Advanced Statistics

Frequency Distributions and Graphs

• When data are collected in original form, they classes and rounding off to the nearest
are called raw data. whole number.
o Select a starting point (usually the
Frequency Distribution
lowest value or any convenient number
• A frequency distribution is the organization less than the lowest value, multiples of
of raw data in table form, using classes and class size);
frequencies. o Add the width to get the lower-class
• Each raw data value is placed into a limits.
quantitative or qualitative category called a o Find the upper-class limits.
class. 2. Tally the data.
• The frequency of a class is the number of data 3. Find the numerical frequencies from the
values obtained in a specific class. tallies.
• Three Types of Frequency Distributions
1. Categorical Frequency Distribution • Rules in constructing grouped frequency
2. Ungrouped Frequency Distribution distribution:
3. Grouped Frequency Distribution 1. There should be between 5 and 20 classes.
Categorical Frequency 2. It is preferable but not absolutely necessary
that the class width be an odd number.
• A categorical frequency distribution is used for 3. The classes must be mutually exclusive.
data that can be placed in specific categories, 4. The classes must be continuous.
such as nominal- or ordinal-level data. 5. The classes must be exhaustive.
6. The classes must be equal in width.
• Procedure for constructing a frequency
distribution for categorical data.
• Open-ended distribution is a frequency
1. Make a table.
distribution with an open-ended class.
2. Tally the data.
3. Count the tallies and place the results in
• Frequency distribution enable the researcher to
frequency column.
see the nature of the data more easily than by
4. Find the percentage of values in each class
𝑓
looking at the raw data, especially when there
by using the formula: % = ∙ 100% where are a large number of data values.
𝑛
f = frequency of the class and n = total
number of values. • Frequency distribution will be analyzed by
o The decimal equivalent of percent is looking for peaks and extreme values.
called a relative frequency. o The peaks show which class or classes
5. Find the totals for frequency and percent have the most data values compared to
column. the other classes.
o Extreme values, called outliers, show
Grouped Frequency Distribution large or small data values that are
relative to other data values.
• When the range of the data is large, the data • Class Boundaries or Exact Class Limits – The
must be grouped into classes that are more than upper and lower values of a class for a grouped
one unit in width, in what is called a grouped frequency distribution whose values have one
frequency distribution. additional decimal place more than the data and
end in the digit 5.
• Procedure for Constructing a Grouped
Frequency Distribution • Class Midpoint or Class Mark – A value for a class
1. Determine the classes. in a frequency distribution obtained by adding
o Find the highest and lowest value. the lower- and upper-class boundaries or the
o Find the range. (R = HS - LS) lower and upper limits and dividing by 2.
o Select the number of classes desired.
o Find the class width or class size (i) by • Cumulative Frequency – The sum of the
dividing the range by the number of frequencies accumulated up to the upper
boundary of a class in a frequency distribution.
 • Relative Frequency – The quotient of the • Statistical graphs can be used to describe the
frequency of the class and the total number of data set or to analyze it.
values. • Graphs are also useful in getting the audience’s
attention in a publication or a speaking
• Cumulative-Relative Frequency – The sum of presentation.
the relative-frequencies accumulated up to the
• Graphs are can be used to discuss an issue,
upper boundary of a class in a frequency
reinforce a critical point, or summarize a data
distribution.
set.
Ungrouped Frequency Distribution • Graphs are can also be used to discover a trend
or pattern in a situation over a period of time.
• When the range of the data is relatively small, a • The three most commonly used graphs in
frequency distribution can be constructed using research are as follows:
a single data values for each class. This type of 1. Histogram
distribution is called an ungrouped frequency 2. Frequency Polygon
distribution. 3. Ogives or Cumulative Frequency Graph

• Procedure for Constructing an Ungrouped Histogram

Frequency Distribution.
1. Determine the classes. • A graph that displays the data by using
o Note: If the data are continuous, class contiguous vertical bars (unless the frequency of
boundaries can be used. a class is zero) of various heights to represent the
2. Tally the data. frequencies of the classes.
3. Find the numerical frequencies from the • Steps
tallies. 1. Draw and label the x and y axes.
2. Represent the frequency on the y-axis and
The reasons for constructing a frequency distribution: the class boundaries on the x-axis.
o to organize the data in a meaningful, 3. Draw vertical bars for each class.
intelligible way; • Histograms show the overall shape of the
o to enable the reader to determine the distribution of the variable.
nature or shape of the distribution; • As with frequency distributions, look for classes
o to facilitate computational procedures with the largest frequency (peaks), any gaps in
for measures of average and spread; the data, and the presence of any outliers.
o to enable the researcher to draw charts • Distribution Shapes
and graphs for the presentation of data; 1. The Bell Shaped or Mound-Shaped
and 2. The Uniform or Rectangular Shaped
o to enable the reader to make 3. The J-Shaped
comparisons among different data sets. 4. The Reverse J-Shaped
5. The Positively or Right Skewed Shaped
6. The Negatively or Left Skewed Shaped
Data Presentation 7. The Unimodal, Bimodal, Multimodal Shaped
8. The U-Shaped
• Methods of Presenting Data:
• When analyzing histograms, look at the shape of
1. Textual form – where the data are
the curve.
presented in paragraph form or in sentence
o Does it have one peak or two peaks or
form.
more?
2. Tabular form – where the data are
o Is it relatively flat, or is it U-shaped?
presented in row and columns.
o Are the data values spread out on the
3. Graphical form – where the data are
graph, or are they clustered around the
presented in pictorial or visual form.
center?
o Are there data values in the extreme
• The purpose of graph in statistics is to convey the
ends?
data to the viewers in pictorial form.
o Are there any gaps in the histogram?
• It is easier for most people to comprehend the
o Are the data clustered at one end or the
meaning of data presented graphically than data
other, indicating a skewed distribution?
presented numerically in tables or frequency
distributions.
Frequency Polygon Other Types of Graphs

• A graph that displays the data by using lines that • Dot Diagram – A dot plot, also called a dot chart
connect points plotted for the frequencies at the or strip plot, is a type of simple histogram-like
midpoints of the classes. The frequencies are chart used in statistics for relatively small data
represented by the heights of the points. sets where values fall into a number of discrete
• Steps bins (categories).
1. Find the midpoints of each class. • Bar Chart – A bar chart or bar graph is a chart or
2. Draw the x and y axes. Label the x-axis with graph that represents categorical data with
the midpoint of each class, then use a rectangular bars with heights or lengths
suitable scale on the y-axis for the proportional to the values that they represent.
frequencies. The bars can be plotted vertically or horizontally.
3. Using the midpoints for the x values and the A vertical bar chart is sometimes called a line
frequencies as the y values, plot the points. graph.
4. Connect adjacent point with line segment. • Pictograph or Pictogram – A form of bar graph in
which stylized, easily recognizable figures are
Ogive or Cumulative Frequency Graph
used in place or rectangular bars.
• A graph that represents the cumulative • Pie Graph – A circle that is divided into sections
frequencies for the classes in a frequency or wedges according to the percentage of
distribution. frequencies in each category of the distribution.
• Steps Used to show the relationship between the parts
1. Find the cumulative frequency for each class. and the whole.
2. Draw the x and y axis. Label the x-axis with o Steps
the class boundaries. Use an appropriate 1. Determine the angle sector of the
scale for the y-axis to represent the frequency for each class. D = (f/n)(360)
cumulative frequencies. 2. Each frequency must also be converted
3. Plot the cumulative frequency at each to a percentage. % = (f/n)(100)
upper-class boundary. 3. Using a protractor and a compass, draw
4. Connect adjacent points with line segments. the graph using the appropriate degrees
Then extend the graph to the first lower class found in step 1, and label each section
boundary. with the name and percentage.
• Time Series Graph – Represent data that occur
Relative Frequency Graph over a specific period of time. Used to show a
pattern or trend that occurs over a period of
• The histogram, the frequency polygon, and the
time.
ogive shown previously were constructed by
• Pareto Chart – Used to represent frequency
using frequencies in terms of the raw data.
distribution for a categorical variable, and the
These distributions can be converted to
frequencies are displayed by the heights of
distributions using proportions instead of raw
vertical bars, which are arranged in order from
data as frequencies. These types of graphs are
highest to lowest. Used to show frequencies for
called relative frequency graphs.
nominal or qualitative variables.
• Steps
o Purpose
1. Convert each frequency to a proportion or
1. To display the relative importance of
relative frequency by dividing the frequency
data.
for each class by the total number of
2. To direct efforts to the biggest
observations. (decimal form)
improvement opportunity by
2. Find the cumulative relative frequencies.
highlighting the vital few in contrast to
3. Draw each graph, for the histogram and
the useful way.
ogive use the class boundaries along the x-
o The Pareto principle (also known as the
axis. For the frequency polygon use the
80/20 rule, the law of the vital few, or the
midpoints on the x-axis. The scale on the y-
principle of factor sparsity) states that, for
axis uses proportion.
many events, roughly 80% of the effects
come from 20% of the causes.
 o Steps o Nonlinear Relationship exist when the
1. Arrange the data from the largest to points fall in a curved line. The relationship
smallest to frequency. is described by the nature of the curve.
2. Find the percent of each class and
compute the cumulative percent.
3. List the classes on the horizontal axis (x-
axis) of a graph from highest to lowest.
Label the left vertical axis (y-axis) with
the frequency, then label the right
vertical axis with cumulative
percentages. Draw in the bars for each
class.
4. Draw a line graph of the cumulative
percentages. The first point on the line
graph should line up with the top of the
first bar.

Stem-and-Leaf Plots

• A combination of sorting and graphing.

• A data plot that uses part of the data value as the
stem and part of the data value as the leaf to
form group or classes.
• Steps
1. Arrange the data in order.
2. Separate the data according to the first digit.
3. A display can be made by using the leading
digit as the stem and the trailing digit as the
leaf.
• Double Stem-and-Leaf Plots
• Back-to-Back Stem-and-Leaf Plots

Scatter Plot

• A graph of order pairs of data values that is used

to determine if a relationship exists between the
two variables.
• Steps
1. Draw and label the x and y axes.
2. Plot the points of pairs of data.
3. Analyze the scatter plot.
• Interpretation of Scatter Plot
o Positive Linear Relationship exists when
the points fall approximately in an
ascending straight line and both the x and y
values increase at the same time. The
relationship then is that as the values for
the x variable increase, the values for the y
variable are increasing.
o Negative Linear Relationship exists when
the points fall approximately in a descending
straight line from left to right. The
relationship then is that as the x values are
increasing, the y values are decreasing, or
vice-versa.
o No Relationship exists when there is no
discernable pattern of the points.