Group 2

Summarizing Data Graphically

Introduction to Data Displays

  • This section covers various ways to visually represent data, specifically:

    • Frequency distribution

    • Dotplot

    • Histogram

    • Pie chart

    • Bar chart

    • Stem-and-Leaf Display

Frequency Distribution

  • Definition: A frequency distribution is a tabular summary of data.

    • Components:

    • Classes: These can be numeric intervals or single numeric values.

    • Raw Frequency: Count of data points within each class.

    • Relative Frequency: Proportion or percentage of data points in each class compared to the total count.

      • Example: If 4 students earned an A in a class with 8 total students:

      • Raw frequency: 4

      • Relative frequency:

        • Fraction: rac48rac{4}{8}

        • Decimal: 0.50

        • Percentage: 50%

      • If instead, we had 100 students:

      • Raw frequency: 4

      • Relative frequency:

        • Fraction: rac4100rac{4}{100}

        • Decimal: 0.04

        • Percentage: 4%

General Format of Frequency Distribution
  • Structure:

    • Class

    • Frequency

    • Relative Frequency

Class

Frequency

Relative Frequency

Class 1

Frequency 1

Proportion (percent) 1

Class 2

Frequency 2

Proportion (percent) 2

…

…

…

Class n

Frequency n

Proportion (percent) n

Construction of Frequency Distribution

  1. Decide on the number of classes represented.

    • Depends on data range (span from smallest to largest).

    • For distinct values class, the number is determined by distinct values.

    • Aim for a balance: too few classes obscures trends, too many classes makes it hard to read.

      • Example involving menu prices ($0-$100) grouped into only two classes ($0-$50 & $50.01-$100) is too broad.

  2. Determine class width.

    • Use the formula: extWidth=racextRangeextNumberofClassesext{Width} = rac{ ext{Range}}{ ext{Number of Classes}}

    • Adjust to make it a round number for readability.

  3. Define where classes begin and end.

    • Start and end with nice round numbers to maintain simplicity.

    • For example, using rounded figures for a dataset of boutique prices.

Example: Boutique Prices
  • Data points: 15, 80, 104, 168, 18, 82, 106, 195, 29, 87, 107, 293, 46, 98, 108, 421, 67, 104, 153, 425

  • Range Calculation: 425−15=410425 - 15 = 410

    • Choice of Classes: 5

    • Calculated width: rac4105=82rac{410}{5} = 82. Adjusted to 80 for ease.

Class Construction for the Example
  • Start the first class at 10:

    1. First class: 10-89

    2. Next classes: 90-169, 170-249, 250-329, 330-409, 410-489

  • Important notes:

    • Classes must have no gaps or overlaps.

Frequency Determination
  • Count of the first class (10-89):

    • Data points: 15, 18, 29, 46, 67, 80, 82, 87 → Raw Frequency: 8

    • Total data points = 20 → Relative Frequency: rac820=0.4=40%rac{8}{20} = 0.4 = 40\%

Additional Example: Lunch Prices

  • Data points with two decimal places: 3.20, 6.25, 9.13, 3.89, 7.56, 9.21, 4.87, 7.81, 11.32, 5.19, 8.00, 12.52, 6.21, 8.21, 16.45

  • Range Calculation: 16.45−3.20=13.2516.45 - 3.20 = 13.25

    • Suggested Width: rac13.255=2.65rac{13.25}{5} = 2.65 → Adjusted to 3.00

Frequency Distribution for Lunch Prices

Class

Frequency

Relative Frequency

3.00-5.99

4

rac415=0.2667=26.67%rac{4}{15}=0.2667=26.67\%

6.00-8.99

6

rac615=0.40=40%rac{6}{15}=0.40=40\%

9.00-11.99

3

rac315=0.20=20%rac{3}{15}=0.20=20\%

12.00-14.99

1

rac115=0.0667=6.67%rac{1}{15}=0.0667=6.67\%

15.00-17.99

1

rac115=0.0667=6.67%rac{1}{15}=0.0667=6.67\%

Histograms
  • Definition: A histogram is a graph that displays interval data without gaps between the bars, representing frequencies along the vertical axis.

    • Features:

    • Horizontal axis = Intervals

    • Vertical axis = Frequencies

  • Example Histograms:

    • Histogram for Boutique Prices

    • Histogram for Lunch Prices

Analyzing Symmetry and Skew
  • Symmetric histogram: Higher in the center, tapering equally on both sides.

  • Skewed histogram: Clustered heavily on one end; direction of the tail indicates skew.

    • Example: Boutique prices are skewed right; lunch prices are also skewed right but less severe.

Single Value Frequency Distributions

  • Sometimes values are distinct enough to use single values as classes.

Example: Number of TVs in Households




  • Data points: 4, 6, 2, 1, 2, 3, 2, 1, 3, 2, 2, 4, 5, 1, 5, 2, 3, 3, 5, 2, 2, 3, 1, 1



  • Frequency Table:


    Class

    Frequency

    Relative Frequency



    1

    5

    rac524=0.2083=20.83%rac{5}{24}=0.2083=20.83\%



    2

    8

    rac824=0.3333=33.33%rac{8}{24}=0.3333=33.33\%



    3

    5

    rac524=0.2083=20.83%rac{5}{24}=0.2083=20.83\%



    4

    2

    rac224=0.0833=8.33%rac{2}{24}=0.0833=8.33\%



    5

    3

    rac324=0.125=12.5%rac{3}{24}=0.125=12.5\%






    Bar Charts




    • Definition: A bar chart visually represents single values (stand-alone) with gaps between bars.

    • Similar to histograms but the horizontal axis features individual values instead of intervals.

      • Example Bar Chart for TVs: Showing the frequency distribution with right skew evident.

    Categorical Data Displays

    • Similar to single-valued classes, categorical data can utilize values as classes.

    • Example: Majors of students in statistics class.

    Frequency Table for Majors

    Class

    Frequency

    Relative Frequency

    Biology

    15

    rac1530=0.50=50%rac{15}{30}=0.50=50\%

    Chemistry

    10

    rac1030=0.3333=33.33%rac{10}{30}=0.3333=33.33\%

    Math

    3

    rac330=0.10=10%rac{3}{30}=0.10=10\%

    Nursing

    2

    rac230=0.06667=6.67%rac{2}{30}=0.06667=6.67\%

    Total

    30


    Considerations for Categorical Data Display
    • Ordering Issues: Categorical data does not have a natural order, thus discussions of symmetry skew in bar charts are inappropriate.

    • If data has implied ranking (e.g., military rank), it becomes meaningful to discuss symmetry.

    Stem-and-Leaf Displays

    • A method to display numeric data visually without losing detail; useful for datasets that aren't overly large.

    Example Data: Length of Children's Movies
    • Data points: 83, 88, 120, 64, 69, 71, 76, 74, 75, 76, 75, 75, 79, 80, 78, 78, 83, 77, 71, 83, 80, 73, 72, 82, 74, 84, 90, 89, 81, 81, 90, 79, 92, 82, 89, 82, 74, 86, 76, 81, 75, 75, 77, 70, 75, 64, 73, 74, 94

    • Stem: Tens place; Leaf: Units place

      • Example display would look like:

      • 6 | 4 9

      • 7 | 0 1 1 2 3 3 4 4 5 5 5 5

    Analysis of Stem-and-Leaf Displays
    • The longest leaves indicate the highest frequencies, similar to histograms.

    • Shapes derived from the display can also indicate symmetry or skew; a display can be skewed right if there’s a longer tail on one side.

    Expanded Stem-and-Leaf Displays

    • Method for visualizing data trends by breaking stems into higher (values 0-4) and lower (values 5-9) portions for clarity.

    • Example:

      • High stem and low stem arrangement for marathon time data.

    Comparative Stem-and-Leaf Displays

    • Also called back-to-back plots to compare two data sets.

    • Centers the stem and separates leaves for each group.

    Example of Two Teams' Basketball Scores

    Team 1

    Leaves

    Team 2

    Leaves

    78

    82 106…

    68

    72 64…

    Dotplots

    • A method to visualize data grouping, where dots represent data points at their respective values.

    • Example Scores: Highlighting differences such as 60 and 124 which stand out amongst the rest.

    Pie Chart

    • Best for categorical data or single-valued classes to display parts of a whole.

    • Example: Survey of 40 students showing majors and their relative frequencies either as percentages or decimals.


Summarizing Data Graphically
Introduction to Data Displays
  • This section covers various ways to visually represent data, specifically:

    • Frequency distribution

    • Dotplot

    • Histogram

    • Pie chart

    • Bar chart

    • Stem-and-Leaf Display

Frequency Distribution
  • Definition: A frequency distribution is a tabular summary of data.

    • Components:

    • Classes: These can be numeric intervals or single numeric values.

    • Raw Frequency: Count of data points within each class.

    • Relative Frequency: Proportion or percentage of data points in each class compared to the total count.

      • Example: If 4 students earned an A in a class with 8 total students:

      • Raw frequency: 4

      • Relative frequency:

        • Fraction: rac48rac{4}{8}

        • Decimal: 0.50

        • Percentage: 50%

      • If instead, we had 100 students:

      • Raw frequency: 4

      • Relative frequency:

        • Fraction: rac4100rac{4}{100}

        • Decimal: 0.04

        • Percentage: 4%

General Format of Frequency Distribution

  • Structure:

    • Class

    • Frequency

    • Relative Frequency
















Construction of Frequency Distribution
  1. Decide on the number of classes represented.

    • Depends on data range (span from smallest to largest).

    • For distinct values class, the number is determined by distinct values.

    • Aim for a balance: too few classes obscures trends, too many classes makes it hard to read.

      • Example involving menu prices ($0-$100) grouped into only two classes ($0-$50 & $50.01-$100) is too broad.

  2. Determine class width.

    • Use the formula: extWidth=racextRangeextNumberofClassesext{Width} = rac{ ext{Range}}{ ext{Number of Classes}}

    • Adjust to make it a round number for readability.

  3. Define where classes begin and end.

    • Start and end with nice round numbers to maintain simplicity.

    • For example, using rounded figures for a dataset of boutique prices.

Example: Boutique Prices

  • Data points: 15, 80, 104, 168, 18, 82, 106, 195, 29, 87, 107, 293, 46, 98, 108, 421, 67, 104, 153, 425

  • Range Calculation: 425−15=410425 - 15 = 410

    • Choice of Classes: 5

    • Calculated width: rac4105=82rac{410}{5} = 82. Adjusted to 80 for ease.

Class Construction for the Example

  • Start the first class at 10:

    1. First class: 10-89

    2. Next classes: 90-169, 170-249, 250-329, 330-409, 410-489

  • Important notes:

    • Classes must have no gaps or overlaps.

Frequency Determination

  • Count of the first class (10-89):

    • Data points: 15, 18, 29, 46, 67, 80, 82, 87 → Raw Frequency: 8

    • Total data points = 20 → Relative Frequency: rac820=0.4=40%rac{8}{20} = 0.4 = 40\%

Additional Example: Lunch Prices
  • Data points with two decimal places: 3.20, 6.25, 9.13, 3.89, 7.56, 9.21, 4.87, 7.81, 11.32, 5.19, 8.00, 12.52, 6.21, 8.21, 16.45

  • Range Calculation: 16.45−3.20=13.2516.45 - 3.20 = 13.25

    • Suggested Width: rac13.255=2.65rac{13.25}{5} = 2.65 → Adjusted to 3.00

Frequency Distribution for Lunch Prices






















Practice Problem: Daily Commute Times

  • Problem: Construct a frequency distribution with 3 classes for the following daily commute times (in minutes), starting the first class at 10: 10, 15, 20, 25, 12, 18, 22, 28, 14, 21, 26, 30, 16, 23, 29.

  • Solution:

    1. Range Calculation: 30−10=2030 - 10 = 20

    2. Calculated Width (3 classes): rac203≈6.67rac{20}{3} \approx 6.67. Adjusted to 7 for ease.

    3. Class Construction (starting at 10, width 7):

      • 10-16

      • 17-23

      • 24-30

    4. Frequency Determination: Data points 10, 12, 14, 15, 16, 18, 20, 21, 22, 23, 25, 26, 28, 29, 30 (Total 15 points)

      • Class 10-16: 10, 12, 14, 15, 16 → Raw Frequency: 5; Relative Frequency: rac515≈0.333=33.3%rac{5}{15} \approx 0.333 = 33.3\%

      • Class 17-23: 18, 20, 21, 22, 23 → Raw Frequency: 5; Relative Frequency: rac515≈0.333=33.3%rac{5}{15} \approx 0.333 = 33.3\%

      • Class 24-30: 25, 26, 28, 29, 30 → Raw Frequency: 5; Relative Frequency: rac515≈0.333=33.3%rac{5}{15} \approx 0.333 = 33.3\%

    Frequency Table for Daily Commute Times:

    Class (minutes)

    Frequency

    Relative Frequency

    10-16

    5

    33.3%

    17-23

    5

    33.3%

    24-30

    5

    33.3%

    Total

    15

    100%

Histograms

  • Definition: A histogram is a graph that displays interval data without gaps between the bars, representing frequencies along the vertical axis.

    • Features:

    • Horizontal axis = Intervals

    • Vertical axis = Frequencies

  • Example Histograms:

    • Histogram for Boutique Prices

    • Histogram for Lunch Prices

Analyzing Symmetry and Skew

  • Symmetric histogram: Higher in the center, tapering equally on both sides.

  • Skewed histogram: Clustered heavily on one end; direction of the tail indicates skew.

    • Example: Boutique prices are skewed right; lunch prices are also skewed right but less severe.

Practice Problem: Histogram Interpretation

  • Problem: A histogram illustrating the distribution of personal wealth in a country reveals that the majority of citizens possess modest wealth, while a small fraction holds immense fortunes. How would you describe the skew of this histogram?

  • Solution: This histogram would be skewed right. The largest number of data points (modest wealth) would be concentrated on the left side, and a prolonged "tail" extending towards the right would represent the few individuals with significantly greater wealth.

Single Value Frequency Distributions
  • Sometimes values are distinct enough to use single values as classes.

Example: Number of TVs in Households


  • Data points: 4, 6, 2, 1, 2, 3, 2, 1, 3, 2, 2, 4, 5, 1, 5, 2, 3, 3, 5, 2, 2, 3, 1, 1


  • Frequency Table:





































    • Definition: A bar chart visually represents single values (stand-alone) with gaps between bars.

    • Similar to histograms but the horizontal axis features individual values instead of intervals.

      • Example Bar Chart for TVs: Showing the frequency distribution with right skew evident.

    Categorical Data Displays
    • Similar to single-valued classes, categorical data can utilize values as classes.

    • Example: Majors of students in statistics class.

    Frequency Table for Majors






















    Considerations for Categorical Data Display

    • Ordering Issues: Categorical data does not have a natural order, thus discussions of symmetry skew in bar charts are inappropriate.

    • If data has implied ranking (e.g., military rank), it becomes meaningful to discuss symmetry.

    Stem-and-Leaf Displays
    • A method to display numeric data visually without losing detail; useful for datasets that aren't overly large.

    Example Data: Length of Children's Movies

    • Data points: 83, 88, 120, 64, 69, 71, 76, 74, 75, 76, 75, 75, 79, 80, 78, 78, 83, 77, 71, 83, 80, 73, 72, 82, 74, 84, 90, 89, 81, 81, 90, 79, 92, 82, 89, 82, 74, 86, 76, 81, 75, 75, 77, 70, 75, 64, 73, 74, 94

    • Stem: Tens place; Leaf: Units place

      • Example display would look like:

      • 6 | 4 9

      • 7 | 0 1 1 2 3 3 4 4 5 5 5 5

    Analysis of Stem-and-Leaf Displays

    • The longest leaves indicate the highest frequencies, similar to histograms.

    • Shapes derived from the display can also indicate symmetry or skew; a display can be skewed right if there’s a longer tail on one side.

Practice Problem: Stem-and-Leaf Construction

  • Problem: Construct a stem-and-leaf display for the following test scores: 65, 72, 78, 81, 83, 85, 88, 90, 92, 95, 68, 70, 75, 80, 84, 87, 91, 93, 96, 74.

  • Solution:

    • Stem: Tens place; Leaf: Units place

    • 6 | 5 8

    • 7 | 0 2 4 5 8

    • 8 | 0 1 3 4 5 7 8

    • 9 | 0 1 2 3 5 6

    Expanded Stem-and-Leaf Displays
    • Method for visualizing data trends by breaking stems into higher (values 0-4) and lower (values 5-9) portions for clarity.

    • Example:

      • High stem and low stem arrangement for marathon time data.

    Comparative Stem-and-Leaf Displays
    • Also called back-to-back plots to compare two data sets.

    • Centers the stem and separates leaves for each group.

    Example of Two Teams' Basketball Scores













    Dotplots
    • A method to visualize data grouping, where dots represent data points at their respective values.

    • Example Scores: Highlighting differences such as 60 and 124 which stand out amongst the rest.

    Pie Chart
    • Best for categorical data or single-valued classes to display parts of a whole.

    • Example: Survey of 40 students showing majors and their relative frequencies either as percentages or decimals.