Udemy
    •  
    •  
    •  
    •  
    •  
    •  
    •  
    •  
Turn what you know into an opportunity and reach millions around the world.
Learn More
Your cart is empty.
Keep shopping
Master 2-Stage Least Squares Without Any Mathematics
Rating: 4.5 out of 5(57 ratings)
1,197 students

Master 2-Stage Least Squares Without Any Mathematics

Learn the intuition of logic of 2-Stage Least Squares Without Any Mathematics At All!
Created byDavid Tan
Last updated 3/2016
English

What you'll learn

  • Understand the endogeneity bias
  • Understand how the 2-Stage Least Squares model can mitigate the endogeneity bias
  • Identify potential instrumental variables for the 2-Stage Least Squares model
  • Understand the mechanics of the Durbin Wu-Hausman Test for endogeneity
  • Understand how to test for the strength and validity of instrumental variables
  • Estimate the 2-Stage Least Squares model in Stata and EViews
  • Conduct the diagnostic tests of the 2-Stage Least Squares model in Stata and EViews
  • Excel in their formal learning of the 2-Stage Least Squares model by building upon what is learnt in this course

Course content

6 sections26 lectures2h 1m total length
  • Introduction to Section 11:05

    A quick introduction to this Section and how to use the material in this course.

  • Correlation vs Causation2:57

    Fundamental to the understanding of 2SLS is the difference between causation and correlation. The purpose of OLS and 2SLS regressions is to examine causality, not correlation. So let's make sure we understand the difference!

  • The Assumption of Strict Exogeneity1:30

    The assumption of exogeneity is vital for OLS regressions. In this Lecture, we will understand exactly what exogeneity is and why it is important for OLS regressions.

  • The Endogeneity Bias3:45

    In this Lecture, the endogeneity bias is presented in an intuitive and completely non-technical manner. Using a simple and relate-able example, the endogeneity bias is clearly illustrated and its impact on OLS regression coefficient estimates is discussed.

  • The Fix: An Instrumental Variable1:04

    The instrumental variable (IV): The magic bullet that will eliminate the endogeneity bias. What makes it so special? How do we identify a possible IV?

    Let's find out!

  • Introducing the 2-Stage Least Squares Model4:08

    Lets walk through the logic of the famous 2-Stage Least Squares model, and discuss exactly how it eliminates the endogeneity bias. Its so simple yet elegant!

  • Hunting for an Instrument for Wages and Education1:43

    In this Lecture, we discuss the thought process of finding a potential instrument. We uncover a possible instrument variable for the classic Wages and Education relationship, and walk you through the theory of justifying a potential instrument.

  • Testing the Strength of the Instrumental Variable1:09

    What is a strong instrument? In this Lecture, we discuss how you can formally test the strength of an instrument to ensure that is gets the job done!

  • Testing the Validity of the Instrumental Variable0:55

    This Lecture discusses the Test for Overidentifying Restrictions and why it is vital in checking the validity of the instrumental variable. If an instrument is invalid, then your 2-Stage Least Squares estimates will be incorrect! This test is crucial for your rseearch!

  • Testing for Endogeneity1:47

    The Durbin and Wu-Hausman test is the classic test of endogeneity in a regression model. That is, do we even need 2-Stage Least Squares at all? When should we use 2SLS or a simple OLS regression? The results of this test will inform us which type of regression model is needed.

    This test is important and should be conducted in all research projects employing instrumental variables.

Requirements

  • Basic knowledge of Regression Analysis
  • Basic understanding of P-Values in Hypothesis Testing

Description

Welcome to my on 2-Stage Least Squares (2SLS). This course is carefully designed for students/researchers who are learning 2SLS for the first time and who are not quantitatively inclined.

In fact, this course is entirely NON-MATHEMATICAL!

This course is perfect for learning the intuition and logic of 2SLS and its corresponding diagnostic tests before formally learning the derivation and mathematics from an econometrics course or textbook.

Moreover, this course covers the application of 2SLS and its diagnostic tests using two of the most popular econometrics software packages, Stata and EViews.

At the end of this course, the student will have a clear understanding of why 2SLS is used and how it is implemented, and be able to estimate a 2SLS model using empirical data.

Who this course is for:

  • Any student or researcher who is learning the 2-Stage Least Squares model for the first time
  • Students or researchers who are not quantitatively inclined