statistics

1 HOUR / 2 MARKS
SK AGRAWAL
topics 2
 Introduction to Statistics
 - Data classifications and processing, graphical representation of data,
frequency distributions
 - Measures of central tendency, dispersion and skewness
 - Elementary theory of probability and probability distributions, sampling
and sampling distributions
 - Simple test of significance, regression and correlation, multiple
correlation coefficient
 - Time series
 - Index numbers
What is it 3

Statistics
 is the sub section of Mathematics
 It deals with large quantity of similar numbers
 It also helps us to understand the larger population / data
 It helps to compare different set of similar data
 It helps us to understand the data over a period of time.

Dr Deming was a professor of Maths and Statistics in USA. Japan called him to help improve the
quality of Japanese products. He used STATISTICS for the same.

SIX SIGMA – is also a statistics term.

 4

Sensible
Results

Random Data Analysis

Collate Data 5

 See the data, and have a feel of it.

 Take the smallest value
 Take the largest value
 Make small intervals
 Tabulate. 1,2,1,5,3,8,5,2,4,5,8,4,5,3,6,7,4,2,1,6,9,3,5,4

Minimum = 1, Maximum = 9

Bands
1-3
4-6
7-9
We get the data table 6

Range Tally Number

1-3 7

4-6 9

7-9 6
Different type of data representation 7

Frequency
Distribution

Cumulative
Frequency
Distribution
Graphical Representation 8

Scatter Diagram Histogram

Basic Measurements 9
 Mean (Average) = Total of all values / number of values.
 Median = Central Point of the distribution
 Mode = maximum Point of the distribution.
Formula’s 10

 Means = Sum of all the values / number of values

 Excel command :: Average
 Median – the centre point.
 Put the numbers from low to high.
 Count the number = say n
 n/2 or (n+1)/2 – is the median value
 Excel Command :: Median
 Mode :: The most common point.
 Please see which number maximum no of times.
 In excel command :: Mode
Dispersion 11

Standard Deviation = how much is

the spread of the graph or
distribution.

In a normal distribution
Mean +/- sigma = 68% of data
Mean +/- 2 sigma = 96% of data
Mean +/- 3 sigma = 99.7% of data
Mean +/- 6 sigma = 99.9999% of data
Standard Deviation 12

SD = ((x-mean)/n) ^ 0.5
X = the data
N = number of data

This is quite an important data and is used in all types of data

distributions
Skewness 13

How Much data is shifted towards one side.

Skewness 14
Difference between sigma & skewness 15

Sigma Skewness
Measure how much data is spread Measures if the data is uniform or
not.
Measures amount of variations Measures how much it is left or
right of centre
How much data is distributed Which side it is distributed
Probability 16

Probability is the branch of

mathematics concerning numerical
descriptions of how likely an event is to
occur or how likely it is that a
proposition is true. Probability is a
number between 0 and 1, where,
roughly speaking, 0 indicates
impossibility and 1 indicates certainty.
Probability 17

Chance of an event happening.

Coin Toss Heads Tail
1 ½ ½
2 ½,½ ½, ½
3 ½, ½, ½ ½, ½, ½

No of spin Minimum 1 head

3 HHH, HHT,HTH,THH,HTT,THT,TTH,TTT
4 HHHH, HHHT,HHTH,HTHH,THHH,HHTT…. And so on

Probability of minimum 2 heads in 4 toss is ??

H H H H
H H H T 18
H H T H
H H T T
H T H H
H T H T
H T T H
H T T T
T H H H
T H H T
T H T H
T H T T
T T H H
T T H T
T T T H
T T T T
Probability – one more case 19

 Two dice…
 Probability of.. 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 & 12…. Please calculate
chance of getting a 7.
 20

Throw 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Dice 1 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 6 6 6 6 6 6

Dice 2 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6

Total 2 3 4 5 6 7 3 4 5 6 7 8 4 5 6 7 8 9 5 6 7 8 9 10 6 7 8 9 10 11 7 8 9 10 11 12
Terms of Probability 21

 Mutually Exclusive Events

 ONLY one can happen. Coin toss, Dice throw.
 Probability = Number of favourable events / total possible events.

In this case, the first circle = A+B

Second Circle = B+C
A B C Union = TOTAL of the two = A+B+C
Interception = B
Laws of probability 22

 The addition law

 For mutually exclusive event
 P(A) & P(B) are mutually exclusive
 P(A.U.B) = P(A)+P(B)
 If P(A) & P(B) are not mutually exclusive.
 P(A.U.B)= P(A) + P(B) – P(A.п.B)

 Multiplication Law
 Probability of A happening – provided B happens.
 P(A.I.B)= P(A.п.B)/P(B)
More on probability 23

 Discrete – goes in steps – like toss of a coin

 Continuous – like – temperature in a day. Can have smaller values.

Probability Functions….
• Machine failure
• Accident rates
Probability Distribution 24
 Binomial Distribution
 Very common
 Discrete
 Eg toss of a coin
 Poisson Distribution
 No. of calls in a day / Trucks arriving at a toll booth.
 They are defined by an integer.
 They are discrete random variable.
 The normal distribution
 Continuous distribution
 Good for sampling
 Close to actual life situation.
 One peak, Mean / mode / median at the centre.
Sampling & Test of Significance 25

 Sampling is when – a small set of data is use to predict the outcome

of the full batch.
 Exit Polls – to see the number of seats.
 Destructive Testing – gears, ball bearings

 When we test a sample… is the data good enough representation

of the whole?
 Can we take the decision based on that?
 How much is the error possible?
Sample Selection 26
 Will be decided on
 Uniformity of the over all data – screws manufacturing or election results
 Minimum numbers (Usually 30 plus)
 Criticality of the parameter to be checked.
 Number of parameters to be checked.
 Test of Significance 27
 Does the sample data – represents the full in a reliable way?
 Reliability – is the test of significance.

H0 = null hypothesis – what we expect. Ha = alternative hypothesis – what we want to test and see.
ONLY H0 or Ha – one will be true.
How much variation – will make the either one acceptable?

Example..
Chocolate factory – candy weight 5 gm. QA engineer says that the machine is NOT OK and the data
has changed.
Day 0 – average of 30 p/c – 5.07 gms. Day 4 – average – 5.34 gms , Day 7 average – 6.22 gms.
H0 = 5 gms. Ha – not equal to 5 gms.
Day 0 – may be OK
Day 4 – may be NOT OK
Day 7 – NOT OK.

H0 – is wrong, and Ha is right.

The variation is quite large and is SIGNIFICANT (Important)
 Correlation & Regression 28
Correlation Regression
Data sets Not necessary to be related Some relationship may be there
Range of value (-1) to (+1). (-1) means – one data goes A best fit line is made. Usual is root
up and other comes down. (+1) means mean square value and optimization.
both move is same direction.
Example Positive – height and weight, Al Nino Best fit lines, can be straight or curved
and rain fall lines.
Names of Pearson method, Spearman Rank,
measurement

Both are used to predict future outcomes if a certain data point occurs.
Multiple Correlation & Regression Analysis 29

 Use of more than ONE independent variable to increase the

accuracy of the estimates.
 This is a higher level of analysis.
 Helpful where we can do complex maths.
 We may not get the relationship with one, so, two or more are used.
 Time Series 30

 When a measurement is carried out over a period of time

 Time interval to be same.
 Measurement data to be same in same units.
This gives the historic perspective of the actual situation.

Example:
Per Capita Income of a country .. Tells us how the country is growing in prosperity’
X – Charts on the machine… tells us how the machine is behaving in it’s accuracy.
Stock price – how is it changing over time.
Index 31

 Very important number

 When several data points are joined in a common number, with relevant
importance.
 It is a weighted average.

NIFTY Index.
Shares prices of 50 shares.
Each share has a fixed weight.
Total 50 shares are such that they represent the over all Indian Economics
Some Important Indexes 32
 Public Indexes
 Consumer Price Index
 Whole sale price Index
 Property price Index
 Construction Cost Index.
 Commodity Price index.
 Private Index – made by a company for it’s use.
 Labour productivity index.
 WIP Index
 RMC / Sale index
Index as Time Series 33

Index are very commonly used in time series as

they give a fair description of the situation and a
better decision making data set.

100 years Dow Jones

 34

thanks
SK Agrawal
9873592082
ska@[Link]