Part 5 | Intermediate General Linear Model

// modeling more complex relationships //

Part 5 tackles the messy reality of economic data: relationships that vary across groups, effects that depend on context, and unobserved factors that confound your analysis. You'll learn advanced techniques like fixed effects, interaction terms, and time series methods that let you handle these complexities systematically. These tools are essential for answering sophisticated economic questions, like whether minimum wage effects differ across regions, or how monetary policy impacts vary over business cycles.

Part 5.1 ~ Fixed Effects and Interactions

Use fixed effects to control for unobserved group differences. Create interaction terms to test whether relationships vary across groups. Interpret coefficients in models with multiple categorical predictors.

▶

Concept 5.1 // Fixed Effects and Interactions

Multiple intercepts and tests for differences in slope

Exercise 5.1 // Fixed Effects Analysis

Notebook with fixed effects and interaction examples

Homework 5.1

Due on Friday at 5PM on Gradescope.

Part 5.2 ~ Time Series

Apply lag operators to model temporal dependencies. Test for autocorrelation using residual plots and statistical tests. Use autoregressive models for time-dependent data structures.

▶

Concept 5.2 // Time Series

Models of sequences which are not independent

Exercise 5.2 // Okun's Law Analysis

Time series analysis with economic data

Homework 5.2

Due on Friday at 5PM on Gradescope.

Part 5.3 ~ Causal Modeling

Identify confounding variables that threaten causal inference. Use instrumental variables and natural experiments. Apply directed acyclic graphs (DAGs) to plan statistical controls.

Part 5.3 // Causal Modeling

Why correlation isn't causation and when GLM can help

Exercise 5.3 // Interactive Causal Analysis

Interactive examples of causal reasoning

Homework 5.3

Due on Friday at 5PM on Gradescope.

Part 5.4 ~ Using GLM Appropriately

Check model assumptions using diagnostic plots and statistical tests. Handle autocorrelation and heteroskedasticity in regression models.

Concept 5.4 // Using GLM Appropriately

Constructing the right model

Exercise 5.4 // Interactive Model Building

Interactive examples of appropriate modeling

Homework 5.4

Due on Friday at 5PM on Gradescope.

MiniExams

Part 5 contains multiple MiniExams testing your understanding of advanced linear modeling techniques including fixed effects, time series analysis, and causal inference.

MiniExam 5

Try solving the demo then check against the video.