EKHM65: Econometrics I - Lectures 5, 7, 8, 9, 10, and 11 Flashcards

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/120

flashcard set

Earn XP

Description and Tags

Comprehensive review flashcards covering simple linear regression, OLS properties, assumptions, inference, heteroskedasticity testing, robust standard errors, multiple regression, omitted variable bias, multicollinearity, nonlinear specifications, logarithms, control variable selection, interaction effects, and joint F-tests from EKHM65 Econometrics I.

121 Terms

1
New cards

In the regression line E(Y∣X)=β0+β1XE(Y \mid X) = \beta_0 + \beta_1 X, what does the intercept β0\beta_0 represent?

The expected/average value of YY when X=0X = 0.

2
New cards

In the regression line E(Y∣X)=β0+β1XE(Y \mid X) = \beta_0 + \beta_1 X, what does the slope β1\beta_1 represent?

The change in the expected/average value of YY when XX increases by one unit.

3
New cards

Why does the linear regression model assume a straight line?

It is a simple way to summarize how the average value of YY changes with XX, assuming that the average relationship is approximately linear.

4
New cards

What is the error term uu in the population regression model Y=β0+β1X+uY = \beta_0 + \beta_1 X + u?

The difference between the observed value YY and the conditional mean of YY given XX, defined as u=Y−E(Y∣X)u = Y - E(Y \mid X).

5
New cards

What is the sample residual eie_i (or u^i\hat{u}_i) in Ordinary Least Squares (OLS)?

The difference between the observed value and the fitted value: ei=Yi−Y^i=Yi−β^0−β^1Xie_i = Y_i - \hat{Y}_i = Y_i - \hat{\beta}_0 - \hat{\beta}_1 X_i.

6
New cards

What mathematical criterion does the Ordinary Least Squares (OLS) estimator minimize?

It minimizes the sum of squared residuals: ∑i=1nei2=∑i=1n(Yi−β^0−β^1Xi)2\sum_{i=1}^n e_i^2 = \sum_{i=1}^n (Y_i - \hat{\beta}_0 - \hat{\beta}_1 X_i)^2.

7
New cards

What is the formula for calculating the OLS slope estimator β^1\hat{\beta}_1?

β^1=∑i=1n(Xi−Xˉ)(Yi−Yˉ)∑i=1n(Xi−Xˉ)2\hat{\beta}_1 = \frac{\sum_{i=1}^n (X_i - \bar{X})(Y_i - \bar{Y})}{\sum_{i=1}^n (X_i - \bar{X})^2}.

8
New cards

What is the formula for the OLS intercept estimator β^0\hat{\beta}_0?

β^0=Yˉ−β^1Xˉ\hat{\beta}_0 = \bar{Y} - \hat{\beta}_1 \bar{X}.

9
New cards

In the college GPA dataset with Yˉ=3.057\bar{Y} = 3.057 and Xˉ=1.076\bar{X} = 1.076, what is the calculated value of β^0\hat{\beta}_0?

β^0=3.057−(−0.0895×1.076)=3.153\hat{\beta}_0 = 3.057 - (-0.0895 \times 1.076) = 3.153.

10
New cards

What does the coefficient of determination R2R^2 measure?

The fraction of the sample variation in YY that is accounted for by variation in the fitted values Y^\hat{Y}, calculated as R2=ESSTSSR^2 = \frac{ESS}{TSS}.

11
New cards
<p>In regression analysis, what does the bracketed quantity in this diagram represent?</p>

In regression analysis, what does the bracketed quantity in this diagram represent?

Total variation in YY, or Total Sum of Squares (TSS=∑i=1n(Yi−Yˉ)2TSS = \sum_{i=1}^n (Y_i - \bar{Y})^2).

12
New cards

What is the mathematical definition of Explained Sum of Squares (ESSESS)?

ESS=∑i=1n(Y^i−Yˉ)2ESS = \sum_{i=1}^n (\hat{Y}_i - \bar{Y})^2.

13
New cards

What is Least Squares Assumption #1 for simple linear regression?

The conditional distribution of uu given XX has a mean of zero: E(u∣X=x)=0E(u \mid X = x) = 0.

14
New cards
<p>What condition does this graph illustrate regarding Least Squares Assumption #1?</p>

What condition does this graph illustrate regarding Least Squares Assumption #1?

Zero conditional mean violated: E(u∣X)≠0E(u \mid X) \neq 0, where the actual conditional mean departs from the assumed regression line.

15
New cards

What is Least Squares Assumption #2?

(Xi,Yi),i=1,…,n(X_i, Y_i), i = 1, \dots, n are independently and identically distributed (i.i.d.) draws from the population.

16
New cards

What is Least Squares Assumption #3?

Extremely large or small values of XX and YY are sufficiently rare (outliers are unlikely).

17
New cards
<p>What econometrics concept does this figure demonstrate regarding OLS regression?</p>

What econometrics concept does this figure demonstrate regarding OLS regression?

The sensitivity of OLS regression to an outlier, showing how a single extreme observation rotates the regression line.

18
New cards

What are the three core asymptotic properties of the OLS estimator β^1\hat{\beta}_1 under the standard assumptions?

Unbiasedness (E(β^1)=β1E(\hat{\beta}_1) = \beta_1), Consistency (β^1→pβ1\hat{\beta}_1 \xrightarrow{p} \beta_1 as n→∞n \to \infty), and Asymptotic normality in large samples.

19
New cards

In the Stata output for reg colGPA skipped, what were the sample size, R2R^2, and slope coefficient for skipped?

Number of observations = 141141, R2=0.0685R^2 = 0.0685, and coefficient = −0.0895215-0.0895215.

20
New cards
<p>Using the prediction equation $$\hat{Y} = 3.15 - 0.09X$$, what is the predicted college GPA for a student skipping $$2.0$$ lectures per week?</p>

Using the prediction equation Y^=3.15−0.09X\hat{Y} = 3.15 - 0.09X, what is the predicted college GPA for a student skipping 2.02.0 lectures per week?

Y^=3.15−0.09(2.0)=2.97\hat{Y} = 3.15 - 0.09(2.0) = 2.97.

21
New cards

What do degrees of freedom represent in regression estimation, and what is the formula?

Degrees of freedom indicate the amount of independent information remaining after estimating parameters: df=n−kdf = n - k, where kk is the number of estimated parameters (n−2n - 2 for simple OLS).

22
New cards

In the college GPA regression where β^1=−0.0895\hat{\beta}_1 = -0.0895 and SE(β^1)=0.028SE(\hat{\beta}_1) = 0.028, what is the t-statistic for testing H0:β1=0H_0: \beta_1 = 0?

t=−0.0895−00.028=−3.19t = \frac{-0.0895 - 0}{0.028} = -3.19, which leads to rejecting H0H_0 at the 5%5\% level (∣−3.19∣>1.96\lvert -3.19 \rvert > 1.96).

23
New cards

What is the assumption of homoskedasticity in regression analysis?

The variance of the error distribution is constant across all values of XX: Var(u∣X=x)=σ2Var(u \mid X = x) = \sigma^2.

24
New cards

What does the Gauss-Markov theorem state about OLS under homoskedasticity (BLUE)?

OLS is the Best Linear Unbiased Estimator (BLUE), meaning it has the minimum variance (most efficient) among all linear unbiased estimators.

25
New cards

What happens to the OLS coefficient estimates in the presence of heteroskedasticity?

They remain unbiased, consistent, and asymptotically normal.

26
New cards

What is the primary damage caused by heteroskedasticity in OLS regression?

The conventional standard errors are invalid (typically underestimating the true parameter variance), which inflates t-statistics and makes hypothesis tests unreliable.

27
New cards

What is the mathematical formula for variable standardization?

Zx=X−XˉsxZ_x = \frac{X - \bar{X}}{s_x}, rescaling XX to have a mean of zero and standard deviation of one.

28
New cards

How is the slope coefficient interpreted when the independent variable XX is standardized?

The expected change in YY associated with a one standard deviation increase in XX.

29
New cards

How is the constant interpreted when the independent variable XX is standardized?

The expected value of YY for an individual with the sample-average value of XX (since Zx=0Z_x = 0 at the mean).

30
New cards

In the regression of chborn on std_educyears (n=33,587n = 33,587), what were the coefficient on std_educyears and the constant?

Coefficient = −0.7651867-0.7651867 (one SD increase in education reduces children by 0.770.77) and constant = 3.1315693.131569 children.

31
New cards

What are the null and alternative hypotheses for the Breusch-Pagan test for heteroskedasticity?

H0H_0: Homoskedasticity (constant residual variance) vs. H1H_1: Heteroskedasticity (non-constant residual variance).

32
New cards

What are the five steps to conduct the Breusch-Pagan test for heteroskedasticity?

(1) Estimate main regression, (2) compute squared residuals u^2\hat{u}^2, (3) regress u^2\hat{u}^2 on the explanatory variables, (4) compute LM=n×R2LM = n \times R^2 from that auxiliary regression, (5) compare LMLM to χcrit2(q)\chi_{crit}^2(q).

33
New cards

Why does the Breusch-Pagan test use squared residuals u^2\hat{u}^2 instead of raw residuals?

Because error size (dispersion) matters rather than sign; large positive and negative residuals both indicate high variance.

34
New cards

In the Breusch-Pagan auxiliary regression for chborn on std_educyears, what was the sample size and the auxiliary R2R^2?

Sample size n=33,587n = 33,587 and auxiliary R2=0.0245R^2 = 0.0245.

35
New cards

What was the calculated Breusch-Pagan test statistic LMLM in the chborn regression, and what was the conclusion at the 5%5\% level?

LM=33,587×0.0245=822.9LM = 33,587 \times 0.0245 = 822.9. Since 822.9>χcrit2(1)=3.84822.9 > \chi_{crit}^2(1) = 3.84, the null hypothesis of homoskedasticity is rejected.

36
New cards

What are heteroskedasticity-robust standard errors (Eicker–Huber–White standard errors)?

Standard errors that give more weight to observations with larger squared residuals when estimating parameter variance, correcting inference without altering point estimates.

37
New cards

When , robust is specified in Stata, what happens to the point estimates of the coefficients?

The coefficients remain completely identical; only the standard errors, t-statistics, p-values, confidence intervals, and F-statistic change.

38
New cards

What happens to the Stata output header when the , robust option is used?

The ANOVA table (Source, SS, MS) is replaced by 'Linear regression', and the Adjusted R2R^2 line is omitted from the printed output.

39
New cards

How can Adjusted R2R^2 be displayed in Stata after running a regression with , robust?

By executing the command display e(r2_a).

40
New cards

What causes omitted variable bias (OVB) in multiple regression?

OVB occurs when an omitted variable is correlated both with the included regressor XX and with the dependent variable YY, leading to E(u∣X)≠0E(u \mid X) \neq 0.

41
New cards

What is the population regression function for a model with two explanatory variables?

E(Y∣X1=x1,X2=x2)=β0+β1X1+β2X2E(Y \mid X_1 = x_1, X_2 = x_2) = \beta_0 + \beta_1 X_1 + \beta_2 X_2

42
New cards

How is the coefficient β1\beta_1 interpreted in the model Y=β0+β1X1+β2X2+uY = \beta_0 + \beta_1 X_1 + \beta_2 X_2 + u?

The expected change in YY per one-unit change in X1X_1, holding X2X_2 constant (a partial effect).

43
New cards

Why does plain R2R^2 mechanically increase every time an additional predictor is added to an OLS model?

Because OLS minimizes the sum of squared residuals, it can never perform worse when additional information is available.

44
New cards

What is the formula for Adjusted R2R^2?

Adj. R2=1−[n−1n−k−1]×(SSRTSS)\text{Adj. } R^2 = 1 - \left[\frac{n-1}{n-k-1}\right] \times \left(\frac{SSR}{TSS}\right), where kk is the number of regressors excluding the constant.

45
New cards

What is the difference between perfect multicollinearity and imperfect multicollinearity?

Perfect multicollinearity occurs when one regressor is an exact linear function of others (causing Stata to drop a variable); imperfect multicollinearity occurs when regressors are highly but not perfectly correlated.

46
New cards

What are textbook symptoms of severe imperfect multicollinearity in regression output?

Coefficients with unexpected signs or magnitudes, dramatic shifts in estimates when variables are added/dropped, and large standard errors/insignificant t-tests despite a high R2R^2.

47
New cards

What is the Variance Inflation Factor (VIF), and what are its common rules of thumb?

VIF quantifies how much a coefficient's variance is inflated by correlation with other regressors: VIF=1VIF = 1 means no correlation, VIF≥4VIF \ge 4 warrants scrutiny, and VIF>10VIF > 10 indicates severe multicollinearity.

48
New cards

In the occupational score example, what accounting identity linked age, labor market experience, and years of education?

agei≈experiencei+edui+7\text{age}_i \approx \text{experience}_i + \text{edu}_i + 7

49
New cards

In the occupational score regression with age, experience, and edu, what were the VIF values for these three variables?

age = 1266.081266.08, experience = 1320.361320.36, and edu = 93.4793.47 (Mean VIF = 383.59383.59).

50
New cards

Why did edu have a VIF of 93.4793.47 despite modest pairwise correlations with age (0.0560.056) and experience (−0.210-0.210)?

Because multicollinearity reflects how well a regressor is predicted by a linear combination of all other regressors jointly, not just pairwise correlations.

51
New cards

How did dropping age from the occscore model affect the coefficients on edu and experience?

edu returned to a positive, significant 1.1181.118 (t=18.05t = 18.05), and experience became positive and significant (0.1030.103, t=7.49t = 7.49).

52
New cards

In the census fertility model, how many categories did urban/rural and region of birth have, and how many dummy variables were included?

Urban/rural had 22 categories (11 dummy), and region of birth had 1616 categories (1515 dummies), following the k−1k - 1 rule.

53
New cards

In the census fertility dataset (n=33,587n = 33,587), what percentage of the sample lived in urban versus rural areas?

Urban = 61.49%61.49\% (20,65120,651 observations), Rural = 38.51%38.51\% (12,93612,936 observations).

54
New cards

Which categories were chosen as the omitted reference categories in the census fertility specification?

New England (for region of birth) and rural (for urban residence).

55
New cards

Why did foreign-born regions like Africa (n=5n = 5) and Oceania (n=6n = 6) yield insignificant p-values (0.9860.986 and 0.8830.883) in the census regression?

Their extremely small sample sizes resulted in huge standard errors (0.770.77 and 0.700.70) and very low statistical power, preventing precise estimation.

56
New cards

In the full census fertility model, what does the constant β^0=3.4436\hat{\beta}_0 = 3.4436 represent?

The predicted number of children (3.443.44) for a rural woman born in New England with the sample-average years of education.

57
New cards

How did the coefficient on std_educyears change between the simple model and the multiple regression model controlling for urban residence and region of birth?

It shrank from −0.765-0.765 in the simple model to −0.639-0.639 in the multiple model.

58
New cards

Why did the coefficient on std_educyears shrink in magnitude when urban residence and region of birth were controlled for?

Because more educated women were disproportionately urban and concentrated in lower-fertility regions; the simple model suffered from omitted variable bias.

59
New cards

What was the estimated coefficient on urban in the simple model vs. the multiple regression model?

−1.177-1.177 in the simple model, shrinking to −1.055-1.055 in the multiple regression model.

60
New cards

How did the coefficient for South Atlantic change when comparing the full model to the region-only model?

It more than doubled, increasing from 0.5030.503 in the full model to 1.1371.137 in the region-only model.

61
New cards

What lesson does the shift in region coefficients provide about omitted variable bias?

OVB does not push every coefficient in the same direction or by the same amount; it depends on how the omitted variable correlates with each specific regressor.

62
New cards

What is the formula for predicted children chborn^\widehat{\text{chborn}} in the full census model?

chborn^=3.4436+βregion+βurban×(1 if urban, else 0)+(−0.6387)×std_educyears\widehat{\text{chborn}} = 3.4436 + \beta_{\text{region}} + \beta_{\text{urban}} \times (1 \text{ if urban, else } 0) + (-0.6387) \times \text{std\_educyears}

63
New cards

What is the predicted number of children for a rural woman born in the Mountain region (β=1.0224\beta = 1.0224) with average education?

chborn^=3.4436+1.0224+0+(−0.6387×0)=4.47\widehat{\text{chborn}} = 3.4436 + 1.0224 + 0 + (-0.6387 \times 0) = 4.47 children.

64
New cards

What is the predicted number of children for an urban woman born in the Pacific region (β=−0.3508\beta = -0.3508, βurban=−1.0553\beta_{\text{urban}} = -1.0553) with education 1 SD above average?

chborn^=3.4436−0.3508−1.0553−0.6387(1)=1.40\widehat{\text{chborn}} = 3.4436 - 0.3508 - 1.0553 - 0.6387(1) = 1.40 children.

65
New cards

What is the predicted number of children for a rural woman born in South/Central America (β=1.2471\beta = 1.2471) with education 1 SD below average?

chborn^=3.4436+1.2471+0+(−0.6387×−1)=5.33\widehat{\text{chborn}} = 3.4436 + 1.2471 + 0 + (-0.6387 \times -1) = 5.33 children.

66
New cards

In the occupational score model with region dummies (reference: Northeast), what is the predicted difference between Midwest (β^=−1.93\hat{\beta} = -1.93) and South (β^=−3.06\hat{\beta} = -3.06)?

β^Midwest−β^South=−1.93−(−3.06)=1.13\hat{\beta}_{\text{Midwest}} - \hat{\beta}_{\text{South}} = -1.93 - (-3.06) = 1.13 points higher in the Midwest.

67
New cards

In the linear fertility model with age_marriage, what was the estimated coefficient on age at marriage?

−0.2113-0.2113 (t=−34.06t = -34.06), meaning each additional year delayed is associated with 0.210.21 fewer children.

68
New cards

What restrictive assumption does a single linear term for age_marriage impose?

It forces the decline in children to be identical (0.210.21 children per year) regardless of whether marriage is delayed from 20 to 21 or from 30 to 31.

69
New cards

What is a polynomial regression model?

A regression specification that includes powers of a continuous regressor (such as X2,X3,X4X^2, X^3, X^4) to model nonlinear curves.

70
New cards

Why can OLS estimate a polynomial model directly without departing from the linear regression framework?

Because the model is nonlinear in the variables, but remains linear in the parameters (β\beta coefficients).

71
New cards

In the quadratic marriage model chborn^=10.95−0.43×age_marriage+0.003×age_marriage2\widehat{\text{chborn}} = 10.95 - 0.43 \times \text{age\_marriage} + 0.003 \times \text{age\_marriage}^2, how does the marginal effect differ between age 20–21 and age 30–31?

The predicted decline is 0.280.28 children between ages 20 and 21 (3.84−3.563.84 - 3.56), but only 0.210.21 children between ages 30 and 31 (1.34−1.131.34 - 1.13).

72
New cards

What is the sequential rule for determining how many polynomial terms to include in a regression?

Add the first polynomial X2X^2 and retain if significant; then add X3X^3 and retain if significant; continue sequentially.

73
New cards

What curve shape does adding a cubic term (X3X^3) allow the regression function to take?

An S-shaped curve, allowing the relationship to change direction twice.

74
New cards

What are two dummy-variable alternatives to polynomial regression for modeling nonlinear relationships?

(1) A categorical (binned) specification grouping continuous values into ranges, and (2) A fully flexible specification with a dummy for every unique value.

75
New cards

In the binned age at marriage model (reference: 21–25), what were the coefficients for ages 15–20 and 26–30?

am_15_20 = 1.19841.1984 (t=27.31t = 27.31) and am_26_30 = −0.9675-0.9675 (t=−27.22t = -27.22).

76
New cards

What is the primary advantage and primary drawback of the fully flexible dummy specification?

Advantage: it imposes zero shape assumptions on the relationship; Drawback: it consumes many degrees of freedom, balloons standard errors at sparsely populated values, and risks fitting noise.

77
New cards

What mathematical base does the natural logarithm (ln) use, and for what values is it defined?

Base e≈2.71828e \approx 2.71828, and it is defined strictly for positive values (X>0X > 0).

78
New cards

What is the mathematical approximation for small changes in natural logarithms?

Δln⁡(X)≈%ΔX100\Delta \ln(X) \approx \frac{\% \Delta X}{100}, or ln⁡(x+Δx)−ln⁡(x)≈Δxx\ln(x + \Delta x) - \ln(x) \approx \frac{\Delta x}{x}.

79
New cards

What is the equation and coefficient interpretation of a linear-log model?

Y=β0+β1ln⁡(X)Y = \beta_0 + \beta_1 \ln(X); a 1%1\% increase in XX is associated with a 0.01×β10.01 \times \beta_1 unit change in YY.

80
New cards

In the linear-log model Income^=−912.53+501.79×ln⁡(age)\widehat{\text{Income}} = -912.53 + 501.79 \times \ln(\text{age}), what is the predicted effect of a 1%1\% increase in age?

An increase of approximately 501.79×0.01=$5.02501.79 \times 0.01 = \$5.02 in income.

81
New cards

What is the equation and coefficient interpretation of a log-linear model?

ln⁡(Y)=β0+β1X\ln(Y) = \beta_0 + \beta_1 X; a one-unit increase in XX is associated with an approximately (100×β1)%(100 \times \beta_1)\% change in YY.

82
New cards

In the log-linear model ln⁡(Income^)=5.958+0.021×age\ln(\widehat{\text{Income}}) = 5.958 + 0.021 \times \text{age}, what is the effect of one additional year of age?

An approximately 2.1%2.1\% increase in income.

83
New cards

Why did the sample size fall from 930,339930,339 to 705,151705,151 when switching from the linear-log to log-linear income model?

Because ln⁡(Y)\ln(Y) is undefined at Y=0Y = 0, so logging income dropped all observations with zero recorded earnings (about 25%25\% of the sample).

84
New cards

What happened to the age coefficient when zero-income observations were artificially recoded to 11 (Model 4)?

The age coefficient turned negative (−0.009-0.009), because non-earners anchored at ln⁡(1)=0\ln(1) = 0 skewed older, creating an artificial negative slope.

85
New cards

What is the equation and coefficient interpretation of a log-log model?

ln⁡(Y)=β0+β1ln⁡(X)\ln(Y) = \beta_0 + \beta_1 \ln(X); a 1%1\% change in XX is associated with a β1%\beta_1\% change in YY (elasticity).

86
New cards

In the log-log model ln⁡(Income^)=3.882+0.798×ln⁡(age)\ln(\widehat{\text{Income}}) = 3.882 + 0.798 \times \ln(\text{age}), what is the estimated elasticity?

0.7980.798 (a 1%1\% increase in age is associated with an approximately 0.8%0.8\% increase in income).

87
New cards

What are the five criteria that make an econometric model good?

(1) Plausible substantive relationship, (2) functional form fits data reasonably well, (3) interpretable coefficients, (4) avoids unnecessary complexity, (5) performs well according to relevant diagnostics.

88
New cards

Why can Adjusted R2R^2 not be used to compare Model 1 (Y=IncomeY = \text{Income}) and Model 2 (Y=ln⁡(income)Y = \ln(\text{income}))?

Because Adjusted R2R^2 measures variation explained in YY; differently scaled dependent variables have different total variations and are not directly comparable.

89
New cards

Why is Adjusted R2R^2 non-comparable between Model 1 (n=930,339n = 930,339) and Model 3 (n=705,151n = 705,151)?

Because the sample size changed; the jump in R2R^2 reflects the composition of the restricted sample rather than superior functional form.

90
New cards

What is the strict rule for when Adjusted R2R^2 can be used as a valid model selection tiebreaker?

It can only compare models that share both the exact same dependent variable and the exact same estimation sample.

91
New cards

In causal modeling, what is a confounder and how should it be handled?

A variable ZZ associated with XX that directly affects YY; it should generally be included as a control to prevent omitted variable bias.

92
New cards

In causal modeling, what is a mediator and why can controlling for it be problematic?

A variable lying on the causal pathway between XX and YY (X→Z→YX \to Z \to Y); controlling for it absorbs part of the total causal effect of XX on YY.

93
New cards

In causal modeling, what is reverse causality and why should such variables be excluded?

When the outcome directly affects the control variable (Y→ZY \to Z), making the control endogenous and biasing estimates.

94
New cards

Why was occupational score excluded as a control when estimating the earnings disadvantage of foreign-born men?

Because occupational attainment is a primary mediator through which immigrant status affects earnings; controlling for it eliminates the disadvantage you seek to measure.

95
New cards

Why was number of children excluded as a control variable in the men's earnings model?

Fertility responds to migration and economic status, and affects labor supply, making it a downstream endogenous variable.

96
New cards

In the earnings analysis, what was the foreign-born coefficient in Model 1 (without residential controls) vs. Model 2 (with residential controls)?

Model 1: +79.38 USD+79.38\,\text{USD} (foreign-born advantage); Model 2: −89.91 USD-89.91\,\text{USD} (foreign-born disadvantage).

97
New cards

Why did the foreign-born coefficient flip from positive (+79.38+79.38) to negative (−89.91-89.91) when residential controls were added?

Immigrants were disproportionately concentrated in higher-paying urban areas and regions (Northeast); controlling for geography removed this locational advantage.

98
New cards

In Model 4 (including citizenship), what were the predicted earnings differences for foreign-born citizens and non-citizens relative to native-born men?

Foreign-born citizens earned −50.26 USD-50.26\,\text{USD} less, while foreign-born non-citizens earned −182.55 USD-182.55\,\text{USD} less (−50.26+(−132.29)-50.26 + (-132.29)), a citizenship gap of 132.29 USD132.29\,\text{USD}.

99
New cards

Why is an additive regression model unable to evaluate whether the returns to education are greater in urban areas?

An additive model constrains the education coefficients to be identical across locations, allowing urban residence only to shift earnings by a constant parallel amount.

100
New cards

What is an interaction term in regression, and how is it constructed?

A regressor constructed by multiplying two independent variables together (X1×X2X_1 \times X_2), allowing the effect of X1X_1 on YY to depend on the level of X2X_2.