You are on page 1of 29

Business Statistics for Decision Making

Unit 1
Chapter 3

Complied by Rooshabhkumar Mehta


DATA
CLASSIFICATION,
CHAPTER 3
TABULATION AND
PRESENTATION
Requisites of ideal classification
Requisites of ideal classification

Homogeneity: columns or rows where categories are appearing


supposed to be relevant each time. Say for example, under age title
all the figures should talk about age of respondents only. Under
gender all categories related to gender should appeared only.

Suitability: Data (information) supposed to be aligned with


objective. If the Table designed to show migration reasons then it
should talk about that only.

Stability: If there is no major change in the perspective/ objective,


data should not be classified. This may change your analysis and
interpretation.

Classification should not be rigid.


Basic Classification of the data

Geographical classification
Chronological Classification
Qualitative Classification
Quantitative Classification
Source: https://data.gov.in/catalog/list-msme-registered-units-under-udyam-till-last-
date?filters%5Bfield_catalog_reference%5D=6871862&format=json&offset=0&limit=6
&sort%5Bcreated%5D=desc

Data according to geographical location

Example: Number of MSMEs


Geographical classification
Chronological Data: Time unit wise data
Example: Time Series Data Rainfall by Seasons in Tamil Nadu
: SHB 2020
Qualitative Classification: Data as per qualitative variables (with categories).
Example: census data on Migration as classified by gender

This Photo by Unknown Author is licensed under CC BY


Quantitative Classification: any table classified
according to quantitative categories of variables
(continuous observations)

Example: Age-group wise classification og migrants in


india
Data Visualisation
– Charts and Graphs
Age in years (Ungrouped Data)
Ungrouped Data: Raw data 25 65 34 32 38 38 42
23 34 26 32 54 37 60
that have not been summarized in any way 21 25 34 30 38 43 38
23 23 27 35 60 45 54
25 22 35 35 54 45 60

Grouped Data: Data that have 24


24
24
22
33
29
33
32
66
43
60
39
54
46
been organized into a frequency distribution 23 25 29 27 43 39 43
25 32 29 30 45 42 45
Class Interval Mid Value Frequency 23 34 34 29 40 38 42
33 32 34 29 54 55
20-30 25 34 25 30 30 35 34 39
25 30 32 34 40 39
30-40 35 41 23 30 34 32 40 38
40-50 45 13 23 34 34 35 43 38
Frequency
50-60 55 10 50 41
40 34
60-70 65 2 30
20 13 10
10 2
Sum 100 0

20-30 30-40 40-50 50-60 60-70


The following data 42 70 64 47 66 69 73 38 48 25 55 85 10 24
represent the 45 31 62 47 63 84 16 40 81 15 35 17 40 36
afternoon high 44 17 38 79 35 36 23 64 75 53 31 60 31 38
temperatures for 52 16 81 12 61 43 30 33
50 construction
days during a year
in Peren.

Question 1 : Construct a frequency distribution


for the data using five class intervals.
Question 2 : Construct a frequency distribution
for the data using ten class intervals
One possible 5 class frequency distribution:

Class Interval Frequency


10 - under 25 9
25 - under 40 13
40 - under 55 11
55 - under 70 9
70 - under 85 8
50

One possible 10 class frequency distribution:

Class Interval Frequency


10 - under 18 7
18 - under 26 3
26 - under 34 5
34 - under 42 9
42 - under 50 7
50 - under 58 3
58 - under 66 6
66 - under 74 4
74 - under 82 4
82 - under 90 2
Which one is better ???
5- class Grouping
OR
10 Class Grouping
The ten-class frequency distribution gives a
more detailed breakdown of temperatures,
pointing out the smaller frequencies for the
higher temperature intervals. The five-class
distribution collapses the intervals into
broader classes making it appear that there
are nearly equal frequencies in each class
Hey This sort of
Ungrouped to
Group
transformation
can be done
suing Tally
Mark Method
This Photo by Unknown Author is licensed under CC BY-SA
Bar Charts & Pie Charts for
Frequency Distribution

Education Percentage Cumulative


Background Frequency Frequency

B.Tech 16 8% 16
BBA 35 18% 51
60
55
BCA 43 22% 94
50
B.Sc 55 28% 149 43
40
BA 26 13% 175 35

30 26
B.Com 25 13% 200 25

20 16
200
10

0
B.Tech BBA BCA B.Sc BA B.Com
Relative and Cumulative Frequency
(With Excel Demonstration)

Relative Cumulative
Income Group Mid Value Frequency Frequency Frequency
(Percentage)
Less than Rs 20,000 Rs 10,000 27 8.43 % 27
Rs 20,000 to Rs 40,000 Rs 30,000 46 14.35 % 73
Rs 40,000 to Rs 60,000 Rs 50,000 76 23.75 % 149
Rs 60,000 to Rs 80,000 Rs 70,000 39 12.18 % 188
Rs 80,000 to Rs 100,000 Rs 90,000 98 30.62 % 286
Rs 100,000 to Rs 120,000 Rs 110,000 34 10.62 % 320
120
320
98
100

80
76

60
46
39 34
40
27
20
14% 23% 38% 20% 49% 17%
0
Less than Rs 20,000 Rs 20,000 to Rs 40,000 Rs 40,000 to Rs 60,000 Rs 60,000 to Rs 80,000 Rs 80,000 to Rs 100,000 Rs 100,000 to Rs 120,000
Quantitative Charts

• Histogram

• Frequency Polygon

• Ogive

• Scatter plots….

• Box Plot Charts


Histograms
A histogram is a useful tool for differentiating the
frequencies of class intervals.

It can be drawn with bars

Construction of a histogram involves labeling the x-axis


with the class endpoints and the y-axis (ordinate) with the
frequencies.
Histogram
120

100

80

60

40

20

0
Less than Rs 20,000 Rs 20,000 to Rs 40,000 Rs 40,000 to Rs 60,000 Rs 60,000 to Rs 80,000 Rs 80,000 to Rs 100,000 Rs 100,000 to Rs 120,000
Frequency Polygons

• Construction of a frequency polygon begins


by scaling class midpoints along the
horizontal axis and the frequency scale
along the vertical axis.
• A dot is plotted for the associated
frequency value at each class midpoint.
• Connecting these midpoint dots completes
the graph.
Ogives
Ogives are most useful when the decision maker wants to
see running totals.
Class end points are taken on X- Axis.

Cumulative frequencies can be considered on Y –Axis.

Co-ordinates are presented with dots and then connected


subsequently by line passing through them.
Cross-
Tab
Analysis
• Frequency Polygon/Histogram is a bar chart used for checking
frequency distribution across classes.

• Ogive is useful when you want to see category size (in terms of its
frequency) based on running total.

• Bar chart can be used for the continuous Data (with frequencies)
for checking distribution pattern.

• Pie Chart is useful for detecting stack of its category (or


continuous observation) w.r.t to its proportions.
References
Black, K., 2013. Applied business statistics. 6th ed. Hoboken, N.J.: Wiley.
Keller, G., 2014. Statistics for management and economics. 10th ed. stamford, USA: Cengage Learning.
Kumar, R., 2011. Research methodology. 3rd ed. london: SAGE.
Malhotra, N., 2010. Marketing research. 6th ed. Boston: Pearson.

You might also like