复习

说明：除了明确说明不在教学内容范围内的章节，本大纲上的所有内容都是考试内容，无轻
重之分。

复习 1：本课程的教学内容
Chapter 1 The Nature of Econometrics and Economic Data
1.1 What Is Econometrics?
1.2 Steps in Empirical Economic Analysis
1.3 The Structure of Economic Data
➢ Cross-Sectional Data
➢ Time Series Data
➢ Pooled Cross Sections
➢ Panel or Longitudinal Data
1.4 Causality and the Notion of Ceteris Paribus in Econometric Analysis
1.5 Review of Probability and Statistics
➢ Appendix B & Appendix C
➢ 关键内容：条件期望及其运算（迭代期望法则）、条件方差及其运算

PART 1 Regression Analysis with Cross-Sectional Data

Chapter 2 The Simple Regression Model
2.1 Definition of the Simple Regression Model
2.2 Deriving the Ordinary Least Squares Estimates
➢ Method of moments （矩估计量）
o E(u|x)=0 可以推出很多矩条件
o 最常用的矩条件是：E(u)=0，Cov(x,u)=E(xu)=0
o 思路 1：直接令样本矩等于总体矩。
o 思路 2：首先识别总体参数，即总体参数表示为总体矩的函数；然后再用样本
矩代替总体矩。
➢ Minimize the residual of sum of squares（OLS 估计量，严格推导见 Appendix 2A）
o 通过最小化残差平方和求出的估计量肯定就是 OLS 估计量，这是 OLS 估计量
为什么叫做“OLS”的原因。
o 当方程存在常数项时，用 E(u)=0 和 E(xu)=0 这两个矩条件求出来的矩估计量
就是 OLS 估计量。
o 利用其它矩条件，例如 E(u)=0 和 E(x2u)=0，得到的矩估计量不是 OLS 估计
量。
o OLS 估计量的表达式

2.3 Properties of OLS on Any Sample of Data

➢ Fitted Values and Residuals
➢ Algebraic Properties of OLS Statistics
o 残差与方程中的任何解释变量正交(不要忘记常数项的解释变量为 1)
o 能灵活应用 OLS 的代数性质化简证明中的表达式
➢ Goodness-of-Fit
o R 方衡量的是常数项以外的解释变量对 y 的解释能力。
o 计算 R 方的三个公式。
o 只有方程存在常数项，才肯定有 SST=SSE+SSR 成立。

2.4 Units of Measurement and Functional Form

➢ The Effects of Changing Units of Measurement on OLS Statistics
➢ Incorporating Nonlinearities in Simple Regression
 ➢ The Meaning of "Linear" Regression

2.5 Expected Values and Variances of the OLS Estimators

➢ OLS 的样本性质依赖于模型假设：SLR1-SLR5
➢ Unbiasedness of OLS
o OLS 无偏性依赖于 SLR1-SLR4
o 证明无偏性时要证到无条件期望等于真实值。
➢ Variances of the OLS Estimators
o OLS 估计量的条件方差具体表达式的推导依赖 SLR1-SLR5
o OLS 估计量的条件方差中含有 X，所以条件方差不等于无条件方差
o OLS 是 BLUE 依赖 SLR1-SLR5
➢ Estimating the Error Variance
o 扰动项方差的估计量的无偏性依赖于 SLR1-SLR5

2.6 Regression through the Origin and Regression on a Constant

➢ 过原点回归的 OLS 估计量等价于利用 E(xu)=0 求出的矩估计量。
➢ 过原点回归要小心使用，因为真实模型存在常数项的话，那么过原点回归相当于遗漏
了常数项对应的解释变量，此时 OLS 估计是有偏的。
➢ 在方程拥有常数项时，假设 E(u|x)=0，而不是假设 E(u|x)=c(c 为非零的常数)总是合
理的，因为总可以适当变化方程使得常数项包含了 c，从而得到新的扰动项服从零条
件均值，此时除了常数项有偏，斜率估计量是无偏的。但是如果强制方程不包含常数
项，那么 E(u|x)=c 成立意味着 OLS 估计的所有系数都是有偏的。
➢ Y 只对常数项进行回归，那么常数项的 OLS 估计量就是 y 的均值，残差就是错误!未找
到引用源。,此时 R 方为零。

Chapter 3 Multiple Regression Analysis: Estimation

3.1 Motivation for Multiple Regression
➢ The Model with Two Independent Variables
➢ The Model with k Independent Variables（注意：解释变量的数目是 k+1 个）
y 0 x
1 1 ... k xk u
3.2 Mechanics and Interpretation of Ordinary Least Squares
➢ Obtaining the OLS Estimates
➢ Interpreting the OLS Regression Equation
➢ On the Meaning of "Holding Other Factors Fixed" in Multiple Regression
➢ Changing More Than One Independent Variable Simultaneously
➢ OLS Fitted Values and Residuals
➢ A "Partialling Out" Interpretation of Multiple Regression

➢ Comparison of Simple and Multiple Regression Estimates

➢ Goodness-of-Fit

3.3 The Expected Value of the OLS Estimators

➢ OLS 的无偏性及其依赖的模型假设：MLR1-MLR4
➢ Including Irrelevant Variables in a Regression Model
➢ Omitted Variable Bias: The Simple Case
➢ Omitted Variable Bias: More General Cases

3.4 The Variance of the OLS Estimators

 ➢ OLS 的条件方差推导依赖于模型假设：MLR1-MLR5
➢ The Components of the OLS Variances
➢ Multicollinearity
➢ Variances in Misspecified Models
➢ Estimating σ2 Standard Errors of the OLS Estimators
3.5 Efficiency of OLS: The Gauss-Markov Theorem
3.6 Appendix 3A 第三章所有重要的证明都在附录中。
➢ 请在第二章简单线性回归模型下重复所有的证明，包括估计量无偏性、方差的推导、
方差估计量无偏性的证明，高斯马尔科夫定理的证明。
➢ 与第二章相比，第三章独有的一些证明分析请注意掌握
o 多元回归系数的经济含义、其中偏回归意义下的排除效应解释
o 遗漏变量下的偏误分析及方差分析

Chapter 4 Multiple Regression Analysis: Inference

4.1 Sampling Distributions of the OLS Estimators

4.2 Testing Hypotheses about a Single Population Parameter: The t Test

➢ Testing against One-Sided Alternatives
➢ Two-Sided Alternatives
➢ Testing Other Hypotheses about βj
➢ Computing p-Values for t Tests
➢ A Reminder on the Language of Classical Hypothesis Testing
Economic, or Practical, versus Statistical Significance

4.3 Confidence Intervals

4.4 Testing Hypotheses about a Single Linear Combination of the Parameters

4.5 Testing Multiple Linear Restrictions: The F Test

➢ Testing Exclusion Restrictions
➢ Relationship between F and t Statistics
➢ The R-Squared Form of the F Statistic
➢ Computing p-Values for F Tests
➢ The F Statistic for Overall Significance of a Regression
➢ Testing General Linear Restrictions

4.6 Reporting Regression Results

Chapter 5 Multiple Regression Analysis: OLS Asymptotics

5.1 Consistency
➢ 大数定理及其运算规则
➢ OLS 估计量的一致性依赖于模型假设：MLR1-MLR3 再加上 MLR4’
➢ Deriving the Inconsistency in OLS (MLR4’不成立，此时扰动项与解释变量相关)
5.2 Asymptotic Normality and Large Sample Inference
➢ 中心极限定理及其运算规则
➢ 简单线性回归下的渐近正态分布证明（Appendix 5A）
➢ Other Large Sample Tests: The Lagrange Multiplier Statistic

5.3 Asymptotic Efficiency of OLS

Chapter 6 Multiple Regression Analysis: Further Issues

6.1 Effects of Data Scaling on OLS Statistics
➢ Beta Coefficients
➢ Data Scaling on dependent variable
➢ Data Scaling on independent variable

6.2 More on Functional Form

➢ More on Using Logarithmic Functional Forms
o 精确百分比变动与近似百分比变动的计算
o 半弹性模型(logy on x)的精确百分比变动与近似百分比变动的计算
▪ 近似计算： % log y = 1*100*x ，即 x 变动一个单位，y 变动的百分比
是 1*100 。
▪ 精确计算： % log y = 100 * (exp(1 * x) −1) ，即 x 变动一个单位，y 变
动的百分比是100(exp(1 ) −1) .
➢ Models with Quadratics
➢ Models with Interaction Terms

6.3 More on Goodness-of-Fit and Selection of Regressors

➢ Adjusted R-Squared
➢ Using Adjusted R-Squared to Choose between Nonnested Models
o nested models: one model (the restricted model) is a special case of the
other model (the unrestricted model).
o Nonnested models：neither equation is a special case of the other.
➢ Controlling for Too Many Factors in Regression Analysis
➢ Adding Regressors to Reduce the Error Variance

6.4 Prediction and Residual Analysis

Chapter 7 Multiple Regression Analysis with Qualitative Information: Binary (or

Dummy) Variables
7.1 Describing Qualitative Information
7.2 A Single Dummy Independent Variable
➢ Interpreting Coefficients on Dummy Explanatory Variables When the Dependent
Variable Is log(y)
7.3 Using Dummy Variables for Multiple Categories
➢ Incorporating Ordinal Information by Using Dummy Variables
7.4 Interactions Involving Dummy Variables
➢ Interactions among Dummy Variables
➢ Allowing for Different Slopes
➢ Testing for Differences in Regression Functions across Groups
7.5 A Binary Dependent Variable: The Linear Probability Model
7.6 More on Policy Analysis and Program Evaluation

Chapter 8 Heteroskedasticity
8.1 Consequences of Heteroskedasticity for OLS
8.2 Heteroskedasticity-Robust Inference after OLS Estimation
➢ Computing Heteroskedasticity-Robust LM Tests
8.3 Testing for Heteroskedasticity
➢ The BP test for Heteroskedasticity
➢ The White Test for Heteroskedasticity
8.4 Weighted Least Squares Estimation
➢ The Heteroskedasticity Is Known up to a Multiplicative Constant
 ➢ The Heteroskedasticity Function Must Be Estimated: Feasible GLS
➢ What If the Assumed Heteroskedasticity Function Is Wrong?
Prediction and Prediction Intervals with Heteroskedasticity（不在教学范围内
➢ ）
8.5 The Linear Probability Model Revisited

Chapter 9 More on Specification and Data Issues

9.1 Functional Form Misspecification
➢ RESET as a General Test for Functional Form Misspecification
➢ Tests against Nonnested Alternatives（不在教学范围内）
9.2 Using Proxy Variables for Unobserved Explanatory Variables
➢ Using Lagged Dependent Variables as Proxy Variables
➢ A Different Slant on Multiple Regression
9.3 Properties of OLS under Measurement Error
➢ Measurement Error in the Dependent Variable
➢ Measurement Error in an Explanatory Variable
9.4 Missing Data, Nonrandom Samples, and Outlying Observations
➢ Missing Data
➢ Nonrandom Samples
➢ Outliers and Influential Observations

Chapter 15 Instrumental Variables Estimation and Two Stage Least Squares

15.1 Motivation: Omitted Variables in a Simple Regression Model
➢ Statistical Inference with the IV Estimator
➢ Properties of IV with a Poor Instrumental Variable and detecting Weak Instruments
➢ Computing R-Squared after IV Estimation
15.2 IV Estimation of the Multiple Regression Model
15.3 Two Stage Least Squares
➢ A Single Endogenous Explanatory Variable
➢ Multicollinearity and 2SLS
➢ Multiple Endogenous Explanatory Variables
➢ Testing Multiple Hypotheses after 2SLS Estimation
15.4 IV Solutions to Errors-in-Variables Problems
15.5 Testing for Endogeneity and Testing Overidentifying Restrictions
➢ Testing for Endogeneity
➢ Testing Overidentification Restrictions
15.6 2SLS with Heteroskedasticity
15.7 Applying 2SLS to Time Series Equations
15.8 Applying 2SLS to Pooled Cross Sections and Panel Data
15.9 Appendix 15A Assumptions for Two Stage Least Squares
复习 2：估计量样本性质的证明

➢ 有限样本性质的证明（模型假设的特点是条件的描述都是基于大 X，推导工具是条
件期望）
o OLS 估计量的无偏性（同样思路证明有偏及其偏差）
o OLS 估计量条件方差的推导（在扰动项条件同方差、条件异方差和序列相
关的情形下）
o OLS 估计量的精确样本分布，t 统计量的精确样本分布
o 高斯马尔科夫定理
o GLS 估计量的有限样本性质是相应转换方程的 OLS 估计量的有限样本性质
➢ 大样本性质的证明（模型假设的特点是条件的描述都是基于当期 x，推导工具是大
数定理与中心极限定理）
o OLS 估计量的一致性（同样思路证明不一致性及其偏差）
o OLS 估计量的渐近正态分布及其渐近方差的推导
o OLS 扰动项方差估计量的一致性
➢ IV 估计量的大样本性质
o IV 估计量的一致性
o IV 估计量的渐近正态分布及其渐近方差的推导