📊

Statistics Basics and Causality

Jun 18, 2025

Overview

This lecture concludes Chapter One by reviewing foundational concepts in statistics, exploring how variables are defined and compared, and introducing treatment and outcome variables for causal studies.

Key Concepts Reviewed

  • Population is the entire group of interest in a study.
  • A sample is a subset taken from the population for analysis.
  • Variables are characteristics measured; they can be numerical or categorical.
  • Categorical variables classify data into groups, like "Yes" or "No".

Collecting and Organizing Data

  • How data is collected determines the validity of the results.
  • The method of data collection influences whether relationships or causality can be established between variables.

Causality and Experimental Design

  • Causality refers to determining if one variable (the treatment) causes a change in another (the outcome).
  • Treatment variable is the starting variable, representing the "if" in an "if-then" statement.
  • Outcome variable is the end result, representing the "then" part.
  • To study causality, divide subjects into treatment and control groups.
  • Treatment group receives the characteristic of interest; control group does not.

Example: Online Homework Study

  • Treatment variable: whether or not a student uses online homework (categorical: Yes/No).
  • Treatment group: students using online homework.
  • Control group: students using traditional homework.
  • Outcome variable: student’s overall grade at the end of the course.

Comparing Groups and Improvement

  • Improvement is determined by comparing the outcome variable (grades) between treatment and control groups.
  • The outcome variable itself is just the individual result (grade) for each student.
  • True improvement analysis and statistical confidence are covered in later chapters.

Key Terms & Definitions

  • Population — the entire group of interest in a study.
  • Sample — a subset of the population used for analysis.
  • Variable — a characteristic measured in a study.
  • Categorical Variable — a variable with categories such as "Yes" or "No".
  • Treatment Variable — the variable manipulated in an experiment (the "if" part).
  • Outcome Variable — the result measured to assess the effect of the treatment (the "then" part).
  • Treatment Group — group receiving the treatment.
  • Control Group — group not receiving the treatment.

Action Items / Next Steps

  • Review chapters 7 and 8 for methods to statistically compare groups and assess improvement.
  • Be prepared to identify treatment and outcome variables in new scenarios.