Please complete before the next class.

Make sure that you’ve completed all previous HW first.

In class, we talked through chapter 17 of the notes “Econometric Notation” and the multiple regression model [in-class notes coming soon].

Then we talked through how to interpret a regression model. See this handout.

For the exam, you should be able to:

  1. Compute the slope, intercept, and RMS error using the avg and SD of X, the avg and SD of Y, and the correlation.
  2. Explain the difference between FPP’s notation (\(y = mx + b\)) and the econometric notation (\(y = \beta_0 + \beta_1x + u\)).
  3. Understand the notation for the multivariable regression model (\(y = \beta_0 + \beta_1x_1 + \beta_2x_2 + ... + \beta_kx_k+ u\)).
  4. Interpret the slope coefficient \(m\) or \(\beta_1\). A one-unit increase in \(x\) leads to a \(m\) or \(\beta_1\) unit increase in \(y\), on average.
  5. Explain how to use the multiple regression model to draw causal inferences. Control for confounding variables! And not colliders or mediators!
  6. Explain the sign-and-significance method and use it to interpret a regression model. For the linear regression model, we can interpret the slope directly.

Have a look at Model 1 (the left-most column) in Table 3 on p. 341 of this paper. Assuming that the author has included the appropriate controls, use the sign-and-significance method to interpret the coefficient for Oil.

  1. Are their any non-linear terms?
  2. If not, is the coefficient statistically significant? What does this mean?
  3. If so, what’s the sign of the coefficient? What does this mean?

Try the table below as well.

Computing Assignment 2

We’ve (finally) covered the material you need for Computing Assignment 2. It’s officially due April 22, but I recommend doing it soon.

Final Exam

I posted a practice quiz for probability and inference that you can take to prepare for the final. It draws upon a large bank of questions, so you can take it multiple times and get practice and feedback. (The regression material is also included on the Final, but is not covered on the practice quiz.)

Make sure you can tackle the regression questions above as well.

Creative Commons License
Carlisle Rainey