Lecture 8: Multivariate OLS in Political Science Research

Multivariate Regression

Statistical method including multiple independent variables.

Endogeneity

Bias caused by omitted variable correlation.

Omitted Variable Bias

Error from excluding relevant variables in analysis.

Random Assignment

Method to ensure unbiased treatment effects in experiments.

Independent Variable (IV)

Variable manipulated to observe effects on dependent variable.

Dependent Variable (DV)

Outcome variable affected by independent variables.

Hypothesis (H0)

Null hypothesis stating no effect exists.

Hypothesis (HA)

Alternative hypothesis suggesting a significant effect exists.

Control Variables

Factors included to isolate the effect of IV.

Descriptive Statistics

Summary statistics to explore and describe data.

Statistically Significant

Results unlikely due to chance, typically p < 0.05.

Robustness Checks

Tests ensuring results hold under various conditions.

Reverse Causation

Situation where outcome influences the independent variable.

Measurement Error

Inaccuracies in data collection affecting results.

Political Participation

Engagement in political activities like voting.

Victimization

Experience of being a victim of crime.

Confounder

Variable influencing both independent and dependent variables.

Placebo Test

Test to check if results hold without treatment.

Empirical Measurement

Collecting data to quantify variables in research.

Research Question

Inquiry guiding the focus of a study.

Data Sources

Origins of data, such as surveys or records.

Bateson's Research

Study linking crime victimization to political activism.

Model Specification

Defining the structure and variables of a regression model.

Political Violence

Use of force to achieve political goals.