signup and view all the flashcards

i.i.d observations

Independent and identically distributed observations; a critical assumption in many statistical models, implying each observation is drawn from the same probability distribution.

Signup and view all the flashcards

Partial Derivative

The rate of change of a function with respect to one of its variables, while holding all other variables constant.

Signup and view all the flashcards

Randomized Controlled Experiment

A research method used to study treatment effects where subjects are randomly assigned to different groups: a treatment group and a control group

Signup and view all the flashcards

Variance of x

The sample variation in the independent variable. It's not zero in a typical dataset.

Signup and view all the flashcards

Zero Conditional Mean Assumption

The error term (u) is unrelated to the independent variable (x). Knowing x gives no info about u.

Signup and view all the flashcards

Zero Mean Assumption

The average value of the error term (u) in the population is zero.

Signup and view all the flashcards

Outliers

Extreme values in the dataset (x or y). Can lead to inaccurate results.

Signup and view all the flashcards

Sampling

Collecting data over time for the same entity (panel data and time series).

Signup and view all the flashcards

Panel Data

Data collected over time for multiple subjects (entities).

Signup and view all the flashcards

Time Series Data

Data over time for a single entity.

Signup and view all the flashcards

Independent Variable

The variable that is being manipulated in an experiment.

Signup and view all the flashcards

Regression Analysis

A statistical analysis technique used to model the relationship between a dependent variable and one or more independent variables.

Signup and view all the flashcards

Public Expenditure

Government spending on public services like healthcare and education.

Signup and view all the flashcards

Life Expectancy

The average number of years a person is expected to live.

Signup and view all the flashcards

p-value (0.001)

The probability of observing results as extreme or more extreme if there's no actual effect in the population.

Signup and view all the flashcards

Significant result (***)

A statistically meaningful relationship detected in the analysis between variables.

Signup and view all the flashcards

Multiple R-squared

A measure of the proportion of the variance in the dependent variable that is predictable from the independent variables.

Signup and view all the flashcards

Residual

Difference between observed value and predicted value from a model.

Signup and view all the flashcards

F-statistic

A statistical measure that helps determine if there is a significant relationship between the set of independent variables and the dependent variable.

Signup and view all the flashcards

Independent Variable

Variable whose effect is being studied.

Signup and view all the flashcards

Dependent Variable

Variable of interest or whose value is being predicted or explained.

Signup and view all the flashcards

Error Variance (σ²)

The variability of the unobserved errors in a regression model.

Signup and view all the flashcards

Residuals (ûi)

Estimated differences between observed and predicted values in a regression model. They are an approximation of the unobserved errors.

Signup and view all the flashcards

Estimated Error Variance (σ̂²)

An estimate of the true error variance calculated using the residuals.

Signup and view all the flashcards

Degrees of Freedom (df)

Number of independent pieces of information available to estimate a parameter.

Signup and view all the flashcards

Unbiased Estimator

An estimator that, on average, equals the true value of the parameter being estimated.

Signup and view all the flashcards

R-squared

The proportion of the total variation in the dependent variable explained by the model. A value between 0 and 1.

Signup and view all the flashcards

R-squared Formula (equation 29)

R^2 = SSR/SST = 1 - SSE/SST. Where SSR is the sum of squares due to regression, SSE is the sum of squared errors, and SST is the total sum of squares.

Signup and view all the flashcards

R-squared Interpretation

Higher R-squared indicates a better fit, meaning the model explains more of the variation in the dependent variable.

Signup and view all the flashcards

R-squared and Correlation

R-squared is the square of the correlation coefficient between the actual and predicted values of the dependent variable.

Signup and view all the flashcards

Limitations of R-squared

R-squared generally increases with more independent variables, making it a poor comparison tool for models with varying numbers of predictors.

Signup and view all the flashcards

Assumption SLR.3

Sample variation in the independent variable (x) is not zero.

Signup and view all the flashcards

Assumption SLR.4

The error term's expected value is zero given any value of the independent variable.

Signup and view all the flashcards

Unbiased OLS Estimators

OLS estimators (β̂0 and β̂1) have expected values equal to the true population parameters (β0 and β1).

Signup and view all the flashcards

Biased Estimators

If any SLR assumptions fail, the OLS estimators might not accurately reflect the population parameters.

Signup and view all the flashcards

OLS Estimator β̂1 formula

The formula for estimating the slope coefficient (β1) using Ordinary Least Squares is: Σ(xi - x̄)yi / Σ(xi - x̄)²

Signup and view all the flashcards

Finite Fourth Moments

Technical assumption that the fourth moments of x and y are finite: Ex^4 and Ey^4 exist.

Signup and view all the flashcards

Conditional Mean of Error Term

The expected value of the error term (u) is zero when given a specific value of the independent variable (x).

Signup and view all the flashcards

Unbiasedness Proof

The proof involves rewriting the OLS estimator in terms of population parameters and showing that the expectation of the estimator equals the true parameter.

Signup and view all the flashcards

Summation Property

The sum of deviations from the mean of x is zero: Σ(xi - x̄) = 0

Signup and view all the flashcards

SLR.4 Likely to Fail Reason

SLR.4 (zero conditional mean) can fail due to omitted variables or measurement errors.

Signup and view all the flashcards

OLS estimator (β̂1)

A method for estimating the slope coefficient (β1) in a simple linear regression model by minimizing the sum of squared differences between observed and predicted values.

Signup and view all the flashcards

Frisch-Waugh theorem

Theorem showing that regressing y on X1 and X2 gives the same result as regressing the residuals of X2 on X1 on y.

Signup and view all the flashcards

Partial out effects

The process of isolating the effect of a given variable on another, holding constant other variables.

Signup and view all the flashcards

Regression of y on residuals of X2 on X1

Regressing y on the residuals from a regression of X2 on X1 gives the effect of X2 on y, while holding X1 constant.

Signup and view all the flashcards

OLS Estimator (y-X2βˆ2) on X1

y-X2βˆ2 represents data after subtracting the effect of X2. Regressing y-X2βˆ2 on X1 to estimate the isolated effect of X1.

Signup and view all the flashcards

Matrix M1

Matrix used to remove the influence of X1 when analyzing the effect of X2

Signup and view all the flashcards

M1 = IN − X1(X1'X1)^-1X1'

Formula shows how to construct the matrix M1, which isolates X2's effects.

Signup and view all the flashcards

M2 = IN − X2(X2'X2)^-1X2'

Formula to create matrix M2, which removes X2's effects.

Signup and view all the flashcards

OLS estimator

The ordinary least squares estimator, a way to find the best-fitting line through a set of data points by minimizing the sum of squared distances.

Signup and view all the flashcards

β̂ = (X'X)^-1 X'y

Formula for calculating the OLS estimator of the regression coefficients.

Signup and view all the flashcards

û = y − Xβ̂

Formula for the residuals in linear regression.

Signup and view all the flashcards

Geometric interpretation of OLS

OLS estimation projects data points onto the space spanned by the regressors, finding the best fit line.

Signup and view all the flashcards

Partitioned Regression

Dividing the independent variables into categories for analysis.

Signup and view all the flashcards

Unbiased OLS Estimators

Ordinary Least Squares (OLS) estimators are unbiased if their expected values equal the true population parameters.

Signup and view all the flashcards

MLR.4 (Zero Conditional Mean)

The error term's expected value is zero, given any values of the independent variables.

Signup and view all the flashcards

Homoskedasticity (MLR.5)

The error term has a constant variance, regardless of the values of the independent variables.

Signup and view all the flashcards

Variance of OLS Estimator

A measure of the variability of the OLS estimators around their true values.

Signup and view all the flashcards

Variance-Covariance Matrix of Error Term

A matrix showing the variances of individual errors and covariances between pairs of errors. This is the variance-covariance matrix.

Signup and view all the flashcards

OLS Estimator Formula

β̂ = (X'X)^−1X'y. This formula calculates the OLS estimator for the population parameters.

Signup and view all the flashcards

Unbiasedness Derivation

Derivation demonstrating that the expected value of the OLS estimators equals the true population parameters, using the OLS formula and zero conditional mean.

Signup and view all the flashcards

Assumption MLR.5

Homoscedasticity, a secondary assumption for obtaining variance formulas in Linear Regression for OLS estimators.

Signup and view all the flashcards

Var(β̃) - Var(β̂)

Difference in variance between an alternative estimator (β̃) and the OLS estimator (β̂), equal to σ^2DD' where D is a matrix.

Signup and view all the flashcards

Positive Semidefinite Matrix

A matrix M such that z'Mz ≥ 0 for all non-zero vectors z.

Signup and view all the flashcards

β̃

Alternative estimator of β, which is a vector of estimated coefficients.

Signup and view all the flashcards

β̂

Ordinary Least Squares (OLS) estimator of β, the vector of coefficients in a regression model.

Signup and view all the flashcards

Unbiased Estimator β̃

Alternative estimator β̃ is unbiased if E(β̃) = β.

Signup and view all the flashcards

DX = 0

Condition for β̃ to be unbiased; alternative estimator β̃ is unbiased if DX equals zero.

Signup and view all the flashcards

Coefficient of Determination (R²)

Proportion of variance in the dependent variable explained by the model.

Signup and view all the flashcards

SST

Total sum of squares, measuring total variation in the dependent variable.

Signup and view all the flashcards

SSE

Sum of squared errors, measuring unexplained variation in y.

Signup and view all the flashcards

SSR

Sum of squared regression, explained variation in y.

Signup and view all the flashcards

Observations in Correlation vs. Causation

Individual persons (i) in a dataset where each person has a value for whether they confused correlation and causation (x) and a result of whether they end up dying (y).

Signup and view all the flashcards

Dependent Variable (y)

The variable being studied, whose value is expected to depend on others in the dataset, in this case whether a person ends up dying (0 or 1).

Signup and view all the flashcards

Independent Variable (x)

The variable hypothesized to affect the outcome, in this example whether a person confuses correlation and causation (0 or 1).

Signup and view all the flashcards

Correlation vs. Causation

Correlation means two variables occur together—causation implies one causes the other. Important distinction to avoid misinterpretations of data.

Signup and view all the flashcards

Instrumental Variable Estimation

Method used when variables directly affecting the outcome are correlated with the error term.

Signup and view all the flashcards

Panel Data

Data collected over time for multiple subjects or entities. Enables analysis of time and individual effects.

Signup and view all the flashcards

Fixed Effects

Econometric method to control for unobserved (and potentially time-invariant) factors affecting the outcome that might vary by individual entity.

Signup and view all the flashcards

Cost and Demand Conditions

Essential variables in economics to accurately model complex relationships between variables regarding market prices and quantity demands.

Signup and view all the flashcards

CEF-decomposition property

Yi = E[Yi |Xi] + ui, where Yi is a random variable, Xi is a random variable, E[Yi |Xi] is the conditional expectation of Yi given Xi, and ui is the error term.

Signup and view all the flashcards

E[ui |Xi] = 0

The error term (ui) is mean-independent of the independent variable (Xi). Knowing Xi doesn't help predict ui.

Signup and view all the flashcards

E[h(Xi) · ui] = 0

The error term (ui) is uncorrelated with any function of the independent variable (Xi).

Signup and view all the flashcards

Bias direction (cost factors)

If zi represents cost factors, the estimated coefficient (α1) of the number of firms (Ni) in the price equation (pi = α0 + α1 Ni + ui) will be upward biased.

Signup and view all the flashcards

Bias direction (demand factors)

If zi represents demand factors, the estimated coefficient (α1) of the number of firms (Ni) in the price equation (pi = α0 + α1 Ni + ui) will be downward biased.

Signup and view all the flashcards

β̂2 calculation

β̂2 = (X20 M1 X2 )−1 (X20 M1 y )

Signup and view all the flashcards

Conditional expectation

The expected value of a random variable given the value of another random variable.

Signup and view all the flashcards

Error term (ui)

The difference between the observed value (Yi) and its conditional expectation (E[Yi |Xi]).

Signup and view all the flashcards

ln(income)

The natural logarithm of income, used as a regressor in a linear-log model.

Signup and view all the flashcards

Linear-log model

A regression model where one variable is in log form and the other is linear.

Signup and view all the flashcards

Log-linear model

A regression model where one variable is linear and the other in natural log form.

Signup and view all the flashcards

Log-log model

A regression model where both variables are in natural log form.

Signup and view all the flashcards

1% income increase

Associated with a 0.36 point increase in test score (in the provided example).

Signup and view all the flashcards

Elasticity (log-log)

Percentage change in one variable due to a 1% change in another in a log-log model.

Signup and view all the flashcards

Percentage change in Y

Calculated as 100 * (change in Y / Y).

Signup and view all the flashcards

OLS estimation in log models

Ordinary Least Squares can be used to estimate models with log transformations.

Signup and view all the flashcards

Log-Log Model

A regression model where both the dependent and independent variables are transformed using the natural logarithm.

Signup and view all the flashcards

Log-Point Changes vs. Percentage Changes

Small log-point changes approximate percentage changes, but larger log-point changes do not.

Signup and view all the flashcards

Dummy Variable

A variable that takes on values of 0 or 1, often used to represent categorical data.

Signup and view all the flashcards

Dummy Variable Trap

Perfect multicollinearity arises when all dummy variables representing categories are included in a regression, causing problems in interpretation.

Signup and view all the flashcards

Solutions to Dummy Variable Trap

Omitting one category or the intercept can address perfect multicollinearity issues in dummy variable regression models.

Signup and view all the flashcards

Wage Regressions with ln(wage)

Regression models using the natural log of wages (ln(wage)) as the dependent variable; these often analyze factors like education or experience affecting wages.

Signup and view all the flashcards

Percentage Change Calculation

Calculate the percentage change from log values using the formula: 100 * (exp(βj) - 1).

Signup and view all the flashcards

Mutual Exclusive Variables

Variables representing categories where every observation falls into exactly one category (e.g., freshmen, sophomores, juniors, seniors).

Signup and view all the flashcards

Linear-log regression

A regression model where the independent variable is transformed using a natural logarithm. Useful for understanding percentage changes in the dependent variable.

Signup and view all the flashcards

Log-linear regression

A regression model where the dependent variable is transformed using a natural logarithm. Useful for examining percentage changes of dependent variable.

Signup and view all the flashcards

Log-log regression

A regression model where both the dependent and independent variables are transformed using natural logarithms. Useful to study percentage changes of both dependent and independent variables.

Signup and view all the flashcards

Interpretation of β1 (slope)

The change in the dependent variable as the independent variable changes, adjusted based on the specific model type (linear-log, log-linear, log-log).

Signup and view all the flashcards

Linear-log interpretation of β1

A 1% increase in the independent variable is associated with a β1 / 100 unit increase in the dependent variable.

Signup and view all the flashcards

Log-linear interpretation of β1

A one-unit increase in the independent variable is associated with a 100β1 % increase in the dependent variable.

Signup and view all the flashcards

Log-log interpretation of β1

A 1% increase in the independent variable is associated with a β1 % increase in the dependent variable.

Signup and view all the flashcards

Approximation of ln(x + Δx) - ln(x)

For small changes in x, the difference in natural logarithms is approximately equal to Δx/x.

Signup and view all the flashcards

Percentage change in x

The change in x expressed as a percentage of the original value of x.

Signup and view all the flashcards

Linear-log population regression function

A linear relationship between the dependent variable and the natural logarithm of the independent variable, useful for modeling contexts where percentage changes in the independent variable relate to changes in the dependent variable.

Signup and view all the flashcards

Adjusted R-squared (women)

Proportion of variance in the dependent variable explained by the independent variables in a regression model for women, adjusted for the number of predictors.

Signup and view all the flashcards

Education (reference: compulsory school)

Wage differences based on different levels of education, compared to individuals with compulsory education.

Signup and view all the flashcards

Professional experience

Wage effect based on the number of years of professional experience.

Signup and view all the flashcards

Partnership (men)

Wage difference between married and unmarried men.

Signup and view all the flashcards

Firm (Ratio of women to men)

How the proportion of women in a firm affects wages for both men and women.

Signup and view all the flashcards

Significant at 95% level

Statistical significance level implying the relationship is unlikely to be due to chance, with only a 5% margin of error.

Signup and view all the flashcards

Constant (women/men)

Expected wage for an individual without any other characteristics.

Signup and view all the flashcards

Duration of employment

Impact of employment duration on wages, considers how long someone has held a position.

Signup and view all the flashcards

Nonlinear Regression

A regression model where the relationship between variables isn't a straight line.

Signup and view all the flashcards

Polynomial Regression

Approximating a relationship with a polynomial equation; powers of x.

Signup and view all the flashcards

Logarithmic Transformation

Changing a variable by taking its logarithm in a regression model. Interpretation of coefficients.

Signup and view all the flashcards

Nonlinear functions of one variable

Regression models where relationships are not straight lines.

Signup and view all the flashcards

Quadratic Function

A polynomial function with highest power 2.

Signup and view all the flashcards

Polynomial Regression

A statistical model where the relationship between variables is polynomial (involving squared or higher terms of independent variables).

Signup and view all the flashcards

OLS Estimation in Polynomial Regression

Using Ordinary Least Squares to estimate the coefficients in a polynomial regression model after introducing new regressors.

Signup and view all the flashcards

Interpreting Polynomial Regression Function

Interpreting the estimated polynomial regression function involves plotting predicted values against the independent variable, and computing predicted changes in dependent variable for different values of the independent variable.

Signup and view all the flashcards

Testing Linearity vs Polynomial

Testing the hypothesis that a linear model is suitable, including testing if population regression coefficients on higher degree terms are zero.

Signup and view all the flashcards

Polynomial Model Specification

Determining the appropriate degree for a polynomial regression model by analyzing the data and assessing the sensitivity of the results.

Signup and view all the flashcards

Estimated Regression Function

A mathematical formula (obtained from regression analysis) that estimates the relationship between the dependent and explanatory variables.

Signup and view all the flashcards

Quadratic Specification

A polynomial regression model with a squared term of the independent variable (e.g., income^2).

Signup and view all the flashcards

Cubic Specification

A polynomial regression model including an independent variable raised to the third power (e.g., income^3).

Signup and view all the flashcards

F-test statistic

A statistical measure used to determine if a set of independent variables has a significant relationship with a dependent variable.

Signup and view all the flashcards

F-test statistic formula 1

(SSRr − SSRur )/q / (SSRur )/(N − K − 1)

Signup and view all the flashcards

F-test statistic formula 2

(SSRr − SSRur )/q = v/(N − K − 1)
= 1/σ2

Signup and view all the flashcards

Unfeasible test

A test that cannot be directly calculated because of the complexity of the necessary calculations.

Signup and view all the flashcards

Null hypothesis (H0) in F-test

The hypothesis in an F-test that states there is no significant relationship between the independent variables and the dependent variable.

Signup and view all the flashcards

Test statistic (z)

It quantifies the variation in a parameter estimate from its hypothesized value by comparing the estimated value to its hypothesized value and incorporating the effect of variability of the estimate.

Signup and view all the flashcards

z ~ N(0, σ2 R(X'X)-1 R')

The test statistic follows a normal distribution under the null hypothesis, with a mean of zero and a variance determined by the model's structure.

Signup and view all the flashcards

t-test

A statistical test used to determine if a single coefficient (e.g., βj) is significantly different from a hypothesized value.

Signup and view all the flashcards

F-statistic

A statistical measure used to determine if a set of linear restrictions on regression coefficients are valid.

Signup and view all the flashcards

F-statistic formula

F = (R^2/K) / ((1 - R^2) / (N-K-1))

Signup and view all the flashcards

Linear Restrictions

Specific constraints placed on the coefficients of a regression model.

Signup and view all the flashcards

Restricted Model

A regression model with specific linear restrictions applied.

Signup and view all the flashcards

Unrestricted Model

A complete regression model without any specific linear restrictions.

Signup and view all the flashcards

F-statistic

A statistical measure used to determine if the change in SSR (Sum of Squared Residuals) when moving from an unrestricted to a restricted model is significant enough to warrant the restrictions

Signup and view all the flashcards

F-statistic formula (restricted vs unrestricted)

(SSRr SSRur ) / q divided by SSRur/(N K 1)

Signup and view all the flashcards

Numerator df (q)

The number of restrictions being tested in the F-statistic.

Signup and view all the flashcards

Denominator df (N K 1)

The degrees of freedom in the unrestricted model.

Signup and view all the flashcards

Rejecting H0 (F-statistic)

Reject the null hypothesis if the F-statistic exceeds a critical value using a given significance level ().

Signup and view all the flashcards

R^2 form of F-statistic

Alternative F-statistic formula using R-squared values for the restricted and unrestricted models.

Signup and view all the flashcards

Overall signicance test

Statistical test checking if all coefficients in a model are jointly significant.

Signup and view all the flashcards

F-distribution

The probability distribution of the F-statistic when the null hypothesis is true.

Signup and view all the flashcards

Coefficients of Regressors (Male)

The estimated coefficients of variables 'years', 'age', and 'tenure' when considering only male workers in the dataset.

Signup and view all the flashcards

Hourly Wage Rate

The wage rate, measured in euros per hour, for a worker.

Signup and view all the flashcards

Years of Education

Number of years of formal education for an individual.

Signup and view all the flashcards

Age (years)

The age of a worker in years.

Signup and view all the flashcards

Tenure (years)

Period of employment at a company.

Signup and view all the flashcards

Model (7)

A proposed regression model with interactions between variables and potentially for describing wage.

Signup and view all the flashcards

Female

A binary variable (0 or 1). Female = 1 If Female, 0 otherwise

Signup and view all the flashcards

Interaction Effect

The effect of two variables working together on a third variable.

Signup and view all the flashcards

ûr

Deviation of observed value from the fitted value in a restricted regression, specifically associated with model r.

Signup and view all the flashcards

SSRr

Sum of squared residuals in restricted model (r).

Signup and view all the flashcards

SSRur

Sum of squared residuals in an unrestricted model (ur).

Signup and view all the flashcards

F-test

Statistical test used to compare restricted and unrestricted regression models, evaluating the significance of restrictions.

Signup and view all the flashcards

H0: Rβ̃ = r

Null hypothesis in F-test: linear restrictions on regression coefficients (β̃) hold true.

Signup and view all the flashcards

R

Matrix of known constants defining linear restrictions on regression coefficients.

Signup and view all the flashcards

r

Vector of known constants defining linear restrictions on regression coefficients.

Signup and view all the flashcards

Linear Restrictions

Specific relationships or constraints imposed on regression coefficients.

Signup and view all the flashcards

Marginal Effect of Age on Log Wage (Female)

The expected change in log wage for a one-unit increase in age, holding other factors constant. Includes effects of age interaction terms.

Signup and view all the flashcards

Marginal Effect of Age on Log Wage (Male)

The estimated change in log wage per unit increase in age, all else equal, for male workers.

Signup and view all the flashcards

β̂2 (Log Wage Equation)

Coeff for the linear age effect in the log wage equation.

Signup and view all the flashcards

β̂8 (Log Wage Equation)

Coefficient for age effect specific to females in the log wage equation.

Signup and view all the flashcards

Population Regression Line (log wage)

The theoretical line illustrating the average relationship between variables in the log wage model when all other factors are considered constant.

Signup and view all the flashcards

Partial derivative of log wage equation

Rate of change in log wage with respect to age.

Signup and view all the flashcards

β̂3 (and β̂9) (Log Wage Equation)

Coefficient for the quadratic term of 'age' in the log wage equation, either male or female.

Signup and view all the flashcards

β̂7 (and β̂10) (Log Wage Equation)

Coefficients for variables influencing wage other than age (e.g., experience, years of work)

Signup and view all the flashcards

Test Statistic (z)

A statistic used to determine if a set of coefficients differs significantly from hypothesized values

Signup and view all the flashcards

OLS Estimator (β̂)

The ordinary least squares estimator of the coefficients in a regression model.

Signup and view all the flashcards

F-Test

A statistical test used to determine whether a set of independent variables significantly predicts a dependent variable.

Signup and view all the flashcards

F-statistic Calculation

Calculated by comparing the explained variation (SSRr) against the unexplained variation (SSRur), divided by their respective degrees of freedom.

Signup and view all the flashcards

Null Hypothesis (H0)

A statement of no effect or no difference between groups, used as a benchmark for hypothesis testing.

Signup and view all the flashcards

Chi-squared Distribution (χ2q)

A probability distribution frequently used in statistical hypothesis tests.

Signup and view all the flashcards

t-test

A statistical test used to assess the significance of a single coefficient in a regression model.

Signup and view all the flashcards

F-Distribution

A probability distribution used in hypothesis tests involving multiple variables,comparing multiple models.

Signup and view all the flashcards

One-sided t-test

A statistical test used to determine if a coefficient is significantly greater than or less than a specific value, often zero.

Signup and view all the flashcards

Two-sided t-test

A statistical test to see if a coefficient is significantly different from a specific value, usually zero.

Signup and view all the flashcards

t-statistic

A measure of how many standard errors a coefficient's estimate is away from zero.

Signup and view all the flashcards

Rejection of null hypothesis

Conclusion that there's enough evidence to reject the assumption that the coefficient is not different from zero.

Signup and view all the flashcards

Statistical Significance (α%)

The level of confidence in a statistical claim. If a result is statistically significant, it is unlikely to occur just by chance.

Signup and view all the flashcards

Testing a linear combination of parameters

Testing hypotheses about a single linear combination of parameters in a regression model, such as β1 = −β2 or β1 = θ1 − β2.

Signup and view all the flashcards

Multiple linear restrictions

Jointly testing multiple hypotheses about parameters in a regression model, often used for exclusion restrictions (checking if a group of parameters are all zero).

Signup and view all the flashcards

Restricted model

A regression model where a subset of parameters are set to specified values, usually zero, under the null hypothesis.

Signup and view all the flashcards

Exclusion restrictions

Hypotheses in regression models about parameters being zero. Used to test if certain variables have no impact on the outcome.

Signup and view all the flashcards

Joint significance

Assessing whether a group of parameter estimates in a regression model are all simultaneously different from zero (not just one).

Signup and view all the flashcards

p-value

Probability of observing results as extreme or more extreme if the null hypothesis is true (no effect).

Signup and view all the flashcards

H0: β1 = -β2

Null hypothesis asserting that the coefficient for the first independent variable is the negative of the coefficient for the second independent variable.

Signup and view all the flashcards

Compute β1=-θ1-β2

Calculating a linear combination of parameters to test hypotheses in a regression analysis. For example, to calculate coefficient β1 in the context of the test of θ1 =0

Signup and view all the flashcards

F-statistic

A statistical measure used to determine if the change in SSR (sum of squared residuals) is big enough to warrant including additional variables in a regression model.

Signup and view all the flashcards

F-statistic formula

(SSRr - SSRur)/q / SSRur/(N-K-1) where r=restricted , ur= unrestricted, q = number of restrictions, N = number of observations, K = number of unrestricted variables

Signup and view all the flashcards

Rejection of null hypothesis H0

Reject the null hypothesis (that the restricted model is as good as the unrestricted model) if the F-statistic is greater than a critical value.

Signup and view all the flashcards

Numerator df (q)

The degrees of freedom related to the restrictions being tested in your model.

Signup and view all the flashcards

Denominator df

The degrees of freedom of the unrestricted model (N − K − 1). It signifies the freedom the observations have to determine the unrestricted model.

Signup and view all the flashcards

Overall Significance

A test of whether all coefficients in a model are jointly significant.

Signup and view all the flashcards

R-squared form of F-statistic

An alternative formula for the F-statistic that uses R-squared values (R^2) to calculate model improvements.

Signup and view all the flashcards

Critical Value

A specific value from an F-distribution that separates the acceptance region from the rejection region.

Signup and view all the flashcards

Chow test

A statistical test for structural breaks in a regression model, analyzing if the relationship between variables changes over time or across different groups.

Signup and view all the flashcards

Structural break

A significant change in the relationship between variables in a regression model, often over time or across different groups.

Signup and view all the flashcards

SSR(1)

Sum of squared residuals from regression using the first sample.

Signup and view all the flashcards

USSR

Sum of squared residuals from regressions using both samples (restricted).

Signup and view all the flashcards

Chow-test F-Statistic

a statistical measure that helps determine if there is a significant relationship between the set of independent variables and the dependent variable in a structural break scenario, calculated by comparing models with and without a structural break.

Signup and view all the flashcards

Effect of using only female employees' data

Analyzing the regression results with data for female employees only.

Signup and view all the flashcards

Marginal effect of age on ln(wage) for males

The theoretical change in ln(wage) for male employees when age increases, in terms of unknown parameters.

Signup and view all the flashcards

Marginal effect of age on ln(wage) for females

The theoretical change in ln(wage) for female employees when age increases, in terms of unknown parameters.

Signup and view all the flashcards

Equality of marginal effects (males vs. females)

Testing whether the marginal effect of age on ln(wage) is the same for male and female employees using an F-test.

Signup and view all the flashcards

Identical regression equations for male and female employees.Null Hypothesis

The null hypothesis that regression equations for male and female employees are the same. This is tested by comparing the restricted OLS estimation to the unrestricted.

Signup and view all the flashcards

Consistency (estimator)

An estimator is consistent if, as the sample size (N) increases, it gets closer and closer to the true value of the population parameter it estimates.

Signup and view all the flashcards

Large sample properties

Properties of estimators that hold as the sample size (N) becomes very large, like consistency.

Signup and view all the flashcards

OLS Estimator Consistency

Under the Gauss-Markov assumptions (MLR.1-MLR.5), the OLS estimator is consistent for the true parameters (β) as the sample size (N) increases.

Signup and view all the flashcards

Consistency Meaning

As sample size (N) gets larger, the distribution of the estimator collapses toward the true parameter value.

Signup and view all the flashcards

Probability Limit (plim)

A concept used to establish consistency; it reflects the value that a statistic gets closer to when the sample size is very large.

Signup and view all the flashcards

Gauss-Markov Assumptions

A set of assumptions (MLR.1-MLR.5) necessary for the OLS estimator to be the best linear unbiased estimator (BLUE).

Signup and view all the flashcards

Law of Large Numbers

A statistical theorem stating that the sample average of a random variable converges to its expected value as the sample size increases.

Signup and view all the flashcards

SLR model consistency

OLS estimator in a simple linear regression (SLR) is consistent under specified assumptions

Signup and view all the flashcards

Consistency of OLS Estimator

The OLS estimator converges to the true population parameter as the sample size (N) grows.

Signup and view all the flashcards

plim β̂1

Probability limit of the estimated slope coefficient in a linear regression in large samples.

Signup and view all the flashcards

Zero Conditional Mean

The expected value of the error term (u) is zero given any values of the independent variables (X).

Signup and view all the flashcards

MLR.4

Assumption that the error term's expected value is zero for any values of the independent variables in multiple linear regression.

Signup and view all the flashcards

OLS Estimator Formula

β̂ = (X'X)^−1 X'y; a formula used to calculate the estimated coefficients in a linear regression.

Signup and view all the flashcards

Residuals

Difference between the observed value of the dependent variable and the predicted value from the linear regression model.

Signup and view all the flashcards

Homoskedasticity

The error term (u) in a regression model has a constant variance for all values of the independent variables.

Signup and view all the flashcards

OLS (Ordinary Least Squares)

A method used to estimate parameters in a linear regression model by minimizing the sum of squared residuals.

Signup and view all the flashcards

Sample Variance (x)

The sample variation in the independent variable, expressed as a measure of the spread of the data around the mean.

Signup and view all the flashcards

Unbiased Estimator

An estimator whose expected value is equal to the true value of the parameter it estimates.

Signup and view all the flashcards

LM Statistic

A statistical test used to determine if a set of independent variables significantly impacts a dependent variable in a regression model, in contrast to the F test or t-test for one exclusion, that may offer a similar result in large samples, but not identical.

Signup and view all the flashcards

Asymptotic Efficiency

A characteristic of OLS estimators where they have the smallest asymptotic variances compared to other consistent estimators under the Gauss-Markov assumptions, indicating they perform best in large sample estimations.

Signup and view all the flashcards

OLS Estimator (asymptotic)

Ordinary Least Squares estimators have the smallest asymptotic variances among consistent estimators when certain assumptions, such as homoskedasticity, are met.

Signup and view all the flashcards

LM Test Distribution

The distribution of the LM statistic follows a chi-squared distribution (χ2q) with a given degrees of freedom, allowing calculation of critical values or p-values for hypothesis testing.

Signup and view all the flashcards

Homoskedasticity

A crucial assumption in regression analysis (especially OLS) where the variance of the error term is constant across all values of the independent variables, ensuring the OLS estimators remain efficient

Signup and view all the flashcards

Asymptotic Normality of OLS

Under certain assumptions, the OLS estimators become approximately normally distributed as the sample size increases (approaching infinity).

Signup and view all the flashcards

Central Limit Theorem (CLT)

The CLT states that the sample mean of a large number of independent observations from any population (with finite mean and variance) will be approximately normally distributed.

Signup and view all the flashcards

Asymptotic Standard Error

An estimate of the standard deviation of an estimator, derived when considering the behavior of an estimator as the sample size approaches infinity.

Signup and view all the flashcards

Gauss-Markov Assumptions

The set of assumptions (like linear relationship, no multicollinearity, etc.) that ensure OLS estimators are optimal.

Signup and view all the flashcards

OLS Estimator Asymptotic Distribution

The OLS estimators are asymptotically normal if the errors are independent and identically distributed (i.i.d.) and the Gauss-Markov assumptions hold.

Signup and view all the flashcards

Asymptotic Normality

In large samples (large n), the distribution of estimated coefficients (like β̂) approximates a normal distribution, even if the underlying data isn't normally distributed.

Signup and view all the flashcards

t-test (asymptotically)

Even with large samples, the t-statistic test can still be used to determine if a coefficient is significantly different from zero, provided certain conditions (homoskedasticity) are met.

Signup and view all the flashcards

Large sample properties

As the sample size increases, estimators (like OLS coefficients) become more reliable and approach the true values.

Signup and view all the flashcards

Homoskedasticity

A condition in regression analysis where the variance of the error term is constant across all observations in the sample.

Signup and view all the flashcards

β̂ ∼ N(β, Q⁻¹)

The estimated parameter β̂ is asymptotically normally distributed with mean β and variance Q⁻¹.

Signup and view all the flashcards

Asymptotic Standard Error

A standard error used when the error term isn't normally distributed, and is derived as sample size increases.

Signup and view all the flashcards

Lagrange Multiplier (LM) Statistic

Alternative test statistic for multiple exclusion restrictions when using asymptotic normality and large samples.

Signup and view all the flashcards

LM Statistic (testing)

Tests if certain coefficients in a multiple regression are zero (null hypothesis).

Signup and view all the flashcards

Restricted Model

Regression model with some coefficients set to zero according to the null hypothesis.

Signup and view all the flashcards

Auxiliary Regression

Regression of model residuals on the independent variables; used in LM statistic.

Signup and view all the flashcards

Random Assignment

Assigning participants to different groups (treatment or control) randomly, ensuring groups are similar, eliminating bias.

Signup and view all the flashcards

Causal Effect

The effect a cause has on something else. Measured in Experiments

Signup and view all the flashcards

Experiment

A study designed to evaluate the effect of a treatment or intervention on an outcome variable.

Signup and view all the flashcards

Quasi-experiment

A research design that mimics an experiment by using naturally occurring variation or groups.

Signup and view all the flashcards

Internal Validity

The extent to which a study accurately measures what it intends to measure, or if the results are due to the treatment.

Signup and view all the flashcards

External Validity

The extent to which the results of an experiment can be generalized to other populations or settings.

Signup and view all the flashcards

Clinical Drug Trial

An example of an experiment, often used to assess whether a proposed drug or treatment works in reducing illness.

Signup and view all the flashcards

Program Evaluation

Field of statistics focused on analyzing and evaluating the outcome of a program or policy

Signup and view all the flashcards

School Fixed Effects

A statistical technique in econometrics that controls for unobserved, time-invariant characteristics specific to schools, reducing bias in estimated effects like teacher experience.

Signup and view all the flashcards

Bias in Teacher Experience Estimate

Without controlling for school differences (fixed effects), the estimated effect of a teacher's experience is likely overestimated, leading to an inaccurate measure of the true causal effect.

Signup and view all the flashcards

Tennessee Class Size Experiment

A famous experiment examining the effect of class size on student achievement, finding small, sustained but not cumulative, effects.

Signup and view all the flashcards

Threats to Internal Validity

Factors that could affect the accuracy of estimated causal relationships in an experiment, like incomplete treatment or non-compliance by participants.

Signup and view all the flashcards

Average Treatment Effect

The average difference in potential outcomes between a treated group and a control group.

Signup and view all the flashcards

Randomized Controlled Experiment

A research design where participants are randomly assigned to different groups. It's ideal for eliminating bias.

Signup and view all the flashcards

Instrumental Variables (IV)

A technique used when the explanatory variable is correlated with the error term (endogeneity), allowing the estimation of causal effects.

Signup and view all the flashcards

Panel Data Regression

A statistical method using data collected over time from many entities to detect and analyze relationships.

Signup and view all the flashcards

Vaccine Effectiveness

Comparing average health outcomes (y) between those who received a vaccine (x=1) and those who did not (x=0). Effective if E(y|x=1) > E(y|x=0).

Signup and view all the flashcards

Scenario 1 (Everyone gets Vaccine)

All individuals receive the vaccine. This prevents studying the effect of a vaccine on those who did not get it. Violates the sample variation assumption.

Signup and view all the flashcards

Scenario 2 (People Choose Vaccine)

Individuals choose to receive a vaccine based on their health status. Treatment (taking the vaccine) relates to the error term (outcome), thus violating the random sampling assumption.

Signup and view all the flashcards

Gauss-Markov Assumptions

Key assumptions in linear regression models, ensuring unbiased and efficient parameter estimates.

Signup and view all the flashcards

Random Sampling (SLR.2/MLR.2)

Observations (x, y) are independent and identically distributed (i.i.d). Choosing to or not to take a vaccine is a choice, not random here. Choice affects results.

Signup and view all the flashcards

Violation of SLR.3 (MLR.3)

The assumption that the independent variable (x) has variability (variance of x) is violated in a case where there is no variation, such as where everyone takes the vaccine.

Signup and view all the flashcards

Random Assignment (Experiments)

A method used to avoid potential bias in experiments by randomly assigning participants to treatment and control groups.

Signup and view all the flashcards

Experimental Design

A method of carefully controlling and measuring factors to determine causal relationships between variables and avoiding systematic bias. A critical part of the scientific method.

Signup and view all the flashcards

Difference-in-Difference

An econometric method that estimates causal effects by comparing changes in an outcome variable across different groups over time. It's useful when evaluating a policy's impact.

Signup and view all the flashcards

Interactions between variables

Relationships between two or more independent variables, where the effect of one variable can depend on the value of another.

Signup and view all the flashcards

Binary-continuous interactions

Interactions between a variable with two categories (binary) and a continuous variable. The effect of the continuous variable depends on the binary variable's value.

Signup and view all the flashcards

Card and Krueger (1994)

A famous study often used to illustrate the difference-in-differences method, investigating the effect of minimum wage increases on employment.

Signup and view all the flashcards

Binary variables

Variables that can only take on two values (e.g., male/female, present/absent, treated/control).

Signup and view all the flashcards

Interaction Term

A term in a regression model that allows the effect of one variable to depend on the value of another variable.

Signup and view all the flashcards

Binary Variable

A variable that can only take on two values, typically 0 or 1, representing the presence or absence of a characteristic.

Signup and view all the flashcards

Effect of D1i

The change in the dependent variable (y) when D1 changes from 0 to 1, holding the rest of the variables constant.

Signup and view all the flashcards

Interaction Effect (β3)

The increment to the effect of one binary variable (D1) when another binary variable (D2) is equal to 1.

Signup and view all the flashcards

Interpreting Regression Coefficients

Understanding the relationship between independent variables and predicted dependent variable by comparing different scenarios (cases).

Signup and view all the flashcards

Expected Value (E(yi))

The average value of the dependent variable (y) for a given set of conditions.

Signup and view all the flashcards

Regression Model with Interaction

Model showing how the effect of one or more variables changes depending on the value of other variables in the equation.

Signup and view all the flashcards

Interaction Interpretation

The change in impact of one variable when another variable changes from 0 to 1.

Signup and view all the flashcards

Binary-continuous interaction

A regression model where the effect of a continuous variable (x) on a dependent variable (y) depends on the value of a binary variable (D).

Signup and view all the flashcards

Regression model

An equation that describes the relationship between a dependent variable and one or more independent variables.

Signup and view all the flashcards

D=0 (group)

Observations with the binary variable D equal to 0. This represents one case for analysis.

Signup and view all the flashcards

D=1 (group)

Observations with the binary variable D equal to 1. This represents the other case for analysis.

Signup and view all the flashcards

Different intercepts, same slope

The regression line for D = 0 crosses the y-axis at a different point than the regression line for D = 1, but both have the same slope.

Signup and view all the flashcards

Different intercepts, different slopes

The regression lines for D = 0 and D = 1 have different y-intercepts and different slopes.

Signup and view all the flashcards

Same intercept, different slopes

The regression lines for both D = 0 and D = 1 intersect the y-axis at the same point (same intercept), but have different slopes.

Signup and view all the flashcards

∆y/∆x

The change in the dependent variable (y) divided by the change in the independent variable (x), representing the effect of x on y.

Signup and view all the flashcards

Regression Coefficients

Numerical values that represent the influence of independent variables on the dependent variable.

Signup and view all the flashcards

Independent Variables (Regression)

Variables used to predict or explain the dependent variable.

Signup and view all the flashcards

Dependent Variable (Regression)

The variable that is being predicted or explained in a regression model.

Signup and view all the flashcards

Difference-in-Differences Estimation

Method for estimating causal effects comparing changes in outcome across different groups over time.

Signup and view all the flashcards

Parallel Trends Assumption

Crucial assumption in DID, stating that the trends in the outcome variable would have been similar in the treatment and control groups in the absence of treatment.

Signup and view all the flashcards

No Anticipation Assumption

Assumption in DID where the treatment group has no prior knowledge/expectation of the treatment.

Signup and view all the flashcards

Regression Equation

Mathematical equation describing the relationship between the dependent and independent variable(s).

Signup and view all the flashcards

Regression Analysis

Statistical method for analyzing the relationship between variables.

Signup and view all the flashcards

Difference-in-Differences (DID)

Econometric method to estimate causal effects by comparing changes in an outcome variable across groups over time, often used with treatment and control groups.

Signup and view all the flashcards

DID Estimate (Card & Krueger 1994)

The estimated effect of a treatment (policy shift in Card and Krueger's case) calculated using DID approach on a given variable. From Table 3, it's calculated as (NJ After - NJ Before) - (PA After - PA Before) for employment.

Signup and view all the flashcards

Treatment Group (PA/NJ)

Group receiving the treatment, e.g., stores in New Jersey that were allowed to have a treatment, example of a store in NJ having a policy change at the beginning of the year.

Signup and view all the flashcards

Control Group (PA/NJ)

Group not receiving the treatment, e.g., stores in Pennsylvania that didn't have the treatment, example of a store in PA that didn't have a policy change at the beginning of the year.

Signup and view all the flashcards

FTE Employment

Full-time equivalent employment, a measure of workforce size in a firm or industry.Used in table 3

Signup and view all the flashcards

Observations (Card and Krueger)

Measurements on the subject for an observation point, like store FTE employment, before and after the policy shift, for a store in PA or NJ. Used in Table 3 of the study.

Signup and view all the flashcards

Before/After

Time periods before and after an event is introduced to compare changes in a variable. In the Card and Kruegers table 3 example, employment is observed before versus after a policy change.

Signup and view all the flashcards

Difference in Differences Estimate

The difference between the difference in the outcome variable between the treatment and control groups after and before treatment. From Table 3 in the study, it represents the effect on jobs of the policy change on store jobs.

Signup and view all the flashcards

DID estimator

A method in econometrics that estimates the average treatment effect on the treated (ATT) by comparing changes in an outcome variable across different groups over time after a treatment is implemented.

Signup and view all the flashcards

ATT

Average Treatment Effect on the Treated. The average impact on a treated group compared to a similar untreated group.

Signup and view all the flashcards

Parallel Trends

Assumption in DID that, in the absence of the treatment, the trends of the outcome variable would be similar across the treatment and control groups.

Signup and view all the flashcards

No Anticipation

Assumption in DID that the treatment doesn't affect outcomes before its implementation.

Signup and view all the flashcards

β̂ DID

The difference-in-differences estimator that measures the impact of a treatment on an outcome.

Signup and view all the flashcards

Treatment Group

The group that receives the treatment or policy being evaluated in a study.

Signup and view all the flashcards

Control Group

The group that does not receive the treatment or policy being evaluated.

Signup and view all the flashcards

Time period

Different time periods analyzed to measure change before and after a treatment.

Signup and view all the flashcards

Regression representation

A strategy that uses regression analysis to estimate the treatment effect.

Signup and view all the flashcards

Fixed Effects

Variables in a regression model accounting for unobserved characteristics that don't vary over time.

Signup and view all the flashcards

Two-way fixed effects

A regression technique controlling for both time-invariant and time-varying factors unique to each observation.

Signup and view all the flashcards

Panel Data

Data collected over multiple periods (time) from the same subjects.

Signup and view all the flashcards

Repeated Cross-Section

Studies that collect data from a new set of subjects with similar characteristics in numerous time periods.

Signup and view all the flashcards

Selection Bias

Bias arising from systematic differences between the treatment and control groups before the treatment is implemented.

Signup and view all the flashcards

Multicollinearity

High correlation among independent variables in a regression model.

Signup and view all the flashcards

Variance of β̂j

How spread out the estimated coefficients are around their true value, impacted by multicollinearity.

Signup and view all the flashcards

Rj²

R-squared value from a regression of one independent variable on the others.

Signup and view all the flashcards

Determinant (X'X)

A value derived from the matrix of independent variables (X) that indicates their linear dependence.

Signup and view all the flashcards

High R² and High Standard Errors

A good overall fit to the data (high R²) but unreliable estimations of individual coefficient effects in a regression.This is often due to high multicollinearity.

Signup and view all the flashcards

Heteroskedasticity

A situation in regression analysis where the variance of the error term is not constant across all values of the independent variables.

Signup and view all the flashcards

Variance of error term

The spread or dispersion of the errors around the regression model, which is the difference between actual and predicted values.

Signup and view all the flashcards

Variance-Covariance Matrix

A matrix showing the variances of the error terms for each observation and the covariances between them.

Signup and view all the flashcards

Heteroskedasticity Detection

Analyzing the data to identify if the variance of the error term changes as the independent variable(s) change.

Signup and view all the flashcards

Why Worry About Heteroskedasticity?

Standard error estimation and hypothesis testing are incorrect if heteroskedasticity occurs., standard OLS estimators are still unbiased, but their standard errors are biased.

Signup and view all the flashcards

Multicollinearity

High correlation between independent variables in a regression model.

Signup and view all the flashcards

Variance Inflation Factor (VIF)

A measure of multicollinearity; high VIF values (e.g.,> 10) indicate serious correlation between predictors.

Signup and view all the flashcards

VIF > 10

Indicates serious multicollinearity problems in regression analysis, potentially making coefficients unreliable.

Signup and view all the flashcards

Omitted Variable Bias

Bias in estimated coefficients when a relevant variable is left out of a regression model.

Signup and view all the flashcards

Heteroskedasticity

The variance of the error term in a regression model is not constant across observations.

Signup and view all the flashcards

Homoskedasticity

The variance of the error term in a regression model is constant across all observations.

Signup and view all the flashcards

Multicollinearity and Sample Size

Multicollinearity may not be a major concern if sample size (N) is high (as variance of the regression coefficient decreases with N).

Signup and view all the flashcards

Multicollinearity Solution

Include all relevant variables in the model and increase the sample size (N) to reduce the impact of multicollinearity.

Signup and view all the flashcards

Heteroskedasticity

The variance of the error term in a regression model is not constant across all observations.

Signup and view all the flashcards

Robust Standard Errors

Standard errors of regression coefficients that are not affected by heteroskedasticity.

Signup and view all the flashcards

White estimator

A consistent estimator for the variance-covariance matrix of OLS estimator under heteroskedasticity.

Signup and view all the flashcards

Sum of Squared Residuals (SSRj)

The sum of squared differences between observed values and predicted values in a regression.

Signup and view all the flashcards

OLS Residual (ûi)

Difference between observed value (yi) and predicted value from the OLS regression. It measures the error of prediction.

Signup and view all the flashcards

Variance of β̂

Expected value of the squared difference between an estimated coefficient β̂ and the true coefficient β.

Signup and view all the flashcards

Variance-Covariance Matrix

A matrix detailing the variance of each coefficient and the covariances between different coefficients. It describes the uncertainty in the estimated regression coefficients.

Signup and view all the flashcards

Inference under Heteroskedasticity

Drawing conclusions about relationships and significance in the presence of non-constant error variances in regression models.

Signup and view all the flashcards

Clustering in Econometrics

A technique used when observations within groups (clusters) are correlated. Standard errors need adjustment to account for this correlation.

Signup and view all the flashcards

Standard Errors (Clustering)

Adjusted standard errors considering the dependency between observations within clusters. They account for the correlation within clusters.

Signup and view all the flashcards

Incorrect Cluster Inference

Assuming observations within clusters are independent (when wrong), resulting in underestimating the standard errors and invalid conclusions.

Signup and view all the flashcards

Clustered Standard Errors Computation

In STATA, use the vce(cluster clusterid) option in the reg command for regression to appropriately estimate standard errors when you have correlated observations.

Signup and view all the flashcards

Cluster ID

A variable used to identify the group (cluster) to which each observation belongs. For example, a school ID.

Signup and view all the flashcards

Weighted Sum of Squared Residuals

The sum of squared residuals, where each residual is weighted by 1/hi (the inverse of the leverages).

Signup and view all the flashcards

Transformed Variables

Variables modified to accommodate heteroscedasticity, weighting by 1/√hi.

Signup and view all the flashcards

Weighted Least Squares (WLS)

A regression method that minimizes a weighted sum of squared residuals

Signup and view all the flashcards

Heteroskedasticity

Unequal variances in the error term of a regression model.

Signup and view all the flashcards

Generalized Least Squares (GLS)

A method to estimate regression coefficients when the error variances are not constant.

Signup and view all the flashcards

Feasible Generalized Least Squares (FGLS)

GLS estimation where parameters like error variances need estimation.

Signup and view all the flashcards

Form of heteroskedasticity (model)

A flexible formula describing how the error variance varies with the independent variables (e.g. Var(u|x) proportional to exp(δ0 + δ1 x1 +...+ δk xk).

Signup and view all the flashcards

Linearizing heteroskedasticity model

Converting a non-linear heteroskedasticity model to linear for estimation.

Signup and view all the flashcards

Strict Stationarity

A time series where the probability distribution of the variable remains constant over time, and the correlation structure between neighboring values is also constant.

Signup and view all the flashcards

Weak Stationarity

A time series where the mean and variance are constant over time, and the covariance between values depends only on the time difference, not the specific time.

Signup and view all the flashcards

Weakly Dependent Time Series

A stationary time series where the correlation between values becomes negligible as the time difference increases.

Signup and view all the flashcards

White Noise

A stationary time series where the values are uncorrelated, have zero mean, and constant variance.

Signup and view all the flashcards

IID Noise

Independent and identically distributed (IID) noise that has a constant probability distribution across all observations.

Signup and view all the flashcards

HAC Covariance Matrix

A more general covariance matrix estimator (Newey-West) that accounts for heteroscedasticity and autocorrelation in regression residuals.

Signup and view all the flashcards

Newey-West estimator

A consistent estimator of the covariance matrix of the OLS coefficients that accounts for heteroscedasticity and autocorrelation.

Signup and view all the flashcards

Truncation lag (q)

A parameter in the Newey-West estimator that represents the number of autocorrelations considered in evaluating the dynamics of the OLS residuals.

Signup and view all the flashcards

Trending Time Series

Economic time series that exhibit a clear upward or downward trend over time.

Signup and view all the flashcards

Linear Trend

A trend model where the dependent variable changes at a constant rate over time.

Signup and view all the flashcards

Exponential Trend

A trend where the dependent variable changes at a rate proportional to its current level.

Signup and view all the flashcards

Iterative Estimation

A method of finding estimates of a parameter by repeatedly refining the estimate until a certain tolerance level is met. Used in non-linear models to reach convergence.

Signup and view all the flashcards

Sum of Squared Residuals (SSR)

A measure of the difference between actual values and predicted values in a regression model.

Signup and view all the flashcards

Autocorrelation Absence Test

A statistical test to determine if there's a correlation between consecutive data points in a time series.

Signup and view all the flashcards

Box-Pierce Statistic

A portmanteau test statistic (QBP) used to check for autocorrelation in a time series; measures correlations up to a specified lag.

Signup and view all the flashcards

Ljung-Box Statistic

Another portmanteau test statistic (QLB) which assesses the absence of autocorrelation in time series data, often used in small samples; variant of Box-Pierce.

Signup and view all the flashcards

ρ̂j

Estimated autocorrelation at lag j.

Signup and view all the flashcards

χ2(k−p)

Chi-squared distribution with (k-p) degrees of freedom; used in assessing the null hypothesis of no autocorrelation in autocorrelations tests, especially with Box-Pierce and Ljung-Box tests

Signup and view all the flashcards

Null Hypothesis (Autocorrelation)

H0: all autocorrelation coefficients (ρ1, ρ2, …, ρk) are zero

Signup and view all the flashcards

Alternative Hypothesis (Autocorrelation)

H1: at least one ρj ≠ 0, for some j from 1 to k.

Signup and view all the flashcards

Portmanteau Test

A test that assesses the null hypothesis of no autocorrelation up to a specific lag (k); combines multiple autocorrelations into a single measure (like Box-Pierce or Ljung-Box methods)

Signup and view all the flashcards

Ergodicity (Definition)

A weakly stationary time series is ergodic for the expectation µ if the sample mean of the series converges to µ as the sample size gets larger.

Signup and view all the flashcards

Ergodicity Sufficient Condition

A weakly stationary process is ergodic for the second moment because the sample autocovariance converges to the true autocovariance.

Signup and view all the flashcards

MA(1) Process

A time series where the current value depends on the current error term and the previous error term, with a coefficient.

Signup and view all the flashcards

AR(1) Process

A time series where the current value depends on the previous value, plus an error term. For weak dependence, |ρ| < 1 is required.

Signup and view all the flashcards

Weak Dependence

Correlation between observations decreases as the time difference increases. The process is stationary, meaning the statistical properties like mean and variance do not change over time.

Signup and view all the flashcards

AR(1) Model

A model where the current value of a variable is predicted by the previous value plus an error term.

Signup and view all the flashcards

Efficient Markets Hypothesis (EMH)

The theory that asset prices reflect all available information, making it impossible to consistently achieve above-average returns.

Signup and view all the flashcards

AR(2) Model

A model that predicts the current value of a variable based on its two preceding values, plus an error term.

Signup and view all the flashcards

Testing EMH using AR(2)

AR(2) models test EMH by examining if the relation between the current and prev 2 data points is zero.

Signup and view all the flashcards

Null Hypothesis (H0: β1 = β2 = 0)

The assumption in the AR(2) model that there's no relationship between the current and previous two return periods.

Signup and view all the flashcards

AR(1) Model

A time series model where the current value of a variable depends linearly on its previous value.

Signup and view all the flashcards

Strict Exogeneity

Errors are unrelated to past or current values of the independent variables in a model.

Signup and view all the flashcards

Weak Dependence

The current value of a variable is less dependent on its distant past.

Signup and view all the flashcards

Lagged Dependent Variable

An independent variable in a regression that is lagged (a previous period's value).

Signup and view all the flashcards

|β1| < 1

A necessary condition for OLS to work well (consistent) in an AR(1) model.

Signup and view all the flashcards

OLS estimator of β1

Estimate of the relationship strength using Ordinary Least-Squares.

Signup and view all the flashcards

Bias of β̂1

Systematic error in estimating β1 (e.g., underestimation)

Signup and view all the flashcards

AR(p) model

A model where a variable depends on its p previous values (order p).

Signup and view all the flashcards

Study Notes

Course Introduction

  • Introductory Econometrics course taught by Simon Martin at the University of Vienna during Winter Term 2024-25.
  • Course materials are courtesy of Tomaso Duso, Martin Halla, Liyang Sun, Jesse Shapiro, and Andrea Weber.
  • The slides are based on Wooldridge (2022) and Pearson's Stock and Watson (2020) materials.

Course Aims and Content

  • Provides understanding of standard econometric methods.
  • Enables comprehension of modern empirical economic literature and conducting independent empirical analysis.
  • Covers cross-sectional, time series, and panel data.
  • Includes in-depth knowledge of Least Squares Estimation, Instrumental Variables Estimation, and Maximum Likelihood methods.
  • Relevant to students in Applied Economics, Banking and Finance, Research in Economics and Finance, and Philosophy and Economics master's programs.
  • Assumes prior knowledge in statistics, probability theory, and linear regression.

Course Logistics

Assignments and Evaluation

  • Unexcused absence from the first session results in deregistration.
  • Students must notify the instructor if unable to attend the initial session to remain enrolled.
  • Assessment:
    • Two tests (midterm, final), each 45% of the grade.
    • Homework assignments (2 exercises in groups of up to 4), each 5% of the grade.
    • Dates for tests: November 15, 2024, and January 31, 2025, each 60 minutes long.
    • Retake exam option available for those who fail one exam or miss one exam date.
    • Registration for the retake exam is due February 6, 2025.
  • Exam questions cover general course material, analytical derivations, and interpretations of empirical results.

Example: Field Experiments

  • General procedure: treatment vs control group.
  • Random (or quasi-random) assignment.
  • Example: Online dating data (Fong 2024), evaluating "network effects".

Example: Policy Evaluation

  • Example: Identifying Agglomeration Spillovers (Greenstone, Hornbeck, and Moretti 2010).
  • Analysis of winners and runners-up of large plant openings.

Example: Cash Transfer Program

  • Example from Kenya, relating to a cash transfer program.
  • Visualizing data with maps exhibiting the share of households belonging to age-set societies.

Example: Forecasting

  • Methods discussed provide a foundation for forecasting tools (e.g., demand, banking).
  • Graph of financial data with US recessions highlighted.

Why Attend This Class?

  • Gain practical data analysis skills.
  • Learn to extrapolate, analyze data, and form educated claims regarding observations.
  • Includes a variety of relevant questions; such as:
    • Does higher education cause higher income?
    • Does beauty increase chances of employment?
    • Does regulation reduce prices in telecom markets?
    • Does minimum wage affect employment?
  • Understand the implications and use of formal language and derivations.
  • Master essential technical tools for analysis.
  • Important skill set for assessing economic policy.
  • This knowledge is useful when writing a thesis about an empirical topic.

Literature

  • Main textbooks and resources used in the course.
    • Stock and Watson (2020), Introduction to Econometrics
    • Wooldridge (2020), Introductory Econometrics
    • Additional resources: Angrist and Pischke (2009), Mostly Harmless Econometrics; Greene (2019), Econometric Analysis; Cunningham (2021), Causal Inference; Wooldridge (2010), Econometric Analysis of Cross Section and Panel Data.
  • Online resources: Hanck et al. (2020) Introduction to Econometrics with R, Heiss (2020) "Using R for Econometrics

Course Plan

  • Course structure overview of topics, from basic to advanced econometric tools.
  • Topics include linear regression, multivariate regression, OLS estimator, difference-in-difference, experiments, endogeneity, instrumental variables, systems of equations, quasi-experiments, maximum likelihood estimation, discrete choice models, panel data methods, sample selection, forecasting, and dynamic causal effects.

Unit 1 Topics

  • Covers introduction, identification, estimation, testing, data structures, and appendices.

Outline

  • Covers Introduction, Identification, Estimation, Testing, Data structure, Appendix (construction of means, confidence intervals, and statistical testing).

Purpose of Scientific Research

  • Covers the purpose of scientific research and the importance of research questions, theoretical models, empirical models, data, assumptions, model, research design, and methodology.

Objectives and methods of economic research

  • Describes the aims, objectives, methodologies in economics.
  • Covers the idea of shortages and the use of models for theory and analysis.
  • Discusses the combination of related fields: Economic theory, Mathematics, and Statistics

History of Econometrics

  • Discusses the evolution of econometrics.
  • Includes the founding of the Econometric Society in 1933.
  • Includes famous Nobel prize-winning economists and achievements.

Why do we need econometrics?

  • Two key purposes: forecasting and causality
  • Forecasting: understanding the future (e.g. market and interest rate predictions)
  • Causality: understanding cause-and-effect relationships (e.g., what causes changes in a phenomenon)

Correlation vs. Causality

  • Distinguishing between correlation (statistical association) and causality (cause-and-effect relationship) is important to accurately develop economic policies and models.

Causation and Effect

  • Defining causation, and examples, and their relationship.

How can we Estimate Causal Effects?

  • The methods of evaluating whether actions taken had a particular effect on the study, avoiding bias or omitted variables.

Empirical Work

  • Discusses the process of deriving conclusions about populations from observed samples, involving identification, estimation, and testing.

Identification

  • Discusses the process of determining the appropriate assumptions within economics theory.
  • This step involves balancing the assumptions; which must include accurate descriptions on the independent and dependent variables.

To estimate, estimator, estimation

  • Discusses various estimation techniques in economics.
  • Provides relevant real-world examples, such as wage regression, demand models, and the Phillips curve.

Selection of the estimator

  • Discusses deciding on the most suitable estimator based on unbiasedness, efficiency, and consistency.

Unbiasedness

  • Defining an unbiased estimator within a specified value.

Efficiency

  • Describing estimators which provide minimized variance and deviation around the expected value.

Mean squared error

  • Defining MSE, and calculation methods, which covers both efficiency and bias.

Consistency

  • Discussing how the estimator fits or converges to the actual value in larger samples.

Data Requirements

  • Outlines three main criteria for valid data used in econometrics: Objectivity, Reliability, and Validity.
  • It suggests using descriptive statistics, such as finding the minimum and maximum values and examining data consistency.
  • It also suggests eliminating data points that appear erroneous.

Data - Units of Observation

  • This section is essential in determining the type of observations or units to be observed, such as individuals, households, companies, cities, regions, federal states, and countries. The correct choice in units of observation depends on the economic question being asked.

How to get data?

  • Provides a list of sources and types of economic data.
    • National statistical agencies
    • International organizations (e.g., World Bank, WTO, etc.)
    • Databases (e.g., ICPSR, NBER).

Data Structure

  • Defining the different types of data and their relevance to economic analysis.
    • Cross-sectional data: observation on many variables at a single time period.
    • Pooled cross-sectional data: observation on the same variables at different time periods.
    • Time series data: observations for a single variable over multiple time periods.
    • Panel data/ Longitudinal data: observations on the same variables over multiple time periods for multiple subjects.

Cross-sectional data

  • Explains cross-sectional data types and characteristics.
  • Examines when assumptions about independence across observations may be violated.

Pooled cross-sectional data

  • Detailing how pooled cross-sectional data allow for the tracking of changes or trends.

Data maintenance - GSOEP

  • Discussing data sources, surveys, types of questions, and how the surveys have been conducted and adjusted (expanded) over time.

Data scale

  • Discusses the categories of data types (nominal, ordinal, cardinal/interval).

Data preparation: Example from the current population survey

  • Describes a specific data source: The Current Population Survey (CPS), and its use and significance.

CPS raw data

  • Presents sample GSOEP data.

1st step: variable codification / 2nd step: variable names

  • Converting raw data into organized data and naming each column.

Prepared data set

  • Gives a specific data example from the study.

3rd step: descriptive statistics

  • Calculating relevant statistical features of the gathered data.

Appendix: Inference for the 1st moment of the population distribution

  • Calculates the expected value of the population.

Appendix: Variance of the sample mean

  • Calculating the variance of the sample mean and highlighting how it decreases with larger samples.

Appendix: Inference for the 2nd moment of the population distribution

  • Calculating the mean squared error associated with the sample from the population.

Appendix: Confidence interval - I, II, III, IV, V

  • Explains the method using confidence intervals for finding the right estimations from confidence levels.

Appendix: Hypotheses testing & confidence intervals

  • Explains the method of generating accurate economic hypothesis testing.

Appendix: The t-test

  • Explaining t-test methodologies, including the regions of rejection.

The t-test

  • Provides a visual interpretation of the t statistic using a graph associated with the regions of rejection.

Appendix: t-test for the BMI

  • Discussing a specific application of the t-test to evaluate whether a certain parameter is significantly different from an expected theoretical value.

Appendix: One sided t-test

  • Discusses one sided statistical tests.

Studying That Suits You

Use AI to generate personalized quizzes and flashcards to suit your learning preferences.

Quiz Team

More Like This

Use Quizgecko on...
Browser
Browser