Statistics@MIT
14.381 -- Statistical Methods in Economics
Course Description: The first part of the course provides an introduction to statistical theory. A brief review of probability will be given mainly as a background material, however it is assumed to be known. The second part of the course provides an introduction to regression analysis, focusing on specification (approximation) analysis and estimation and inference in large samples.
More detailed listings of topics includes:
Part 1. Samples and their characteristics: sample vs population, histogram, sample moments, order statistics. Types of convergence and limit theorems. Data compression via sufficiency and ancillarity. Point estimates and their comparison, including Rao-Cramer bound, information matrix, consistency, asymptotic normality, asymptotic efficiency. Method of moments and maximum likelihood. Testing, including size and power, uniformly most powerful tests, and confidence sets construction.
Part 2. Specification analysis (building functional forms for regression functions using approximation theory, splines, power series, fourier, and other bases). Regression in finite samples (Gauss-Markov optimality, finite-sample inference under normality and non-normality) and large samples (consistency and asymptotic normality for independent and time series data).
This class is at the Graduate level
Instructor: V. Chernozhukov, A. Mikusheva
Prerequisites: 18.02, permission of instructor
Back to Classes
Course Description: The first part of the course provides an introduction to statistical theory. A brief review of probability will be given mainly as a background material, however it is assumed to be known. The second part of the course provides an introduction to regression analysis, focusing on specification (approximation) analysis and estimation and inference in large samples.
More detailed listings of topics includes:
Part 1. Samples and their characteristics: sample vs population, histogram, sample moments, order statistics. Types of convergence and limit theorems. Data compression via sufficiency and ancillarity. Point estimates and their comparison, including Rao-Cramer bound, information matrix, consistency, asymptotic normality, asymptotic efficiency. Method of moments and maximum likelihood. Testing, including size and power, uniformly most powerful tests, and confidence sets construction.
Part 2. Specification analysis (building functional forms for regression functions using approximation theory, splines, power series, fourier, and other bases). Regression in finite samples (Gauss-Markov optimality, finite-sample inference under normality and non-normality) and large samples (consistency and asymptotic normality for independent and time series data).
This class is at the Graduate level
Instructor: V. Chernozhukov, A. Mikusheva
Prerequisites: 18.02, permission of instructor
Insider's Wisdom
The class used to start with probability, but now that is a prerequisite and probability is quickly reviewed in the first two lectures. This class doesn't do mathematical proofs but sketches some of them. The goal of the class is to provide a foundation for future statistics and econometrics classes by giving the students a general education on the concepts and methods of statistics. For that reason, this class can be characterized as more of an "engineering" based statistics class than a "mathematical" one.
This class generally has about 40 students, half of which are from the Econ department.
Suggestions,feedback? Please send your thoughts to statistics@mit.edu