Lesson 2 Flashcards
What does multiple regression provide?
On average” effects, not individualized predictions.
What is a Key-X variable
The explanatory variable whose causal effect we aim to identify.
What are control variables?
Variables related to both the dependent variable and Key-X, used to isolate causal effects.
Why is holding other factors constant important?
It helps isolate the causal effect of Key-X by removing influences from control variables.
What is Condition 5 in regression models?
Key-X must be uncorrelated with the error term for valid causal inference.
What happens if Condition 5 is violated
The model’s causal claims become unreliable; must redesign or acknowledge limitations.
What is the difference between imprecision and inaccuracy
precision is random error; inaccuracy is systematic bias (e.g., omitted variables
How can causal effects change over time?
Due to changes in the economy, markets, social factors, and environment.
What do regression flowchart symbols represent?
Rectangles = Variables, Ovals = Unobserved factors, Solid arrows = Causal effects, Dashed arrows = Potential problems.
What are mediating factors and mechanisms?
Variables through which Key-X influences the outcome; explain how effects travel.
What does regression analysis do?
Quantifies the relationship between a dependent variable (Y) and explanatory variables (X).
What is the simple linear regression model formula?
Y = β0 + β1X + ε
what does the intercept (β0) represent?
The value of Y when X = 0.
What does the slope (β1) represent?
The average change in Y for a one-unit increase in X.
What is the role of the error term (ε)?
Captures influences on Y not included in the model.
What are the main assumptions for causal claims in regression?
No omitted variables, no reverse causality, no confounding factors.
What is a good regression model?
includes key variables, leaves out irrelevant ones, and avoids correlation between X and the error term.
What is R² (R-squared)?
Measures the goodness of fit; closer to 1 means a better model.
What are residuals in regression?
Differences between actual and predicted Y.
What does “holding other factors constant” mean?
Controlling for other variables to isolate the effect of the main explanatory variable.
