AP Statistics Unit 2 Notes

Overview of Unit 2: Exploring Two Variable Data

  • Definition: Unit 2 focuses on two variable data, where one sample is analyzed for two variables to examine relationships between them.

  • Examples of Data:

    • Sample of frogs measuring length and weight.

    • Hospital patient data tracking age and length of stay.

    • Children’s age and their TV watching habits.

  • Major Themes: Determine relationships between pairs of variables.

Key Concepts

Analyzing Two Categorical Variables
  • Two-Way Tables: A tool to organize categorical data.

  • Procedures:

    • Marginal Relative Frequencies: Values in the margins of the table that represent proportions of the total.

    • Example: Proportion of students tardy at least once.

    • Joint Relative Frequencies: Focus on the combination of two categories in the table to compute proportions.

    • Example: Proportion of students tardy who rode the bus.

    • Conditional Relative Frequencies: Adjust the denominator to only include a subset based on a condition.

    • Example: Given a student is tardy, what proportion rode the bus?

Graphical Representations
  • Segmented Bar Graphs: Visual representation of categorical data showing relative frequencies within categories.

    • Breakdown can be made based on conditions.

  • Determining Association:

    • If marginal frequencies are similar across conditions, no association exists. Example: Consistent tardy percentages regardless of transport method suggests no relation.

    • If marginal frequencies vary with conditions, an association is indicated.

Analyzing Two Quantitative Variables
  • Scatterplots: Essential tool for depicting relationships between two quantitative variables.

    • Variables: One becomes explanatory (x-axis) and the other response (y-axis).

  • Key Traits of Scatterplots:

    1. Direction: Positive (both increase) or negative (one increases, the other decreases).

    2. Form: Linear, quadratic, or other patterns observed in data spread.

    3. Strength: How closely data points follow a clear trend (strong, moderate, weak).

    4. Unusual Features: Outliers or gaps in data.

Correlation and Regression
  • Correlation Coefficient (R): Quantifies strength and direction of a linear relationship.

    • R varies from -1 to 1, closer values indicate stronger correlations.

  • Linear Regression Equation: ŷ = a + bx

    • : Predicted value, a: y-intercept, b: slope, x: explanatory variable.

  • Making Predictions: Predict response variable based on linear regression model.

  • Extrapolation vs Interpolation: Making predictions within the range of observed data vs outside the observed range (not recommended).

Residual Analysis
  • Residuals: Differences between actual and predicted values; helps analyze the fit of the model.

    • Residual = Actual - Predicted (y - ŷ).

  • Residual Plot: Should show no pattern indicating the appropriateness of the linear model.

  • Determining Best Fit: The line of best fit should minimize the sum of the squared residuals.

Coefficient of Determination
  • : Explains the variation in the response variable due to the explanatory variable.

    • Interpretation: Higher values (closer to 1) indicate better predictions.

Outliers and Influential Points
  • Outliers: Points that deviate significantly from the trend in scatterplots, affecting correlation and regression model.

  • High Leverage Points: Points that exert significant influence on the slope of the regression line, potentially altering its direction and fit.

Important Notes

  • Relationship detected does not imply causation.

  • Review computer output analysis for regression models, focusing on interpreting slope, y-intercept, R², and standard deviation of residuals (S).

Study Strategies
  • Utilize two-way tables and segmented bar graphs for categorical data analysis.

  • Draw and interpret scatterplots for quantitative data.

  • Practice calculations for R, R², and regression coefficients.


Definition: Unit 2 focuses on two variable data, where one sample is analyzed for two variables to examine relationships between them.

Examples of Data:
  • Sample of frogs measuring length and weight.

  • Hospital patient data tracking age and length of stay.

  • Children’s age and their TV watching habits.

Major Themes:

Determine relationships between pairs of variables.

Key Concepts

Analyzing Two Categorical Variables

  • Two-Way Tables: A tool to organize categorical data.

    • Procedures:

    • Marginal Relative Frequencies: Values in the margins of the table that represent proportions of the total.

      • Example: Proportion of students tardy at least once.

    • Joint Relative Frequencies: Focus on the combination of two categories in the table to compute proportions.

      • Example: Proportion of students tardy who rode the bus.

    • Conditional Relative Frequencies: Adjust the denominator to only include a subset based on a condition.

      • Example: Given a student is tardy, what proportion rode the bus?

Graphical Representations

  • Segmented Bar Graphs: Visual representation of categorical data showing relative frequencies within categories.

    • Breakdown can be made based on conditions.

  • Determining Association:

    • If marginal frequencies are similar across conditions, no association exists. Example: Consistent tardy percentages regardless of transport method suggests no relation.

    • If marginal frequencies vary with conditions, an association is indicated.

Analyzing Two Quantitative Variables

  • Scatterplots: Essential tool for depicting relationships between two quantitative variables.

    • Variables: One becomes explanatory (x-axis) and the other response (y-axis).

    • Key Traits of Scatterplots:

    • Direction: Positive (both increase) or negative (one increases, the other decreases).

    • Form: Linear, quadratic, or other patterns observed in data spread.

    • Strength: How closely data points follow a clear trend (strong, moderate, weak).

    • Unusual Features: Outliers or gaps in data.

Correlation and Regression

  • Correlation Coefficient (R): Quantifies strength and direction of a linear relationship.

    • R varies from -1 to 1, closer values indicate stronger correlations.

  • Linear Regression Equation: y^=a+bx\hat{y} = a + bx

    • y^\hat{y}: Predicted value, a: y-intercept, b: slope, x: explanatory variable.

  • Making Predictions: Predict response variable based on linear regression model.

  • Extrapolation vs Interpolation: Making predictions within the range of observed data vs outside the observed range (not recommended).

Residual Analysis

  • Residuals: Differences between actual and predicted values; helps analyze the fit of the model.

    • Residual = Actual - Predicted (yy^y - \hat{y}).

  • Residual Plot: Should show no pattern indicating the appropriateness of the linear model.

  • Determining Best Fit: The line of best fit should minimize the sum of the squared residuals.

Coefficient of Determination

  • R²: Explains the variation in the response variable due to the explanatory variable.

    • Interpretation: Higher values (closer to 1) indicate better predictions.

Outliers and Influential Points

  • Outliers: Points that deviate significantly from the trend in scatterplots, affecting correlation and regression model.

  • High Leverage Points: Points that exert significant influence on the slope of the regression line, potentially altering its direction and fit.

Important Notes
  • Relationship detected does not imply causation.

  • Review computer output analysis for regression models, focusing on interpreting slope, y-intercept, R², and standard deviation of residuals (S).

Study Strategies
  • Utilize two-way tables and segmented bar graphs for categorical data analysis.

  • Draw and interpret scatterplots for quantitative data.

  • Practice calculations for R, R², and regression coefficients