Professional Documents
Culture Documents
You have studied in Unit 7 that the frequency distributions can be presented in the form of
tables. In fact, frequency distribution can also be presented in the form of graphs. Compared to
tabular presentation, graphs of Cequency distribution are helpful in identifying characteristics
and relationships. Graphs of frequency distribution are also useful in locating the positional
averages such as mode, median, quartiles, etc. Let us study the basic principles to be followed
while preparing the graphs of frequency distribution.
Principles of Constructing Graphs of Frequency Distribution
Graphs of frequency distribution are constructed on the basis of the following principles :
1) The values of the variable are shown on X axis. For example, in a frequency distribution
showing age (in years) of the students, the age will be shown on X axis.
2) The values derived from the frequency of the items/classes are shown o n Y axis.
3) It is not necessary that horizontal axis should start from zero. But vertical axis is not broken
in these graphs. False base line cannot be taken.
4) The scales on the two axes be taken such that they reflect the data significantly. These
scales should be clearly specified.
5 ) The graph should have a concise and self-explanatory title.
9.7.1 Histogram
A histogram is a series of rectangles each proportionate in width to the magnitude of a class
interval and proportionate in area to the number of frequencies concerning the class intervals.
In this graph the items/class intervals are $ken on X axis and values derived from frequency
are taken on Y axis as height of the rectangle. To derive the height, smallest class interval will
be taken as one unit width. The width of all other class intervals will be determined in terms
of this. The frequency will then be divided by this number. The value so obtained is called as
frequency density or adjusted frequency. This represents frequency per unit class interval.
When class intervals are all equal, every class interval will have a width of one unit. SO
dividing frequency (which is represented by the area) by one, we get the height (i.e., frequency
density) equal in number to frequency. So in case of equal class intervals, height may be taken
proportionate to frequency. But actually in concept it is area which is proportionate to the
frequency.
In constructing histogram there should not be any gap between two successive rectangles. If
the data is given in inclusive class interval or in discrete series, it should be first converted to
continuous data with exclusive type class intervals. A histogram helps in determining the mode
of the data, about which you will learn Iater in this course.
Illustration 7
The bllowing distribution relates to the marks secured by the students of a college in
statistics. Present it graphically.
No. of Students : 5 15 25 50 40 30 10 5
Solution
The data in this illustration has been given in the form of continuous frequency distribution
with exclusive type. So the given class intervals, as it is, will be taken on X axis. As all the
class intervals are of the same sue, by denoting their length by one unit, each class interval will
be of one unit width. Dividing frequency (i.e,,area of rectangle) by this width, the height of
the rectangles will be equal to their frequenciic.
~d NO.of Students Width Units Frequency Density
Graphic &.tq
(Area) (Base) (Height)
The height will be taken on Y axis. Now on both X axis and Y axis mark the scale in a
suitablemanner. To construct the histogram. we will construct rectangles on the base bar as
class intervals, with height equal to the value derived as above. The histogram,is drawn and
presentedjn Graph 9.8.
Gnph : 9.8 : Histogram Showing the Marks of Students in Statistics (Class Intervals with Equal Width)
For Class 30-40, the adjusted frequency or frequency density will be = 5/10 x 20 = 10.
For Class 40-60, the adjusted frequency or frequency density will be = 5/20 x 16 = 4.
Let us also calculate the frequency density by the method pointed out in the Illustration 7 and
study whether the two methods give the same value for height q r not.
Graph 9.9 : Histogram Showing the Marks of Students in Accountancy (Class Intervals with Unequal Width).
No. of Students Width Units Frawncy Density Craphk Presenlpbion
Mnrb
(Arm) (Base) (lFi@gbh:)
Illustration 9
Draw a histogram for the following data relating to the age (in 'years on the nearest birthday)
for members of a club.
Age in Years : 10-14 15-24 25-39 ' 40-59
No. of Members : 5 12 24 12
Solution
In this illustration the class intervals are of unequal size, and are given in inclusive form.
Converting them in 'exclusive' form to have no gaps in between and taking the smallest class
interval size (i.e., 5) as one unit width, the data is as follows :
Age in Years : 9.514.5 14.5.24.5 24.5-39.5 39.5-59.5
No. of Members :
(Area) 5 12 24 12
Width Units
(Base) 1 2 3 4
Frequency Density :
(Height) 5 6 8 3
AGE I N YEARS
Coktion, cladtkatiw and Look at Graph 9.10 carefully the histogram for the data is presented. In this graph one big
?resentatlon of Data square is taken to represent 5 years. The points 9.5, etc., are marked on a thick line to make
the plotting easy. A break is shown on X axis near the origin to indicate that this part of the X
axis is not on the same scale as the rest of'the axis.
9.7.2 Frequency Polygon
Frequency polygon is another way of showing frequency distribution graphically, There are
two ways of constructing a frequency polygon. In the first method, along with the given
class intervals, two more class intervals with zero Frequencies are taken one at the beginning of
the given classes and the other at the end. Then the frequency densities are plotted against the
mid points of all the class and they are joined by straight line. Another way of preparing a
frequency polygon is to construct a histogram in a.nomal way. Here put a dot against the mid
points of the tops and the first and last vertical lines of the histogram. Then join the different
dots by straight line segments to get the frequency polygon. These methods point out that the
frequency polygon originates at a point lying to the left of the lower limit of the first classinterval
equal to one-haIf of the width of the class interval. Similarly, it is extended at the point lying at
half-the class interval width away to the right of the upper boundary of the last class interval.
It should be noted that the area under the two histograms and frequency polygon is always the
same and they represent the total frequency.
Illustration 10
Draw a histogram and a frequency polygon for the following data :
Class Interval : 10-20 20-30 30-40 40-50 50-60 60-70
Frequency : 10 . 20 30 25 10 5
Solution
The calculations for consttucting histogram and frequency polygons are as follows :
10-20 10 I 10 15
20-30 20 1 20 25
30-40 30 I 30 35
40-50 25 I 25 45
BY HISTOGRAM BY DIRECT
C L A S S IN TE R V AL S C LA S S INT ER VAL S
----.- .
f i e two class intervals, one before and one after the given class intervals, will be 0-10 and Graphic PresentaHon
70-80,since there are no items for these two class intervals. The frequencies for them will be taken
as zero. Now the frequency polygon is constructed by two methods : 1) By using histogram
and joining the mid points of the top (i.e., A, B, C , D, E, F, G and H). 2) By direct method i.e.,
"sing mid points of given class intervals and two extra class intervals, and joining them (i.e.,
A, B, C, D, E, F, G and HI.
Study Graph 9.1 1 carefully. You will note the following points :
1) In cases of histogram, height at various points within a class intervil is same throughout
the class interval, while in frequency polygon it is changing. This means that in a histogram
frequencies are uniformly distributed within a class interval, while in a frequency pQlygdn
frequency density is different at different points.
2) The frequency polygon shows a gradual rate of rise and fall of frequency density, while in
histogram it suddenly jumps at the end points of the class'intervals.
Thus, histogram is used when rise and fall of frequencies from one group to another are to be
shown permanently, and when gradual rise and fall are to be shown then frequency polygon is
used.
Illustration 11
Draw a Frequency polygon for the following distribution.
Frequency : 10 20 30 40 20 10 5
Solution
In this illustration, the frequency polygon can easily be constructed on the basis of the mid
points. The length of the first class interval is the difference between two successive mid points
i.e. 12.5 - 7.5 = 5. Similarly, the length of class interval is same throughout the data. So the
mid points of two class intervals, one before and one after the given data, will be 2.5 and 42.5,
and the frequency density will be numerically the same as the given frequencies. Study
/carefullythe frequency polygon presented in Graph 9.12.
Age (in Years) , : 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80.
No, of Residents : 150 300 500 800 loo0 900 400 100
Solution
In this case all the class intervals are equal. We can construct histogram in the same manner as
described earlier. Frequency polygon can now be drawn with the help of histogram. Now
study Graph 9.13 carefully. You will notice that the first vertical line of histogram is the Y axis
itself. So when we join the mid point of this first vertical line with the mid point of the top of
the fmt rectangle, the polygon line will go to negative side of X axis. Normally, this implies
that there are negative items in the data. In fact, there are no negative values in the data.
~ctually,frequency polygon is allowed to go on negative side so that areas of histogram and
polygon are the same and represent the total frequency. To draw the frequency curve we have
to follow the general pattern of the frequency polygon and draw a smooth line which is
sometimes below the polygon and sometimes above it so that the areas of polygon and the
curve are equal. In a case like the present one, adjustment for negative side is done by starting
the frequency curve from '0'.
Gmph 9.13 : HUogram, Frequency Polygon and Frequency Curve for Age of Residents
9.7.4 Ogive or Cumulative Frequency Graph Graphic Ekscntation
Graph 9.14 : Less than Ogive Showing the Marks of Students in Accountancy
Collection, QassiRwtion and Solution
Presentation of Data First, we have to convert the above distribution 'into "less than" and "more than" cumuiative
frequency distributions. The two cumulative frequency distributions have been constructed
below :
Construction of Cumulative Frequency Distribution
0-10 . 10 10
,. 150
Now study Graph 9.14 carefully. To draw less than ogive, less than cumulative frequencies are Graphic Presentation
plotted at upper limits of the various class intervals. So the points (10,10), (20,30), (30,65),
(40,95), (50,115), (60,130), (70,140), and (80,150) are plotted as points A, B, C, D,E, F, G,
and H. Every plotted point represents the number of students having marks below a particular
upper limit of the class interval. If you look at the limits of class intervals, you will find that
the points' have been plotted for all the values except the lower limit of the first class interval
(i.e., '0').There are no students having marks below 0. So to plot points for all the given class
limits, the point (0,O) is also plotted as point P. Now all the points are joined by straight lines.
Study Graph 9.15 carefully. The more than ogive is drawn by plotting points corresponding to
more than cumulative frequencies against the lower limits of the various class intervals and the
(i.e., 80). Look at Graph 9.16 which shows both the "less than" and "more than" ogives in the
same graph.
Graph 9.16 : Leu; than Ogive and More than Ogive Showing the Marks of Students in Accountancy
You will notice from Graph 9.16 that the "more than ogive" and "less than ogive" intersect
at a point which corresponds to 50%of the frequency. In this case that point is 150/2 or 75
frequency.This will be always the case.
Collection, Chssiflcatlon and Check Your Progress B
Presentation of Data
1) Distinguish between a histogram and a historigram.
3) How do you adjust the frequencies for constructing a histogram, when the class intervals
have unequal width?
.......................................................................................................................
.......................................................................................................................
.......................................................................................................................
4) What is a frequency curve?
.......................................................................................................................
.......................................................................................................................
......................................................................................................................
.......................................................................................................................
5) Distinguish between a "less than" and "more than" ogives?
Graphic presentation renders comparison of data easier, helps in establishing trends of past
performance and makes it possible to determine positional averages. Graphs are prepared on
the basis of coordinated system of plotting points and joining them by lines. Graphs are of two
types : 1) graphs of time series, and 2) graphs of frequency distribution.
Graphs of time series, also called Historigram, depicts chronological data. Time, being the
independent variable, is always taken on X axis and the dependent variable is taken on Y axis.
The Y axis or vertical axis normally starts with zero but can also be broken and a false base
line can be taken.
This type of graph can either be prepared for one dependent variable or more than one
dependent variable. A historigram can also be prepared for negative values. In case two
dependent variables arc given with two different units of measurement, the data can be
represented by means of mixed graphs. Two vertical axes are taken in this case and two curves
are drawn on the bases of these axes for the two variables. Another type of historigram is a
range graph which is prepared to show the range of data. This graph is prepared to show how
the extreme values (i.e., maximum and minimum values) of a dependent variable are changing
with time.
Graphs of frequency distribution represent the data of frequency distribution. In this type of
graph the value of a variable is taken on X axis and the values derived from frequencies on Y
axis. Vertical axis is not broken in this type of graph. Graphs of frequency distribution are
classified as : 1) histograms, 2) frequency polygon, 3) frequency curve, and 4) ogive.
A histogram is a series of rectangles each proportionate in width to the magnitude of class
interval and proportionate in area to the frequency pertaining to that class interval. When class
intervals are all equal, the height of the rectangles will also be numerically proportionate to the
frequency. In case the class intervals are unequal the height will be proportional to frequency
density.
A frequency polygon is constructed by plotting the frequencies density against the mid points
of classes. It can also be prepared first by preparing a histogram and then by putting dots
against the mid points and then joining these dots by straight lines. It is extended on both the
sides of histogram. The area under a histogram and a frequency polygon is always the same. A
frequency curve is prepared by free hand to smoothen the frequency polygon. It is
generalisation of frequency polygon.
An ogive or a cumulative frequency graph depicts cuniulative frequencies. We can either
prepare a "less than" ogive when cumulative frequencies are in ascending order or a "more
than" ogive, when cumulative frequencies are in descending order. An ogive helps us to locate
median and other partition values.
- --
9.9 KEY WORDS
Adjusted Frequency or Frequency Density : Represents frequency per unit class interval.
False Base Line : Js taken by breaking y-axis in case of a historigram.
Frequency Curve : A curve constructed to smoothen the frequency polygou.
Frequency Polygon : A graph of frequency distributiod constructed by plotting the
frequencies density against the mid-points of the class.
Histogram : A graph of frequency distribution where rectangles are awn with area
proportionate to the frequency of a class interval and the class interval a5 the base.
Cdsection, CLasdfication and Historigram : A graph of time series.
b n t a t i o n of Dats
Mixed Graph : A graph constructed to show the two dependent variables with two different
units of measurement.
Ogive : A grzph of frequency distribution depicting cumulative frequencies.
Originating Point : The poini 9f Inter-section of X axis and Y axis.
Range Graph : A graph showing the range of data between two extreme values of a variable
at different points of time.
X axis : The horizontal axis foi plotting points.
Y axis : The vertical axis for plotting points.
-
9.10 ANSWERS TO CHECK YOUR PROGRESS
Questions
I) Describe the rules of graphic presentation of data.
2) Discuss the importance of graphic presentation of data.
3) What is a graph of time series? Discuss the principles of constructing a graph of time series.
4) What is a graph of a frequency distribution? Discuss the principles of construction of
graphs of frequency distribution.
5) Describe the different types of graphs based on frequency distribution.
6) Explain different types of graphs of time series.
Exercises
I) The following data relates to th'e number of man-days lost in a workshop on account of
power frlilure. Show it by means of a suitable graph.
2) Draw a historigram for the following data relating to the financial outlay on education
during the Various Five Year Plans.
(1 n crorer)
lJopul;~tio~l 27.9 31 9 36.1 43.9 54.8 68.5
Drav: a suitable graph for the above data.
4) Draw a suitable graph for the following data relating to the balance of trade of a country :
I3;ilancc of l ' r d e
('000 h) 4 3 15. I1 --4 -2 13 I It;