Notes on Relative Frequency, Bar Charts, Pareto, and Pie Charts

Relative Frequency, Bar Charts, and Pie Charts

Key idea: Percentages are parts of a whole; in any data set, the parts should sum to 100% (or 1 in decimal form). The decimal form of 100% is $1.$ In decimal form, 100% = $1.0$ ; in percent form, 100% = $100\%$ .
Relative frequency basics
- Relative frequency for a category = (frequency of that category) / (total number of data values).
- Expressed as: $pi = \dfrac{fi}{N}$ where $f_i$ is the frequency of category i and $N$ is the total sample size.
- Percent version: $\text{Percentage} = pi \times 100\%$ , i.e. $\text{Percentage} = \dfrac{fi}{N} \times 100\%$ .
- The sum of all relative frequencies equals 1: $\sumi pi = 1.$ The sum of all angles in a pie chart equals 360 degrees: $\sumi \thetai = 360^{\circ}.$
Graphical representations of categorical data
- Bar chart (frequency): each category has a vertical bar whose height represents the category's frequency.
- Bar chart (relative frequency): bars depict relative frequencies instead of raw counts.
- Both are useful for comparing categories, but their scale depends on the total sample size for relative frequency charts.
- Ordering and aesthetics:
- There is no inherent natural order for most categorical data (e.g., freshman, sophomore, junior, senior, graduate). You can order bars in any way, but some orders aid interpretation.
- Pareto chart: a bar chart where bars are ordered from tallest to shortest (highest to lowest frequency). This emphasizes most frequent categories and highlights patterns quickly.
- Practical note: The appearance of a bar chart (height, width, spacing) can vary with scaling and drawing choices, but the relative ordering by category should be preserved if the data are the same.
Ordering and interpretation of data
- If you arrange categories in a specific order (e.g., by the period of charge, as mentioned in the transcript), you can produce different visual impressions, though the underlying data are unchanged.
- When you want to identify the largest vs. smallest frequencies, using a descending order (largest to smallest) can make patterns immediately visible.
- The two common methods for organizing categorical data graphically are:
- Bar chart (ordered by category as given, or alphabetically, etc.).
- Pareto chart (ordered by frequency from tallest to shortest).
Transition to pie charts
- A pie chart represents the relative proportion of each category as a sector of a circle.
- To convert a frequency table to a pie chart, you first convert frequencies to relative frequencies, then to angles.
- Steps:
1. Start with a relative frequency table: each category has a proportion $pi = fi / N$ , with $\sumi pi = 1$ .
2. Convert to degrees for a circle: each sector angle $\thetai = pi \times 360^{\circ}$ .
3. The sum of all sector angles should equal $360^{\circ}$ : $\sumi \thetai = 360^{\circ}.$
- Important interpretation: The sector size is proportional to the category's share of the total.
Practical example (data and calculations as described in the transcript)
- The lecturer discusses a data set with categorical levels: freshman, sophomore, junior, senior, graduate, unknown.
- Relative frequency discussion:
- They state the unknown category as 35% in one part of the discussion: "unknown is 35%."
- They then discuss specific category percentages for a pie-chart example: "Sophomore represents 52%. Junior 20%. Senior 25%." and later reference an unknown sector as 5% for the pie chart illustration.
- Note: The transcript contains a moment where the stated percentages (e.g., 35% unknown, 52% sophomore, 20% junior, 25% senior) would sum to more than 100%, indicating either overlapping interpretations or a mixed example. The important takeaway is the method, not the specific numbers:
 - Convert each category's relative frequency to a sector angle via $\thetai = pi \times 360^{\circ}$ .
 - Ensure the percentages sum to 100% (or the relative frequencies sum to 1) so that the sector angles sum to 360°.
- Worked steps shown in the transcript (as a teaching demonstration):
- Convert a relative frequency to a degree sector: compute $\thetai = pi \times 360^{\circ}$ for each category.
- Example notes from the transcript suggest specific steps like identifying the sector corresponding to Unknown, Sophomore, etc., and verifying that the sum of the sectors equals 360°: "$\sumi \thetai = 360^{\circ}$." The student is guided to identify which color/sector corresponds to each category and to confirm the total angle closure.
- The transcript emphasizes:
- How to determine which sector represents which category by the calculation of its angle from the relative frequency.
- The idea that a pie chart is a circular representation with sectors adding up to 360°, and that the sector size is a visual cue of the category’s proportion.
Connections to broader principles
- Visual data literacy: Pictures are said to be worth a thousand words, and good charts reveal patterns quickly (high vs. low frequencies).
- Data preparation: Converting raw frequencies to relative frequencies is a standard step before both bar and pie chart representations.
- Conceptual consistency: Whether using frequency or relative frequency bars, the underlying data are the same; the visual outcome depends on scaling and ordering.
Ethical and practical implications of data visualization discussed
- Ordering choices can influence perception (e.g., presenting data in descending order in a Pareto chart highlights the largest contributors).
- Clarity and honesty in visualization: Ensure that ordering and coloring do not mislead viewers about the magnitude or importance of categories.
- When transitioning from a table to a pie chart, verify that the data sum to the whole (100%) to avoid misrepresentation in the chart.
Summary of formulas to remember
- Relative frequency: $pi = \dfrac{fi}{N}$
- Percentage: $\text{Percentage} = pi \times 100\% = \dfrac{fi}{N} \times 100\%$
- Pie chart angles: $\thetai = pi \times 360^{\circ}$
- Angle sum for pie chart: $\sumi \thetai = 360^{\circ}$
- Relative frequencies sum: $\sumi pi = 1$
Quick recap questions you should be able to answer
- How do you convert a frequency table to a relative frequency table?
- How do you create a bar chart vs a Pareto chart, and what ordering do you use for each?
- How do you convert a relative frequency to a pie-chart sector angle? What should the total degrees be?
- If the percentages don’t sum to 100%, what does that imply about your data or calculations?
Note for students: The transcript includes several practical demonstrations and dialogues (e.g., deciding order, verifying sector sums, and discussing visual interpretation). Use these notes to practice the core steps and check for consistency in your own data visualizations.