Definitions

Statistics is a numerical statement of facts in any department of enquiry placed in relation to each other. -Bowley Statistics are the classified facts representing the conditions of the people in a State specially those facts which can be stated in numbers or any tabular or classified arrangement. -Webster Statistics can be defined as the aggregate of facts affected to a marked extent by multiplicity of causes, numerically expressed, enumerated or estimated according to a reasonable standard of accuracy, collected in in a systematic manner, for a pre-determined purpose and placed in relation to each other. -Secrist Statistics is the science of collecting, organizing , analyzing, interpreting and presenting data.

SCOPE OF STATISTICS
1.Social Sciences
-Man Power Planning -Crime Rates -Income & Wealth Analysis of Society -In studying Pricing, Production, Consumption, Investments & Profits etc.

2.Planning
-Agriculture -Industry -Textiles -Education etc. For ex. Five Year Plans in India.

SCOPE OF STATISTICS contd

3. Mathematics
-Extensive use of Differentiation, Algebra, Trigonometry, Matrices etc in modern business analysis. -Statistics now treated as Applied Mathematics.

4. Economics
- Family Budgeting
-Applied in solving economic problems related to production, consumption, distribution of products as per income & wealth related patterns, wages, prices, profits & individual savings, investments, unemployment & poverty etc.

SCOPE OF STATISTICS contd

5. Business Management
- Trend Analysis - Market Research & Analysis - Product Life Cycle

i)

Marketing
Marketing Policy Decisions depend on forecasting, demand analysis, time & motion studies, inventory control, investments & analysis of consumer data for production & sales.

SCOPE OF STATISTICS contd

ii) Production
- Designs - Methods of Production - Technology Selection - Quality Control Mechanisms - Product Mix - Quantities - Time Schedules for Manufacturing &

Distribution

SCOPE OF STATISTICS contd

iii) Finance
-Correlation Analysis of profits & dividends, assets & liabilities -Analysis of income & expenditure - Financial forecasts, break-even analysis, investment & risk analysis

iv) Sales
-Demand Analysis -Sales Forecasts

v) Personnel
- Wage plans, Incentive plans, Cost of living, Labor turnover ratio, Employment trends, Accidental Rates, Performance Appraisals etc.

SCOPE OF STATISTICS contd

vi) Accounting & Auditing
-Analysis of Income, Expenditure, Investment, Profits and Optimization of Production etc - Forecasting costs of production & price

vii) Other Areas

-Insurance, Astronomy, Social Sciences, Medical Sciences, Psychology, Education etc.

LIMITATIONS OF STATISTICS

Does not study individual items, deals with aggregates. Statistical laws are not exact. Not suitable for the study of qualitative phenomenon. Statistical methods are only means and not end for solving problems.

ROLE OF STATISTICS IN MANAGEMENT DECISIONS

A. Marketing & Sales
- Product selection & competence strategies - Utilization of resources including territory control - Advertising decisions for cost & time effectiveness - Forecasting & trend analysis - Pricing & market research

ROLE OF STATISTICS IN MANAGEMENT DECISIONS contd

B. Production Management
- Product mix & product positioning - Facility & production planning - Distribution management - Material handling & facility planning - Maintenance policies - Activity planning & resources allocation - Quality control decisions

ROLE OF STATISTICS IN MANAGEMENT DECISIONS contd

C. Materials Management
- Buying policy- Sourcing & Procurement - Material Planning & Lead Times D. Finance, Investments & Budgeting - Profit planning - Cash Flow Analysis - Investment decisions - Dividend policy decisions - Risk Analysis - Portfolio Analysis

ROLE OF STATISTICS IN MANAGEMENT DECISIONS contd

E. Personnel Management - Optimum organization level - Job evaluation & assignment analysis - Social / habit analysis - Salary / wage policies - Recruitment & Training

ROLE OF STATISTICS IN MANAGEMENT DECISIONS contd

F. Research & Development - Area of thrust Analysis & Planning - Project Selection Criteria\ - Alternatives analysis - Trade off analysis - cost & revenue

ROLE OF STATISTICS IN MANAGEMENT DECISIONS contd

G. Defense - Optimization of weapon system - Force deployment - Transportation Cost Analysis - Assignment Suitabilities

Definitions Continued

Observations: Numerical quantities that measure specific characteristics. Examples include height, weight, gross sales, net profit, etc.

Some More Definitions

Raw Data: Data collected in original form.

Classes / class intervals: Subgroups within a set of collected data. Ex.10-20,20-30 etc Width of class-interval = upper limit lower limit Mid Value = (U.L + L.L)/2 Frequency: The number of times a certain value or class of values occurs.

Frequency Distribution Table: The organization of raw data into table form using classes and frequencies.

More Definitions

Cumulative Frequency of a class is the sum of the frequency of that class and the frequencies of all the preceding or succeeding classes which are listed in some sensible order (numerical order, alphabetical order, etc.)

Illustration Individual Series

Marks of ten students of a class in Statistics 15, 35, 55, 67, 78, 84, 79, 90, 89, 94

Illustration Discrete Frequency Distribution

Height
(in inches)

No. of Students 12 18 10 6 4

60 62 64 66 68

Illustration Grouped or Continuous Frequency Distribution

Exclusive Type Class Intervals

ClassIntervals 20-25 25-30

Frequency

8
2

30-35 35-40
40-45

40 23
9

Illustration Grouped or Continuous Frequency Distribution contd

Inclusive Type Class Intervals

ClassIntervals 1-10 11-20

Frequency 2 6

21-30
31-40 41-50

10
15 12

CONVERSION OF INCLUSIVE TYPE CLASS-INTERVALS TO EXCLUSIVE TYPE CLASS INTERVALS

1.

Calculate ADJUSTMENT FACTOR as follows:

A.F= Lower Limit of Next C.I Upper Limit of Previous C.I 2 using the given inclusive type class intervals. 2. Obtain new class intervals as follows: New Lower Limit = Old Lower limit A.F New Upper Limit = Old Upper Limit + A.F

CONVERSION OF INCLUSIVE TYPE CLASS-INTERVALS TO EXCLUSIVE TYPE CLASS INTERVALS contd

ClassIntervals 1-10 11-20 21-30 31-40 41-50 Frequency A.F = (11 10)/2 = 0.5 For 1st C.I i.e 1-10 New L.L = 1(old L.L) 0.5 = 0.5 New U.L=10(old U.L) +0.5 = 10.5 And so on.

2
6 10 15 12

CONVERSION OF INCLUSIVE TYPE CLASSINTERVALS TO EXCLUSIVE TYPE CLASS INTERVALS contd

ClassIntervals 0.5-10.5 10.5-20.5

Frequency 2 6 Now calculations can be made.

20.5-30.5
30.5-40.5

10
15

40.5-50.5

12

Obtaining Cumulative Frequency Distribution

Class Intervals Frequency Less than type Cum.frequency More than type cum.frequency

20-25

15
34 6 10 8 2

15
15 +34 =49 49 + 6 =55 55 + 10 = 65 65 + 8 = 73 73 + 2 = 75

60 + 15 = 75
26 + 34 = 60 20 + 6 = 26 10 + 10 = 20 2 + 8 = 10 2

25-30 30-35 35-40 40-45 45-50

Introduction to Measures of Central Tendency

Also known as averages. Values show a distinct tendency to cluster or group around a value. This behavior is central tendency of data. The value around which the data clusters is the measure of central tendency which represents the whole set of data.

Objectives of Averages

To find out one value that represents the whole mass of data. To enable comparison. To establish relationship. To derive inferences about universe to which sample belongs. To aid decision making.

Requisites of a Good Average

Should be rigidly defined. Should be mathematically expressed. Should be readily comprehensible & easy to calculate. Should be calculated on the basis of all the observations. Should be least affected by extreme values and sampling fluctuations. Should be suitable for further mathematical treatment.

Common Measures of Central Tendency

Arithmetic Mean Geometric Mean Harmonic Mean Median Mode Partition Values like Deciles ,Quartiles & Percentiles.

Averages

Mathematical Averages

Positional Averages

A.M

G.M

H.M

Median

Mode

Arithmetic Mean

Individual Series = x1 + x2 + + xn n For ex. A.M of 3, 6, 24 and 48 = 3 + 6 + 24 + 48 4 = 81/4 = 20.25 Ans.

Arithmetic Mean contd

Discrete Frequency Distribution = f1x1 + f2x2 + ..fnxn = fx N Where N = f1 +f2 ++fn n = no. of observations

Freq.

fx

x1
x2 x3

f1
f2 f3

f1x1
f2x2 f3x3

x4

f4

f4x4

Illustration Height (in inches) X No. of Students f

fX

60 62 64 66 68

12 18 10 6 4 50 = N

60 x 12 = 720 1116 640 396 272 3144 = fx

= 3144 / 50 = 62.88 Ans.

Arithmetic Mean contd

Continuous Frequency Distribution - Direct Method - Assumed Mean Method - Step Deviation Method

Arithmetic Mean Formulae

Direct Method
= f1x1 + f2x2 + ..fnxn = fx N f Where N = f1 +f2 ++fn x = mid value of a C.I = (U.L + L.L) 2

Arithmetic Mean Formulae contd

Assumed Mean Method = A + fd N Where A = assumed mean N=f d=xA x = mid - value

Arithmetic Mean Formulae contd

Step Deviation Method = A + fd x i N where A = assumed mean N=f d=xA i x = mid value i = width of C.I = U.L L.L

Illustration Direct Method

C.I Freq f MidValue X fX = fx f

4-6 6-8 8-10 10-12

6 12 17 10

5 7 9 11 13

30 84 153 110 65 442 = fx

= 442/50 = 8.84 Ans.

12-14 5 Total 50 = f

Illustration Assumed Mean Method

C.I 10-15 Freq. f 2 Mid Values (x) 12.5 d =(x-A) -10 fd -20 = A + fd f

15-20
20-25 25-30 30-35 35-40

7
9 8 6 4 f= 36

17.5
22.5 = A 27.5 32.5 37.5

-5
0 5 10 15

-35
0 40 60 60 fd = 105

= 22.5 + 105 36 = 22.5 + 2.916 = 25.416 Ans.

Illustration- Step Deviation Method

C.I Freq.(f) MidValues d= (x-A) (x) I (i= 5) 10-15 15-20 20-25 25-30 30-35 35-40 200 700 900 800 600 400 f= 3600 12.5 17.5 22.5 = A 27.5 32.5 37.5 -2 -1 0 1 2 3 -400 -700 0 800 1200 1200 fd = 2100 fd = A + fd x i f

= 22.5 + 2100 x 5
3600 = 22.5 + 2.916 = 25.416 Ans.

Illustration
Marks X or more Cum. Freq.

C.I
10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100

Freq.
140-133= 7 133-118=15 118-100=18 100-75=25 75-45=30 45-25=20 25-9=16 9-2=7 2-0=2
Proceed as usual

10 20 30 40 50 60 70 80 90 100

140 133 118 100 75 45 25 9 2 0

What if
C.I 50-59 40-49 Frequency 1 3

30-39 20-29
10-19 0-9 Total

8 10
15 3 N=40

A.F = (L.L of 1st C.I U.L of 2nd C.I)/2 = (50-49)/2 = 0.5 New C.I L.L of new C.I = L.L of original C.I A.F U.L of new C.I= U.L of original C.I + A.F For ex. For 1st C.I,new L.L = 50-0.5 = 49.5 new U.L = 59 +0.5 = 59.5 and so on. Now Continue as usual.

Determining missing frequency when A.M is known Illustration Mean = 16.82 Marks Freq. M.V (x) d= (x A)/i fd

0-5
5-10 10-15 15-20 20-25 25-30 30-35

10
12 16 ? = f4 14 10 8 N = 70 + f4

2.5
7.5 12.5 17.5 = A 22.5 27.5 32.5

-3
-2 -1 0 1 2 3

-30
-24 -16 0 14 20 24 fd = -12

Determining missing frequency when A.M is known Illustration Soln. A + fd x I f = 16.82 (given) , I = 5 Hence 16.82 = 17.5 + ( -12 )x 5 70 + f4 = - 0.68 = - 60 70 + f4

- 0.68 (70 + f4) = - 60 f4 = 12.4/0.68 = 18 approx. Ans.

Some More Applications of A.M

Q1.The avg. marks secured by 50 students was 44.Later on it was discovered that a score 36 was misread as 56. Find the correct average marks secured by the students. Soln. Given N = 50 and mean = 44 = X N X = 44N i.e X = 44x55 X = 2200 Since 36 was misread as 56 Hence correct X = 2200 56 + 36 = 2180 Correct mean = 2180/50 = 43.6 Ans.

Combined A.M

Suppose for k different series with n1,n2nk observations each, the respective A.M s are 1,2,.k. Then the A.M of the new series obtained on combining all the n1,n2,nk observations is obtained using the formula: = n11+n22+.+nkk n1+n2+.+nk

Illustration- Combined A.M There are two branches of a Co. employing 100 and 80employees respectively .If A.Ms of the monthly salaries paid by the two branches are Rs.4570 and Rs.6750 respectively, find the A.M of the salaries of the employees of the Co. as a whole. Soln. Given No. of employees in 1st factory, n1 = 100 Avg. Salary of employees in 1st factory, 1 = Rs. 4750 No. of employees in 2nd factory, n2 = 80 Avg. Salary of employees in 2nd factory, 2 = Rs.6750 Avg salary of the employees of the Co. as a whole = 100 x 4750 + 80 x 6750 = 997000 = Rs. 5538.89 100 + 80 180

Practice Questions- Arithmetic Mean

Weekly Income (in Rs.) 20-25 25-30 30-35 35-40 40-45 45-50

Q1.

No.of workers

200

700

900

800

600

400

Weight (in kgs)

30-34 35-39 40-44 3 5 12

45-49 18

50-54 55-59 60-64 14 6 2

Q2

No.of Students

Q3

Wages(in Rs.) No.of workers

125-175

175-225

225-275

275-325

325-375

375-425

425-475

10

25

35

12

10

Practice Questions- Arithmetic Mean contd

Lifetime (in hrs.) No. of tubes 0 20 60 116

Q4

Less than 300 Less than 400 Less than 500 Less than 600

Less than 700

Less than 800 Less than 900 Less than 1000 Less than 1100 Less than 1200

194
265 324 374 392 400

Merits of A.M

Is rigidly defined and has a definite value. Is based on all the observations. Is capable of algebraic treatments for further data analysis & interpretation. Easy to calculate & simple to understand. For a large no. of observations, A.M provides a good basis of comparison.

Drawbacks of A.M

Being based on all the observations, is considerably affected by abnormal observations. For ex. A.M of 1000, 25, 35 & 40 will be (1000+25+35+40)/4 = 275 which is not at all a representative figure. Cannot be calculated even if a single observation is missing. Cannot be obtained just by inspection as in case of median & mode. May give absurd results. For ex. If avg. no. of children per family is to be calculated and the result is 3.4 children per family, how would you interpret it?

Weighted Arithmetic Mean

Formula Used w = x1w1+ x2w2 +.+xnwn w1+ w2 +.+wn

Illustration Weighted A.M

Designation Monthly Salary (in Rs.) (X) 1500 800 500 250 No. of employees (w) 10 20 70 100 wX 15000 16000 35000 25000

Class I Officers Class II officers Subordinate Staff Clerical Staff

Lower Staff

100

150 350 = w

15000 106000 = wX

Illustration Weighted A.M

Weighted A.M = wX w = 106000 350 = Rs 302.857 Ans.

Median Positional Average

The value of the middle term of a series arranged in ascending or descending order of magnitude. Its value is the value of the middle item irrespective of all other values.

Calculation of Median
Individual Series N = no. of observations or items in the series - Arrange all the items in ascending or descending order of magnitude. Case I N = Odd Median = Value at (N+1) th position in 2 the arranged series. Case II N = Even Median = A.M of values at (N, N+1)th 2 2 position.

Calculation of Median Illustration (Individual Series) Ex.1 Find the median 5, 7, 9, 12, 10, 8, 7, 15,21 Solution: Arranging in ascending order we get 5, 7, 7, 8, 9, 10, 12, 15, 21 Here N = 9 i.e odd Hence Md = (N+1) th item in the arranged order 2 = (9 +1) th item 2 = 5 th item = 9 Ans.

Calculation of Median Illustration (Individual Series)

Ex 2. Find the median 10, 18, 9, 17, 15, 24, 30, 11 Solution Arranging in ascending order we get 9, 10, 11, 15, 17, 18, 24, 30 Here N = 8 i.e even Hence Md = A.M of the ( N , N+1)th items in the 2 2 arranged order. = A.M of (4th, 5th) items = (15 + 17) 2 = 16 Ans.

Calculation of Median
Discrete Frequency Distribution (i) Find less than type cum.frequency. (ii) Find N/2.( N = f) (iii) Find the cum.freq. just greater than N/2. Suppose it is C. (iv) Find the corresponding value of X. (the item) This is median.

Calculation of Median-Illustration (Discrete Freq. Distribution) Height (in inches)

60 62 64 66 68

No. of students
12 18 10 6 4

Cum. Freq.
12 30 40 46 50

Here N = 50 (i) N/2 = 25 (ii) Cum. Frequency just greater than N/2 = 30 (iii)Corresponding value of item is 62. Median = 60 Ans.

N = 50

Calculation of Median

Grouped Frequency Distribution

(i) Find less than type cum.frequency.

(ii) Find N/2.( N = f) (iii) Find the cum.freq. just greater than N/2. Suppose it is X. (iv) Look for the cum.freq. preceding X. Find the corresponding class interval.This is median class Formula Used Md = L1 + N/2 - C (L2 L1)
f

Where L1 = L.L of median class L2 = U.L of median class C =cum.freq. of class preceding the median class. f = frequency of median class.

Calculation of Median-Illustration (Grouped Freq. Distribution)

C.I Freq.(f) Cum. Freq

N/2 = 3600/2 = 1800 Cum.freq. just greater than 1800 is 2600. Hence median class is 25-30. Hence L1 = 25 L2 = 30 C = 1800 f = 800 Md = 25 + 1800 - 1800 (30 25 ) 800 = 25 Ans.

10-15 15-20 20-25 25-30

200 700 900 800

200 900 1800 2600

30-35
35-40

600
400 f= 3600

3200
3600

Calculation of Missing Frequencies when median is known : Illustration : Median = 50

Expenditure No. of Families

Cumulative Freq.

0-20
20-40 40-60 60-80 80-100

14
? = f1 27 ? = f2 15 N = 100

14
14 + f1 41 + f1 41+ f1+f2 56 + f1 + f2

Calculation of Missing Frequencies when median is known : Illustration

Here median = 50 N = 100 N/2 = 50 Hence median class 40-60
Md = L1 + N/2 - C (L2 L1) f 50 = 40 + 50 (14 + f1)(60 40) 27 10 = 720 20 f1 27 f1 = 450/20 = 22.5 = 23 families approx. N = 56 + f1 + f2 100 = 56 + 23 + f2 f2 = 21 Ans. f1 = 23 and f2 = 21

L1 = 40 L2 = 60 f = 27 C = 14 + f1

Practice Numericals - Median

Q1. Age 20-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60

No. of Persons 14 28 33 30 20 15 13 7

Q2. Value
Less than 10 Less than 20 Less than 30 Less than 40 Less than 50 Less than 60 Less than 70 Less than 80

Frequency 4 16 40 76 96 112 120 125

Practice Problems- Median

Q3. Determine the missing frequencies. The median is 46.Also determine the A.M.

Class-Intervals Frequency 10-20 20-30 30-40 40-50 50-60 60-70 12 30 ? 65 ? 25

70-80

18
229 = N

Merits - Median

Is rigidly defined. Can be easily calculated. Not affected by extreme values. Can be located merely by inspection.

Demerits - Median

May not represent the entire series in many cases. Not suitable for further algebraic treatment. More likely to be affected by sampling fluctuations.

Mode

The value occurring the largest no. of times in a series. That is the value having the maximum frequency. Is calculated for discrete and continuous frequency distributions only. For ex. How to obtain the mode for 1,2,3,4,5 ? as the maximum frequency is 1 and each observation has frequency 1.

Mode Discrete Frequency Distribution

The value corresponding to maximum frequency is the mode. For ex. The weight 132 pounds has the maximum frequency 3. Hence 130 pounds is the mode for this frequency distribution.

Wt. in pounds 120

No.of students 1

130
132 135 140 141 Total

3
2 2 1 1 10

Mode Continuous Frequency Distribution

1.Look for the class-interval with maximum frequency. This is the modal class. 2. Note down the following: L1 = lower limit of the modal class. i = width of class-interval f0 = frequency of class preceding the modal class. f1 = frequency of modal class. f2 = frequency of class succeeding the modal class.

Mode: Formula for Continuous Frequency Distribution

Mode = L1 + h(f1 f0) 2f1-f0-f2

Empirical Relationship between Mean, Median & Mode

Mode = 3 Median 2 Mean

Geometric Mean

Individual Series G = (x1.x2.x3xn)1/n log G = 1 (logx1 + log x2 +.+ logxn) n G = antilog ( 1 log x) n

Geometric Mean

Discrete Frequency Distribution G = (x1f1.x2f2.xnfn)1/N

log G = 1( f1logx1 + f2logx2 ++xnlogfn) N G = antilog ( 1 filogxi) N

Geometric Mean
Continuous Frequency Distribution - Formula same as in case of discrete frequency distribution with x (as observations) replaced by x (as mid-values)

Harmonic Mean

Reciprocal of A.M of reciprocals - Individual Series H= 1 1( 1 + 1 +..+ 1 ) n x1 x 2 xn

H= n (1 ) x

Harmonic Mean
-Discrete Frequency Distribution H= 1 1( f1 + f2+..+ fn ) N x 1 x2 xn

H= N (fi ) xi

Harmonic Mean
Continuous Frequency Distribution - Formula same as that of Discrete Frequency Distribution with x (as observations) replaced by x (as mid values).