AP Statistics: Exploring One-Variable Data Notes

Categorical and Quantitative Variables

• Statistics vs. Parameters: Statistics are summary values calculated from a sample ($S$), whereas parameters are summary values calculated from an entire population ($P$).

• Individuals and Variables: An individual is any single entity (such as a person, object, tree, state, or day) being studied. A variable is any characteristic that changes from one individual to another.

• Categorical Variables: Take values that are category names or group labels (e.g., eye color, hair color, or species). ZIP codes are classified as categorical variables because they serve as location category identifiers rather than measured or counted quantities. Summarized using counts (frequency) or proportions (relative frequency), and displayed visually using pie charts or bar graphs / relative frequency bar graphs.

• Quantitative Variables: Take numerical values representing counts or measurements.

  • Discrete Quantitative Variables: Take countable values, typically whole numbers.

  • Continuous Quantitative Variables: Take uncountable values that can theoretically take any value within a given interval depending on measurement precision.

Graphical Displays and Distribution Characteristics

• Graphical Representations: Quantitative data can be displayed using dot plots, stem-and-leaf plots, histograms (frequency or relative frequency), and cumulative graphs.

• Cumulative Graphs: Plot cumulative proportions against values. A steeper slope indicates a higher concentration of data within that interval, while a horizontal line indicates an interval with zero data.

• Describing Distributions: Describing a quantitative distribution requires addressing four main features in context:

  • Shape: Unimodal, bimodal, uniform, symmetric, left-skewed, or right-skewed, noting any gaps or clusters.

  • Center: The typical or middle value of the data (mean or median).

  • Spread: The variability of the data values (range, $IQR$, or standard deviation).

  • Outliers: Extreme values falling far outside the overall pattern.

Measures of Center, Position, and Spread

• Measures of Center:

  • Mean ($\bar{x}$): The arithmetic average of all values; highly sensitive to extreme values and skewness.

  • Median ($M$): The physical middle value of an ordered dataset; resistant to outliers. Position can be determined using $\frac{n + 1}{2}$.

  • Relationship to Shape: Symmetric distributions have $\text{Mean} \approx \text{Median}$. Left-skewed distributions have $\text{Mean} < \text{Median}$. Right-skewed distributions have $\text{Mean} > \text{Median}$.

• Measures of Position:

  • Percentiles: The percentage of data values at or below a specific score.

  • Quartiles: First quartile ($Q_1$) represents the $25^{\text{th}}$ percentile; Median represents the $50^{\text{th}}$ percentile; Third quartile ($Q_3$) represents the $75^{\text{th}}$ percentile.

• Measures of Spread:

  • Range: Calculated as $\text{Maximum} - \text{Minimum}$; heavily affected by outliers.

  • Interquartile Range ($IQR$): Calculated as $Q_3 - Q_1$; measures the range of the middle $50\%$ of the data.

  • Standard Deviation: Measures the typical distance of data values from the mean.

• Outlier Identification Rules:

  • Fence Method: Values above $Q_3 + 1.5 \times IQR$ or below $Q_1 - 1.5 \times IQR$.

  • Standard Deviation Method: Values located more than $2$ standard deviations away from the mean ($\text{Mean} \pm 2 \times \text{Standard Deviation}$).

Summarizing, Transforming, and Comparing Data

• Five-Number Summary and Box Plots: Consists of Minimum, $Q_1$, Median, $Q_3$, and Maximum. Box plots and modified box plots divide data into four intervals, each containing $25\%$ of the dataset. Wider sections indicate greater spread, not more data points.

• Data Transformations:

  • Adding or Subtracting a Constant: Shifts all measures of center and position by that constant, but leaves all measures of spread ($IQR$, range, standard deviation) unchanged.

  • Multiplying or Dividing by a Constant: Multiplies or divides all measures of center, position, and spread by that constant.

• Comparing Distributions: When comparing two or more distributions (e.g., using parallel box plots or back-to-back stem plots), explicitly compare shape, center, spread, and outliers using comparative language (e.g., "greater than", "less than", "higher", "lower") strictly in context.