Summarizing and Graphing Data (Introductory Statistics, Ch. 2)

Summarizing and Graphing Data

Why We Graph Data

Graphing data is a fundamental aspect of statistics, allowing for the organization, communication, and interpretation of data sets. Graphs help reveal patterns, trends, and unusual behaviors within the data.

• Organization: Graphs structure data for easier analysis.

• Communication: Visual representations make data more accessible.

• Revelation: Graphs can highlight behaviors and outliers in the data.

Key Terms

• Frequency: The number of times a value occurs in a data set.

• Variation: The degree to which data values differ from each other.

• Distribution: The pattern of data values over the possible range.

• Outliers/Unusual Values: Data points that fall outside the majority of values.

Frequency Distributions

Frequency Distribution Table

A frequency distribution partitions a data set into several classes and lists the number of values in each class. This method does not retain the original data values but provides a summary for analysis.

Units Produced	Frequency
1-10	3
11-20	34
21-30	45
31-40	48
41-50	56
51-60	59
61-70	52
71-80	42
81-90	31

Example: Pulse Rates

Given pulse rates (beats per minute) of 40 females, a frequency distribution can be constructed to summarize the data:

Pulse Rate	Frequency
60-69	12
70-79	14
80-89	11
90-99	1
100-109	1
110-119	0
120-129	1

Parts of a Frequency Distribution

• Frequency: Number of values in each class.

• Lower Class Limits: Smallest value in each class.

• Upper Class Limits: Largest value in each class.

Class Boundaries

Class boundaries are values used to separate classes in a frequency distribution. For example, boundaries for pulse rates might be 59.5, 69.5, 79.5, etc.

Class Midpoints

The midpoint of a class is the average of its lower and upper limits. For example, the midpoint of 60-69 is $(60+69)/2 = 64.5$.

Pulse Rate	Midpoint	Frequency
60-69	64.5	12
70-79	74.5	14
80-89	84.5	11
90-99	94.5	1
100-109	104.5	1
110-119	114.5	0
120-129	124.5	1

Class Width

Class width is the difference between consecutive lower (or upper) class limits. For example, $70 - 60 = 10$.

Constructing a Frequency Distribution

1. Sort the data and determine the number of classes (typically 5-20).

2. Calculate class width: $cw = \frac{\text{max} - \text{min}}{\text{number of classes}}$ If the result is not a whole number, round up.

3. Choose the minimum value as the first lower class limit.

4. List all lower and upper class limits using the class width.

5. Count and enter the frequency for each class.

Example: Frequency Distribution Construction

Given bear weights, construct a frequency distribution with 6 classes and include midpoints.

Relative and Cumulative Frequency Distributions

Relative Frequency Distribution

Relative frequency is the proportion of the total frequency for each class.

• Calculated as: $\text{Relative Frequency} = \frac{\text{Frequency}}{\text{Total}}$

• Example: 12 out of 40 is $\frac{12}{40} = 0.3$ or 30%.

Pulse Rate	Relative Frequency
60-69	30%
70-79	35%
80-89	27.5%
90-99	2.5%
100-109	2.5%
110-119	0%
120-129	2.5%

Cumulative Frequency Distribution

Cumulative frequency is the sum of the current and all previous frequencies.

Pulse Rate	Cumulative Frequency
Less than 70	12
Less than 80	26
Less than 90	37
Less than 100	38
Less than 110	39
Less than 120	39
Less than 130	40

Data Types and Graphs

Quantitative vs. Categorical Data

• Quantitative Data: Numerical values representing counts or measurements.

• Categorical Data: Names or labels that do not represent counts or measurements.

Frequency Distributions for Both Data Types

Weight (lbs) of Wild Bears	Frequency
26-95	5
96-165	6
166-235	7
236-305	1
306-375	4
376-445	2

Color	Frequency
Red	23
Orange	12
Yellow	7
Green	19
Blue	26
Purple	1

Dot Plots

A dot plot displays each data value as a dot along a scale. Dots representing equal values are stacked. Dot plots can be used for both quantitative and categorical data.

Graphs and Charts for Quantitative Data

Histogram

A histogram uses adjacent bars of equal width to represent frequencies of quantitative data classes. The horizontal axis shows classes or midpoints, and the vertical axis shows frequencies. Histograms do not maintain original data values.

Analyzing Histograms

• Histograms reveal the shape of the data distribution.

• Distributions may be symmetric or skewed.

Skewed vs. Symmetric Distributions

• Symmetric: The distribution matches if folded in half horizontally.

• Skewed: The distribution is not symmetric.

Left and Right Skew

• Skewed Left (Negative): Tail on the left side.

• Skewed Right (Positive): Tail on the right side.

Relative Frequency Histogram

Similar to a histogram, but the vertical axis shows relative frequencies instead of raw counts.

Frequency Polygon

A frequency polygon connects points above class midpoints with line segments. A relative frequency polygon uses relative frequencies for the vertical axis.

Ogive

An ogive is a line graph that depicts cumulative frequencies, useful for understanding how frequencies accumulate across classes.

Stemplot

A stemplot separates each data value into a stem (leftmost digits) and a leaf (rightmost digit), providing a quick visual of data distribution.

Graphs and Charts for Categorical Data

Bar Graph

Bar graphs use bars of equal width with gaps between them. The horizontal axis identifies categories, and the vertical axis shows frequencies. Multiple bar graphs compare two or more data sets.

Histogram vs. Bar Graph

• Histogram: No gaps between bars; used for quantitative data.

• Bar Graph: Gaps between bars; used for categorical data.

Pareto Chart

A bar graph with bars arranged in descending order of frequency, often used to highlight the most significant categories.

Pie Chart

Pie charts represent categorical data as slices of a circle, with slice size proportional to frequency.

Best Practices for Graphs

• Include a clear title.

• Label axes and categories.

• Ensure accuracy and clarity.

• Present data at the forefront for interpretation.

Example Applications

• Constructing frequency distributions for animal weights.

• Analyzing pulse rates using histograms and polygons.

• Comparing categorical preferences with bar and pie charts.

Additional info: These notes cover foundational concepts in summarizing and graphing data, including frequency tables, histograms, and categorical charts, as outlined in a typical college-level statistics course.