UNIT TWO

DATA MANAGEMENT AND REPRESENTATION

Data are a set of facts and provide a partial picture of reality. Whether data are being collected
with a certain purpose or collected data are being utilized, questions regarding what information
the data are conveying, how the data can be used, and what must be done to include more useful
information must constantly be kept in mind.

Since most data are available to researchers in a raw format, they must be summarized,
organized, and analyzed to usefully derive information from them. Furthermore, each data set
needs to be presented in a certain way depending on what it is used for. Planning how the data
will be presented is essential before appropriately processing raw data.

First, a question for which an answer is desired must be clearly defined. The more detailed the
question is, the more detailed and clearer the results are.

Sources and Data Collection

The type of data collected and how the associated sampling takes place depend on the statistical
question asked.

There are a number of aspects of data collection that need to be considered when carrying out a
statistical investigation. Draw the frequency distribution table and used it to analyse data
3. Explain and use appropriate text, graphs and tables to summarise your data.

• Choose the questions (e.g. "Do you like …?", "How tall is …?") that will determine the
type of data collected (e.g. categorical, numerical).
• Choose a sample size that will allow confidence in the conclusion.
Use repeated sampling to demonstrate how much variation occurs in samples of different
sizes.
• Avoid bias in data collection.
Explore sources of bias in various contexts where data are collected.
• Understand the importance of random selection.
Explore various methods of random sampling.
• Know the relationship between samples and populations.
Learn to define populations and samples in various statistical contexts.

Methods of Data Collection:

a. Obtrusive data collection methods that directly obtain information from those
being evaluated e.g. interviews, surveys, focus groups, observation, case study,
questionnaires.
b. Unobtrusive data collection methods that do not collect information directly from
examinees. e.g.: Document analysis, observation at a distance.
Data Collection Tools: Participatory Methods. Records and Secondary Data, Observation,
Surveys and Interviews, Focus Groups, Diaries, Journals, Self-reported Checklists, Expert
Judgment, Delphi Technique. Other Tools. -scales (weight), tape measure, stopwatches, and
chemical tests: i.e. quality of water, - health testing tools: i.e. blood pressure. -citizen report
cards.

DATA MANAGEMENT TECHNIQUES

Data Management is concerned with “looking after” and processing data it involves: Looking after
field data sheets, entering data into computer files, checking and correcting the raw data preparing
data for analysis, documenting and archiving the data and meta-data. Data Management is the
consolidation of data (and meta-data) in a way that is easy to manipulate, retrieve and maintain.

WHY is Data Management Important?

• Ensures data for analysis are of high quality so that conclusions are correct
• Good Data Management allows further use of the data in the future and enables efficient
integration of results with other studies.
• Good Data Management leads to improved processing efficiency, improved data quality,
improved meaningfulness of the data.

Data Management Problems

• Lack of skills –the inability to use software or set up data checking procedures
• Multiple copies of files
• No one with responsibility for checking data
• No clear policy on archiving or making data available
• Lack of documentation
• Multiple entries of the same data
• Hand pre-processing of data
Basically, three data management techniques will be discussed in this unit namely:

Data trimming is the process of removing or excluding extreme values, or outliers, from a
data set. Data trimming is used for a number of reasons and can be accomplished using various
approaches. As social scientists, communication researchers often work with data sets that may
require the removal of outliers to strengthen a statistic and accomplish a number of research
goals. It is important to understand the impact outliers can have on data and the approaches
available to eliminate or censor these extreme values without compromising the data set.

Data Winsorizing or winsorization is the transformation of statistics by limiting extreme values

in the statistical data to reduce the effect of possibly spurious outliers. Winsorization is a way to
minimize the influence of outliers in your data by either:

• Assigning the outlier a lower weight,

• Changing the value so that it is close to other values in the set.
 Basic Method to Winsorize by Hand

1. Analyse your data to make sure the outlier isn’t a result of measurement error or some
other fixable error.
2. Decide how much Winsorization you want. This is specified as a total percentage of
untouched data. For example, if you want to Winsorize the top 5% and bottom 5% of data
points, this is equal to 100% – 5% – 5% = 90% Winsorization. An 80% Winsorization
means that 10% is modified from each tail area.
3. Replace the extreme values by the maximum and/or minimum values at the threshold.

For example:

• The following data set has several (bolded) extremes:

{0.1,1,12,14,16,18,19,21,24,26,29,32,33,35,39,40,41,44,99,125}
Mean = 33.405.
• After modifying the top and bottom 10% (I matched those values to the nearest
extreme):
{12,12,12,14,16,18,19,21,24,26,29,32,33,35,39,40,41,44,44,44}
80% Winsorized mean = 24.95.
Note that winsorizing is not equivalent to simply excluding data, which is a simpler procedure,
called trimming or truncation, but is a method of censoring data. In a trimmed estimator, the
extreme values are discarded; in a winsorized estimator, the extreme values are instead replaced
by certain percentiles (the trimmed minimum and maximum).

Bootstrapping is any test or metric that relies on random sampling with replacement.
Bootstrapping allows assigning measures of accuracy (defined in terms of bias, variance,
confidence intervals, prediction error or some other such measure) to sample estimates. This
technique allows estimation of the sampling distribution of almost any statistic using random
sampling methods. Generally, it falls in the broader class of resampling methods.

Bootstrapping is the practise of estimating properties of an estimator (such as its variance) by

measuring those properties when sampling from an approximating distribution. One standard
choice for an approximating distribution is the empirical distribution function of the observed data.
In the case where a set of observations can be assumed to be from an independent and identically
distributed population, this can be implemented by constructing a number of resamples with
replacement, of the observed data set (and of equal size to the observed data set).

A bootstrap sample is a smaller sample that is “bootstrapped” from a larger sample.

Bootstrapping is a type of resampling where large numbers of smaller samples of the same size
are repeatedly drawn, with replacement, from a single original sample.

For example, let’s say your sample was made up of ten numbers: 49, 34, 21, 18, 10, 8, 6, 5, 2, 1.
You randomly draw three numbers 5, 1, and 49. You then replace those numbers into the sample
and draw three numbers again. Repeat the process of drawing x numbers B times. Usually, original
samples are much larger than this simple example, and B can reach into the thousands. After a
large number of iterations, the bootstrap statistics are compiled into a bootstrap distribution.
You’re replacing your numbers back into the pot, so your resamples can have the same item
repeated several times (e.g. 49 could appear a dozen times in a dozen resamples).

Bootstrapping is loosely based on the law of large numbers, which states that if you sample over
and over again, your data should approximate the true population data. This works, perhaps
surprisingly, even when you’re using a single sample to generate the data.

• An empirical bootstrap sample is drawn from observations.

• A parametric bootstrap sample is drawn from a parameterized distribution (e.g. a normal distribution).

PRESENTATION OF DATA

Ungrouped vs. Grouped Data

Data can be classified as grouped or ungrouped. Ungrouped data are data that are not organized,
or if arranged, could only be from highest to lowest or lowest to highest.
Grouped data are data that are organized and arranged into different classes or categories. This
refers to the organization of data into tables, graphs or charts, so that logical and statistical
conclusion can be derived from the collected measurements.

Principles of Data Presentation

(a) To arrange the data in such a way that it should create interest in the reader’s mind at first sight.
(b) To present the information in a compact and concise form without losing important details.
(c) To present the data in a simple form so as to draw the conclusion directly by viewing at the
data.
(d) To present it in such a way that it can help in further statistical analysis.

Data may be presented in (3 Methods):

A. Textual
B. Tabular
C. Graphical.

Textual Presentation
Data can be presented using paragraphs or sentences. It involves enumerating important
characteristics, emphasizing significant figures and identifying important features of data.
➢ The data gathered are presented in paragraph form.
➢ Data are written and read.
➢ It is a combination of texts and figures.

Example:
Of the 150-sample interviewed, the following complaints were noted: 27 for lack of books in the
library, 25 for a dirty playground, 20 for lack of laboratory equipment, 17 for a not well-maintained
university building.
Example.
You are asked to present the performance of your section in the Statistics test. The following are
the test scores of your class:
34 42 20 50 17 9 34 43
50 18 35 43 50 23 23 35
37 38 38 39 39 38 38 39
24 29 25 26 28 27 44 44
49 48 46 45 45 46 45 46
Solution First, arrange the data in order for you to identify the important characteristics. This can
be done in two ways: rearranging from lowest to highest or using the stem-and-leaf plot.
Below is the rearrangement of data from lowest to highest:
9 23 28 35 38 43 45 48 17 24 29 37 39
43 45 49 18 25 34 38 39 44 46 50 20 26 34
38 39 44 46 50 23 27 35 38 42 45 46 50

OR
In the Statistics class of 40 students, 3 obtained the perfect score of 50. Sixteen students got a
score of 40 and above, while only 3 got 19 and below. Generally, the students performed well in
the test with 23 or 70% getting a passing score of 38 and above.

Graphical presentation kinds of graphs or diagrams

1. Bar graph – used to show relationships/ comparison between groups
2. Pie or circle graph- shows percentages effectively
3. Line graph – most useful in displaying data that changes continuously over time.
4. Pictograph – or pictogram. It uses small identical or figures of objects called isotopes in making
comparisons each picture represents a definite quantity.
a graphical method can be grouped into quantitative and qualitative data.
For quantitative: Histogram
Frequency polygon
Frequency Curve
Line Chart
Normal distribution cure
Cumulative distribution curve
Scatter diagram

For qualitative: Map diagram

Bar chart
Pie chart
Pictogram

Bar Chart/Graph

Data that are from nominal scales or categorical and are represented in graphic form with
the use of bar graphs. Bar graphs give a pictorial description of the data and emphasize how groups
compare with one another. They are used to compare the sizes of the various parts. The height of
the bars is the basis for the comparisons and not the area of the bars. Data is presented in the form
of a rectangular bar of equal breadth. Each bar represents one variant /attribute. The suitable scale
should be indicated and the scale starts from zero. The width of the bar and the gaps between the
bars should be equal throughout. The length of the bar is proportional to the magnitude/ frequency
of the variable. Bar graphs are either column or horizontal. Column graphs are more popular in
education. Column bar graphs are simple, compound (multiple) or component. Examples are
shown below.

Year Wise enrollment of students in government

school
350
300
Number of students

250
200
150
100
50
0
One Two Three Four Five Six Seven Eight Nine
Year

Figure 2 is a compound column bar graph showing school enrolment at Ayeduase Basic School
by gender.

Component Bar Chart

It is known as a composite or stacked bar chart. It is used when a set of data combines to form a
total. The total is the length/height of the bar. It allows for visual comparisons between different
components ie how components contribute to the total of the category.

140

120

100
Enrolment

80

60

40

20

0
1A 1B 1C 2A 2B
Forms

male female

Figure 2: School enrolment at Ayeduase Basic School by gender.

 Constructing bar graphs/charts
1. Draw two axes, a vertical and horizontal. Label the vertical axis by the source of the
values/scores e.g. enrolment, points etc. Label the horizontal axis by the names of the
categories.
2. Divide vertical scale by points considering the lowest value and the highest value. Choose
appropriate scales such that the bars are not too tall or too short and must start with zero.
3. Construct equally wide and equally spaced bars for each category with the height of the bar
being the value/score for the category on the horizontal axis, which has the names of the
categories as the label.
4. Where computer software such as Microsoft Excel and SPSS are not available, it is
recommended that graph sheets be used.
5. Shade/colour the bars to differentiate bars and components.

Strengths and Limitations

1. Bar charts are easy to draw

2. Values can be read easily from the vertical axis.
3. Comparisons can be made easily and the significance of the information is easily grasped.
4. It cannot be effective with interval and ratio scales of measurement
5. Extreme values distort comparisons especially if some bars are too short and some very tall.
6. For component bar graphs, too many subgroups make the graph crowded.

Uses

Teachers can use bar graphs in several ways. Enrolment by classes, courses and subjects and
inter-house competitions can be represented by bar graphs.

Pie Chart
Pie charts use nominal or categorical data. Pie charts are represented in the form of a circle of
3600 sliced into the shape of ‘pies. Each pie is cut from an angle at the centre of the circle. The
angle corresponds to the data for each category or group. Pie charts give a pictorial view and the
contributions of the parts that make a whole. An example is shown below.

Table 1: Performance in Jackson SRC games in Ashanti Region

Centre Total Points Degrees To calculate each degree:
Louis 120 72 For Louis Centre
KASS 100 60 120
X 3600 =72
Agogo 130 90 600

Akrokerri 80 36 For KASS centre

Mampong 170 102

 Total 600 100
X 3600 =60
360 600
Figure3: Performance in
Jackson SRC games in
Ashanti Region

Constructing pie charts

1. Calculate the degree
equivalents for the value of
each category/group by
dividing the total point for
each group by the overall
total points and multiple the
result by 3600.
For example, for Louis
centre above we have:
120
 3600 = 72 0 and
600
for KASS centre, we have
100
 3600 = 60 0
600
2. Use a pair of compass and protractor to draw the circle and the sectors based on the degrees
calculated.
3. Shade/Colour the sectors to differentiate one from the other.

Strengths and limitations

1. Individual parts of the whole are seen and can be compared.
2. It provides a visual impression of the proportion that each part contributes to the overall total.
3. Angles are harder to compare.
4. They are not easy to draw especially where statistical software and computers are not available.
5. Data that are in continuous form and of ratio and interval scales are not appropriate.
6. The values of each component cannot be read from the chart but must be provided.
7. It is not useful where there are many parts as these parts become too small.
8. It only gives a visual impression but not the details of the data.

Uses
Pie charts can be used by teachers and educational practitioners for examination results by the
number of passes in various subjects, school enrolment by class, form or subjects.

Line graphs

Data that are related to time are best used for line graphs. Time could be days, weeks, months
and years. Line graphs show changes in the data over a period of time. Data from interval and
 ratio scales are most appropriate. Line graphs could be simple or compound. Simple line
graphs give a pictorial description of the data. Compound line graphs compare group data over
a period of time.

Examples are shown below.

Table 2: Attendance at monthly teachers’ workshops

Month Total
January 120
February 85
March 100
April 150
May 90
June 85
July 100
August 60
September 90
October 75
November 100
December 150

Figure 4 is a simple line graph showing attendance at a monthly teachers’ workshop.

160
140
Number of attendants

120
100
80
60
40
20

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Figure 4: Attendance at monthly teachers’ workshops

Table 3: Attendance at monthly teachers’ workshops

Attendance
Month Female Male
Jan 60 60
Feb 45 40
March 60 40
 April 80 70
May 50 40
June 40 45
July 50 50
August 25 35
Sep 40 50
Oct 30 35
Nov 50 50
Dec 90 60

Compound line graph showing attendance at a monthly teachers’ workshop by gender.

10
9
8
7
Attendance

6
Female
5 Male

4
3
Figure 5: Attendance at monthly teachers’ workshops by gender
2
1
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Months
Constructing line graphs
1. Draw two axes, a vertical and horizontal. Label the vertical axis by the source of the
values/scores e.g., attendance, enrolment, points etc. Label the horizontal axis by the time
period e.g., months, days, weeks etc.
2. Divide vertical scale by points considering the lowest value and the highest value. Choose
appropriate scales such that the graph is not too tall or too flat and must start with zero.
3. Plot the value/quantity for each time period on the graph and join all the points by a straight
line.
4. Where computer software such as Microsoft Excel and SPSS are not available, it is
recommended that graph sheets be used.

Strengths and Limitations

1. Values can be read easily from the vertical axis.
2. Comparisons can be made easily and the significance of the information easily grasped.
3. They are less appropriate for nominal scale data.
4. Graphs are distorted where there are extreme values.
 Uses
Teachers and educational practitioners can use line graphs in several ways. Examination results
over a period of years in a subject, total school enrolment as well as enrolment by subjects and
courses for a period of time can be represented by line graphs.
Tabulation
Tables are the devices, that are used to present the data in a simple form. It is probably the first
step before the data is used for analysis or interpretation.

General principles of designing tables

a) The tables should be numbered e.g., Table 1, Table 2 etc.
b) A title must be given to each table, which should be brief and self-explanatory. c) The
headings of columns or rows should be clear and concise.
d) The data must be presented according to size or importance chronologically, alphabetically,
or geographically.
e) If percentages or averages are to be compared, they should be placed as close as possible.
f) No table should be too large
g) Most of the people find a vertical arrangement better than a horizontal one because it is
easier to scan the data from top to bottom than from left to right
h) Footnotes may be given, where necessary, providing explanatory notes or additional
information.

Types of tables
1) Simple tables: Measurements of a single set are presented
2) Complex tables: Measurements of multiple sets are presented

Simple Table

When characteristics with values are presented in the form of a table, it is known as a
simple table e.g. Table Infant mortality rate of selected countries in 2004.

Table 5: The Infant mortality rate of selected regions in Ghana in 2004

Name of region Infant mortality rate
Central 90
Western 60
Eastern 26
Northern 60

Frequency distribution table

In the frequency distribution table, the data is first split up into convenient groups (class interval)
and the number of items (frequency) which occur in each group is shown in adjacent columns.
Hence it is a table showing the frequency with which the values are distributed in different groups
or classes with some defined characteristics.

Rules for construction of frequency table

1) The class interval should not be too large or too small

2) The number of classes to be formed more than 8 and less than 15
3) The class interval should be equal and uniform throughout the classification.
4) After construction of the table, proper and clear heading should be given to it
5) The base or source of data should be mentioned with the pattern of analysis in the footnote at
the end of the table
Example of grouped, relative, and cumulative frequency distributions of serum cholesterol levels
in 200 men.

Table 6: Frequency distributions of serum cholesterol levels in 200 men

Interval Frequency f relative f cumulative f
251- 260 5 2.5 200 0
241- 250 13 6.5 195.0
231- 240 19 9.5 182.0
221- 230 18 9.0 163.0
211- 220 38 19.0 145.0
201- 210 72 36.0 107 0
191- 200 14 7.0 35.0
181- 190 12 6.0 21.0
171- 180 5 2.5 9.0
161- 170 4 2.0 4.0

Features
1. Class. A group of scores.
2. Class interval. The range within which a group of scores lie. It has a number at the beginning
and at the end. E.g., 90- 95.
3. Unequal Class Interval. These result where there are differences in the range of the intervals.
E.g., 91 – 95, 91 - 100
4. Open-ended classes. These are classes with a value at the beginning of the end. e.g. 90 and
above, 45 and below, below 46, above 90.
5. Class limits. The endpoints of a class interval. The smaller number is the lower limit and the
bigger number is the upper limit.
6. Class boundaries. The exact or real limits of a class interval. The lower-class boundaries are
obtained by subtracting 0.5 from the lower-class limit. The upper-class boundaries are
obtained by adding 0.5 to the upper-class limits. A class interval with limits of 91 – 95
produces class boundaries of 90.5 - 95.5
7. Class size/class width. The number of distinct/discrete scores within a class interval. They
are obtained by finding the difference between successive lower-class limits or upper-class
 limits in cases of equal class intervals. They can also be obtained by finding the difference
between successive class marks in cases of equal class intervals or between class boundaries
for each interval.
8. Class mark: The midpoint for each class interval.
9. Frequency: The number of distinct scores from the given data that can be found in a class
interval.
10. Cumulative frequency. The successive sum of the frequencies starting from the frequency of
the bottom class.
11. Cumulative percentage frequency. The successive sum of the percentage frequencies
starting from the percentage frequency of the bottom class. It is also obtained by expressing
each cumulative frequency as a percentage.
12. Relative frequency. It is obtained by dividing each frequency by the total frequency.
13. Cumulative relative frequency. The successive sum of the relative frequencies starting from
the frequency of the bottom class.

Constructing a grouped frequency distribution table

1. Draw a table with four columns with the headings – Class, Class mark, Tally, Frequency.
2. Determine the range i.e., the difference between the highest score and the lowest score. For
example, from the raw scores of the 40 students in the Statistics examination, the highest score
is 93 and the lowest is 49. The range becomes 93 – 49 = 44.
3. Arbitrarily decide on class size. Popular sizes are 3, 5, 7, 10. Odd-numbered class sizes make
computations easier. In Education, the most popular sizes are 5 and 10.
4. Determine the approximate number of classes by dividing the range by the class size. E.g.
44
suppose a size of 5 is taken. The approximate number of classes would be = 8.8 which is
5
rounded to 9 classes. Generally, the number of classes is between 5 and 20.
Alternatively, arbitrarily decide on the number of classes. This is normally between 10 and 20 for
large data sets. Determine the approximate class width by dividing the range by the number of
44
classes. Suppose the number of classes taken is 10. The approximate class size becomes, = 4 .4
10
which is rounded to 5.
5. Identify the highest value/score and write it down.
6. To obtain the topmost class, decide on whether to start with the lower or the upper limit of the
class interval. Determine the closest value to the highest value identified in Step 5 that is a
multiple of the class size. Choose one of the values as a limit and use the class size to determine
the other limit. For example, if class size is 5, and the highest score is 83, then the closest
values are 80 and 85. A possible lower limit is 80 and a possible upper limit is 85. Selecting
the lower limit, the topmost class becomes 80-84 or selecting the upper limit the topmost class
becomes 91-95.
7. Complete the first column with the rest of the classes using equal class sizes, then the second
column with the class marks.
8. Tally the scores in the third column which is the tally column. Take the scores one by one and
place slashes (f) or tally marks in the respective classes. Where the tallies are five, bind them
into one unit to facilitate counting.
9. Count the number of slash or tally marks and write them in the frequency column. Add the
frequencies and put the total at the bottom of the frequency column.

Points to note in constructing Frequency Distributions

1. In Education, the highest-class intervals/classes are at the top so this convention must be
followed in constructing the frequency distribution table.
2. Use mutually exclusive classes. Make sure that an observation falls into one and only class.
Classes must not overlap at the class limits. For example, 50 – 60 and 60 – 70 contain
overlapping class limits of 60.
3. There should be no class with a zero frequency. If this occurs, it is recommended that the class
size is changed. Preferably increase class size.
4. Open-ended classes should be avoided. These classes have only the lower limit if it is the class
at the top, or the upper limit if it is the class at the bottom. For example, 51 and above, 20 and
below.
5. Aim at classes with equal sizes or width. This facilitates the interpretation of the information
from the frequency distribution.
6. The number of classes should not be too small (i.e. not less than 5) and not too large (i.e. not
more than 20). Where the number of classes is less than 5, class size should be reduced but
when the number of classes is more than 20, the class size should be increased.

GRAPHIC REPRESENTATIONS OF FREQUENCY DISTRIBUTIONS

Graphs are useful methods for presenting simple data.

1. They have a powerful impact on the imagination of people.
2. Gives information at a glance.
3. Diagrams are better retained in memory than a statistical table.
4. However, graphs cannot be substituted for a statistical table, because the graphs cannot
have mathematical treatment whereas tables can be treated mathematically.
5. Whenever graphs are compared, the difference in the scale should be noted.
6. It should be remembered that a lot of details and accuracy of original data is lost in charts
and diagrams, and if we want the real study, we have to go back to the original data.

Histogram

Histograms use data from the ratio or interval scale and depend on frequency distributions. It uses
the classes and the frequencies from the frequency distribution table. Used for quantitative,
continuous variables. It is used to present variables which have no gaps e.g. age, weight, height,
blood pressure, blood sugar etc. It consists of a series of blocks. The class intervals are given along
the horizontal axis and the frequency along the vertical axis. An example is shown below.
To construct a histogram
1. Draw two axes, a vertical and horizontal. Label the vertical axis by frequency and the
horizontal axis scores/classes.
2. Select an appropriate scale on the vertical axis considering the highest/largest value. When
using a graph sheet, the scale should be such that the bars are not too tall nor too short.
3. Use class midpoints/marks or class boundaries or class limits to label the points on the
horizontal axis.
4. Drawbars of equal width representing the classes from a frequency distribution table with
corresponding heights as the frequencies.

Importance
1. It gives a pictorial description of the raw data, providing information about the nature of the
data.
2. It gives the direction of performance in terms of academic performance (i.e. skewness).

F 40 F 40
r r
e 30 e 30
q q

20 20

10 10

0 5 10 15 20 25 30 0 5 10 15 20 25 30
 Classes Classes

Skewed to the right Skewed to the left

Group performance tends to be low Group performance tends to be high

3. It provides an estimate of the most typical score. This is the intersection of the two diagonals
of the tallest bar.
Frequency Polygon

Frequency polygon uses data from ratio or interval scales and depends on frequency
distributions. It uses the classes and the frequencies from the frequency distribution table. An
example is shown below.

F
r
e
q

Classes

To construct a frequency polygon;

1. Draw two axes, a vertical and horizontal. Label the vertical axis by frequency and the
horizontal axis scores/classes.
2. Select an appropriate scale on the vertical axis considering the highest/largest value. When
using a graph sheet, the scale should be such that the polygon is not too pointed or too short.
3. Use class midpoints/marks or class boundaries or class limits to label the points on the
horizontal axis.
4. Plot at the midpoint of each class or the midpoint of the histogram the relevant heights as the
frequencies. Join the midpoints with a straight line.
5. Where the line has not touched the horizontal axis, extend the line one class in that direction
so that the polygon touches the horizontal axis.

Importance
1. It gives a pictorial description of the raw data, providing information about the nature of the
data.
2. It provides an estimate of the most typical score. This is the point on the horizontal axis where
the highest point of the polygon is located.
 Most typical score

3. It is used to compare the performance of groups. E.g., Performance in a class test for Forms 1
and 2 can be shown as follows.

Form 1

Form 2

10 20

The diagram shows that Form 2 class, which is more to the right, performs better. The most typical
scores, where the highest point of the polygon is located can be used to confirm the comparisons.
Where the total frequencies are not the same, use relative frequencies in place of the actual
frequencies to draw the polygon.

4. It gives the direction of performance (skewness). Consider three classes, A, B, C.

A B C

Positive skewness Normal Negative skewness

Skewed to the right Skewed to the left
Tends to score low marks Tends to score high marks

Cumulative Percentage Frequency Polygon (Ogive)

Ogives are drawn from frequency distribution tables. Data from ratio or interval scales are
most appropriate.
Plot the graph using the upper-class boundaries of each class against the cumulative percentage
frequencies.

C 100
U 80
M 70
% 60
50
F 40
R 30
E 20
Q 10

0 10 20 30 40 50 60 70 80
CLASSES

To construct an ogive,
1. Obtain cumulative percentage frequencies.
2. Plot the cumulative percentage frequencies in each class on the vertical scale. Choose
appropriate scales, on a graph sheet, such that the ogive is not distorted.
1. Label the horizontal axis as scores or classes.
2. Plot at the upper-class boundary of each class the relevant values of the cumulative frequency.
Join the points with a straight line.
5. Extend the line one class to the left so that the polygon touches the horizontal axis.

Importance
1. It is used for comparisons of distributions of performance especially for distributions where
the class/group sizes are not the same. Generally, the graph that moves more to the right has
better performance. The median score obtained at the cumulative frequency of 50 is also used.
Given the following performances in a test, draw two ogives. Which school performed better?

School A School B
Classes Frequency Cum. % Freq Frequency Cum. % Freq
91 - 100 1 100 7 100
81 – 90 2 99 17 95.3
71 – 80 11 97 30 84
61 – 70 24 86 25 64
51 – 60 20 62 15 47.3
41 – 50 16 42 11 37.3
31 – 40 12 26 19 30
21 – 30 8 14 14 17.3
11 - 20 4 6 6 8
1 - 10 2 2 6 4
Total 100 150

2. It is used to determine percentiles and percentile ranks. Later in the course, you will learn how
to obtain the percentiles and percentile ranks

Box and Whisker Plot

It is used to compare distributions by noting the 10th percentile (P10), first quartile (Q1), median
(Q2), third quartile (Q3), and 90th percentile (P90).

A box and whisker plot is drawn below. Later in the course, you will learn how to obtain the
percentiles and quartiles.

Q1 Q2 Q3
P10 P90

An example.

Assume that the following values were obtained for two classes, Form 1A and Form 1B in a class
test in Mathematics.

P10 Q2 P90
Form 2A 15 42 73
Form 2B 29 56 93
The information is presented below by two box and whisker plots.
Form 2A
 Q1 Q2 Q3
P10 P90

Form 2B
Q1 Q2 Q3
P10 P90

0 25 50 75 100
15 29 42 56 73 93
It can be observed that P10, Q1, Q2, Q3, and P90 values are greater in Form 1B than in Form 2A.
This means that performance is better in Form 1B than in Form 2A.
Also, note that the graph for Form 1B has moved more to the right towards higher values than that
of Form 2A.