8.2 Graphical Representations of Data

8.2 Graphical Representations of Data

Quick Summary: Visualizing data transforms raw statistics into accessible insights. This section details essential graph formats (bar graphs, histograms, circle graphs, line graphs, scatter plots), establishes central sector angle calculations for pie charts, analyzes scatter plots and trend lines (correlation types, linear modeling, interpolation versus extrapolation), and provides tools to detect misleading graph distortions.

---

Categorical & Discrete Visualizations: Bar Graphs

Bar graphs (or bar charts) display categorical data or discrete numerical categories using rectangular bars.

• Key Formatting Rules:
  • Bars do not touch (there are distinct gaps between bars) because categories are discrete and separate.
  • The height (or length) of each bar is proportional to the frequency or percentage of that category.
  • The vertical axis ($y$-axis) starts at zero to ensure visual proportions are accurate.
• Grouped / Double Bar Graphs: Compare two or more sub-groups across identical categories (e.g., comparing male vs. female participation across sports).

---

Continuous Numerical Visualizations: Histograms

Histograms represent continuous numerical data that has been grouped into equal-width contiguous intervals called class intervals or bins.

• Key Formatting Rules:
  • Bars must touch with no gaps between them because the horizontal axis represents a continuous number line (gaps only occur if a bin has a frequency of zero).
  • The horizontal axis displays continuous interval boundaries (e.g., $50–59$, $60–69$, $70–79$).
  • The vertical axis represents the frequency (or relative frequency) of data points falling into each bin.

Shape Analysis of Histograms:

• Symmetric / Bell-Shaped: Data is evenly distributed around a central peak.
• Right-Skewed (Positively Skewed): Peak is concentrated on the left; long tail extends to the right.
• Left-Skewed (Negatively Skewed): Peak is concentrated on the right; long tail extends to the left.

---

Proportional Visualizations: Circle Graphs (Pie Charts)

Circle graphs (pie charts) display how a total quantity ($100%$) is partitioned into relative proportional categories.

Sector Angle Calculation Formula

The total angle around the center of a circle is $360^\circ$. The central sector angle $\theta_i$ for category $i$ is calculated using the relative frequency ratio: $\theta_i = \left( \frac{\text{Frequency of Category } i}{\text{Total Sample Size } N} \right) \times 360^\circ = \text{Percentage (decimal)} \times 360^\circ$

Sector Angle Conversion Table Example ($N = 200$):

---

Time-Series Visualizations: Line Graphs

Line graphs track continuous changes in a quantitative variable over continuous time intervals ($x$-axis).

• Data points are plotted on a coordinate grid and connected by straight line segments.
• Slope of line segments indicates rate of change: positive slope = increasing trend; negative slope = decreasing trend; horizontal segment = zero change.

---

Bivariate Data: Scatter Plots & Trend Lines

Scatter plots display bivariate numerical data ($x, y$) to investigate relationships or associations between an independent variable ($x$, explanatory) and a dependent variable ($y$, response).

graph TD
    subgraph ScatterRelationships["Correlation Patterns in Scatter Plots"]
        POS["Positive Linear Correlation<br/>(y increases as x increases | Slope m > 0)"]
        NEG["Negative Linear Correlation<br/>(y decreases as x increases | Slope m < 0)"]
        NONE["No Correlation<br/>(Points scattered randomly | No clear trend)"]
        NONLIN["Non-Linear Association<br/>(Quadratic / Exponential curve)"]
    end

1. Types of Correlation

• Positive Correlation: As $x$ increases, $y$ tends to increase (points slope upward from left to right).
• Negative Correlation: As $x$ increases, $y$ tends to decrease (points slope downward from left to right).
• No Correlation: No systematic relationship between $x$ and $y$ (points appear randomly scattered).
• Strength of Correlation: Strong (points lie very close to a line), Weak (points are widely scattered around a general trend).

2. Line of Best Fit (Trend Line)

A line of best fit is a straight line drawn through data points on a scatter plot that best represents the linear relationship.

• Properties: Passes through the centroid $(\bar{x}, \bar{y})$; minimizes overall distance to all data points; roughly equal number of points above and below the line.
• Algebraic Model: Written in slope-intercept form: $y = mx + b$, where $m$ is the rate of change and $b$ is the initial $y$-intercept.

3. Interpolation vs. Extrapolation

When using a line of best fit to estimate values:

• Interpolation: Estimating a $y$-value for an $x$-value within the domain of observed data points.
  • Reliability: High, because the mathematical model is grounded in existing empirical measurements.
• Extrapolation: Estimating a $y$-value for an $x$-value outside the domain of observed data points (either beyond the maximum or below the minimum $x$).
  • Reliability: Low to moderate; assumes that the linear trend continues indefinitely without change, which is often physically invalid.

---

Detecting Misleading Graph Techniques

Media and reports sometimes use misleading graph constructions to manipulate reader perception. Ontario educators must recognize three primary misleading techniques:

1. Truncated Vertical Axis (Broken Scale)

• Starting the $y$-axis scale at a value significantly greater than zero (e.g., starting at 75 instead of 0).
• Impact: Visually magnifies minor absolute differences, making small variations appear massive.

2. Improper Pictogram & Graphic Scaling

• Using 2D images or 3D objects to represent 1D numerical data values.
• Impact: If a bar height doubles ($2\times$), scaling the width and height of an image proportionally increases its surface area by $2^2 = 4\times$ or volume by $2^3 = 8\times$, creating a visually deceptive exaggeration.

3. Unequal or Inconsistent Axis Increments

• Spacing axis gridlines unevenly or changing scale intervals mid-axis (e.g., $0, 10, 20, 50, 100$).
• Impact: Distorts the slope of trend lines and misrepresents constant rates of change.

---

Multi-Step Worked Graph Construction & Analysis Problem

Problem Statement: An Ontario educator collected data tracking weekly study hours ($x$) and test performance ($y$ percent) for a group of Grade 10 students. The observed study times ranged between $x = 1.0$ hour and $x = 6.0$ hours.

Using linear regression, the line of best fit was determined to be: $y = 5.5x + 50$

In addition, a classroom survey of $N = 250$ students determined their preferred visual study aids:

• Summarized Notes: $100$ students
• Video Tutorials: $75$ students
• Flashcards: $50$ students
• Group Discussions: $25$ students

Tasks:

1. Calculate the central sector angles required to construct a pie chart for all four study aid categories.
2. Using the scatter plot trend line $y = 5.5x + 50$, predict the test score for a student who studies $4.0$ hours. State whether this is interpolation or extrapolation and assess reliability.
3. Predict the test score for a student who studies $12.0$ hours. Identify the mathematical limitation of this prediction.

Step-by-Step Solution:

1. Step 1: Calculate Pie Chart Sector Angles

   • Total sample $N = 250$. Total angle = $360^\circ$.
   • Summarized Notes Sector Angle: $\theta_1 = \frac{100}{250} \times 360^\circ = 0.40 \times 360^\circ = 144^\circ$
   • Video Tutorials Sector Angle: $\theta_2 = \frac{75}{250} \times 360^\circ = 0.30 \times 360^\circ = 108^\circ$
   • Flashcards Sector Angle: $\theta_3 = \frac{50}{250} \times 360^\circ = 0.20 \times 360^\circ = 72^\circ$
   • Group Discussions Sector Angle: $\theta_4 = \frac{25}{250} \times 360^\circ = 0.10 \times 360^\circ = 36^\circ$
   • Check Sum: $144^\circ + 108^\circ + 72^\circ + 36^\circ = 360^\circ$.

2. Step 2: Predict Score for $x = 4.0$ Hours (Interpolation)

   • Substitute $x = 4.0$ into $y = 5.5x + 50$: $y = 5.5(4.0) + 50 = 22 + 50 = 72\%$
   • Classification: Interpolation, because $x = 4.0$ hours lies comfortably within the observed domain ($1.0$ to $6.0$ hours). This estimate is highly reliable.

3. Step 3: Predict Score for $x = 12.0$ Hours (Extrapolation)

   • Substitute $x = 12.0$ into equation: $y = 5.5(12.0) + 50 = 66 + 50 = 116\%$
   • Classification & Limitations: This prediction is extrapolation because $12.0$ hours lies far outside the observed domain ($1.0$ to $6.0$ hours).
   • Limitation: The model yields an impossible test mark of $116%$ (exceeding $100%$ maximum). Linear growth cannot continue indefinitely due to diminished returns and performance caps.