CLASSIFICATION OF STATISTICAL DATA
The data collected and kept in raw form is of does not provide any scientific
and valuable information
The collected data goes through;
1. editing process (rectifying the errors, missing and mismatching
responses)
2. coding process
Then taken into data classification and tabulation process
● The objectives of classifications are;
○ To condense the large mass of data into a comprehensible and
accessible form
○ To eliminate unwanted details from the data collected
○ To understand the realised values and frequencies of different variables
under study
○ To help draw more insights on the potential interrelationship of
different variables
○ To facilitate tabular presentation and statistical analysis of the data
● The primary rules of classification are:
● There should not be any ambiguity in the classifications therefore
inclusion of an item into a class is doubtless,
● Even though classes in special cases can be of unequal in width,
generally, the attempt must be made to construct classes of equal width
while constructing class intervals
● While preparing class intervals, class limits coinciding with raw data
items are generally avoided
● Classification must be exhaustive therefore no raw data is left out
● Classification must suit the objective of the investigation
● Classes must be of the sort that the items included in a class have
certain level of commonality.
TYPES OF CLASSIFICATION OF STATISTICAL DATA
Generally, there are four types of classification of data such as;
● Chronological classification
● Geographical classification
● Qualitative classification
● Quantitative classification
● Chronological Classification
Chronological classification enables the examination of the changes in the
values of the variables over time
● Geographical Classification
If the data collected are linked to any spatial element it can then be
classified with respect to it.
● Qualitative Classification
The data collected can be classified in qualitative terms or based on
attributes.
Attributes are the qualities of a variable.
In case of a variable ‘colour’ its attributes can be any color like;
red, blue, green etc
In case of variable ‘Occupation’ may have its attributes like;
farmer, soldier, teacher, businessman etc.
Different types of qualitative classification
● Simple classification
In this classification, we can classify data based on the attributes of one
variable in one row and with another row of its frequencies.
Education: Primary Highschool Others
Frequency x x x
● Two fold classification
In this classification, we can classify data based on the attributes of two
variables.
Education: Primary Highschool Others
Domicile Rural Urban Rural Urban Rural Urban
Frequency x x x x x x
● Manifold classification
In this classification, we can classify data based on the attributes of more
than two variables.
Education: Primary Highschool Others
Domicile: Rural Urban Rural Urban Rural Urban
Gender F M F M F M F M F M F M
Frequency x x x x x x x x x x x x
● Quantitative Classification
In case of variables on height, weight, Income, temperature etc, they
assume numerical values and are termed as quantitative variables, hence,
the classification is called classification by variables.
● Variables are of two types:
1. Continuous variables
2. discrete variables (discontinuous variables)
● The variable which can assume both integer and fraction values are termed
as continuous variables.
○ Weight and height of people in a study area can be considered as
continuous as the variable takes both integers and fractions.
● If variable takes only integer values they are normally termed as discrete
variables
○ the data collected on number of times households availed medical
facility in six months time period take only whole number
● Both of discrete and continuous variables do not possess any natural
classification like qualitative data
● Their classification will have only numerical values either continuous or
discrete
● They are assigned artificial classification either in the form of Class
intervals (continuous variables) or in terms of specific values or inclusive
classes ( discrete variables)
● Class intervals are the ranges of numerical values in which observations of
the variables are classified according to the numerical values they assume
In these classifications only two things are involved, that is, the variable
in terms of class intervals and the number of times the values in such
intervals are occurring, which is termed as frequencies
● Class intervals can be framed in two ways:
○ Exclusive Class intervals
○ Inclusive class intervals
● Class intervals can be framed in two ways:
○ Exclusive Class intervals
The boundaries of a class intervals is termed as class limits
The lower boundary is termed as lower limit
The upper boundary is termed as upper limit
If in the classification the observations above and equal to the lower limit and
below the upper limit ( excluding the upper limit value), it is termed as
exclusive class intervals
Exclusive class interval
● Class intervals can be framed in two ways:
○ Inclusive Class intervals
If in the classification the observations are grouped in such a
way that each class intervals include the observation which are equal and above
the lower limit and equal and below the upper limit (both limits included) they
are termed as Inclusive class intervals
Inclusive class intervals
STATISTICAL SERIES AND FREQUENCY DISTRIBUTION
A logical arrangement of statistical observations of a variable is termed as
statistical series.
. In this respect the observations can be classified into series of:
1. individual observations, 2. discrete series, [Link] series
In case of series of individual observations, the data is arranged
individually either in ascending or descending order.
It is not very informative in nature as it is mere an orderly presentation of
raw data.
No. tourists visits for 7 days; 100, 50, 50, 33,20, 10, 10
A frequency distribution is a tabular summary of data showing the number of
items in each class or class- intervals
The definition of frequency distribution holds the same for both qualitative and
quantitative data
Thus, based on the nature of values assumed by variable,
1. Ungrouped frequency distribution
2. Grouped frequency distribution (inclusive type classification) for discrete
variables
3. Exclusive type class interval for continuous frequency distribution
Three steps are necessary to define the classes of quantitative type of data:
i) determination of number of classes for the given data
ii) determine the width of the classes
iii) determine class limits
Class limits
It must be chosen so that each data item belongs to one and only one class.
The lower limit identifies the smallest possible data value assigned to the class.
The upper class limit identifies the largest possible data value assigned to the
class.
In case of qualitative data, it is not necessary to specify class limits as each data
item naturally fall into separate class.
But with quantitative data class limits are necessary to determine where each
value belongs
Converting working class into actual class (inclusive class to exclusive class)
Inclusive type classification is also known as working class or normal class and
exclusive type classification is known as actual class.
Steps of converting working class into actual class are as follows;
1. Obtain difference between the upper limit of the previous class and
lower limit of the following class
2. Divide the difference by 2
3. Subtract the value obtained in step 2 from the lower limit of each
class and add the same value to the upper limit of each class