anuj.gupta010@gmail.

com
T56GZSRVAH
Hypothesis Testing

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
Proprietary content. © Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited.
 LET’S SET SOME GROUND RULES

• Come prepared for these sessions by watching the videos.

-Concepts will be covered in the videos.
-Hands-On Application will be covered in Mentor Sessions.
[email protected]
T56GZSRVAH

• Submit all assignments on time.

• Let’s be punctual & respect each
others’ time.

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Content
• Population vs Sample
• Why Sample & not entire population?
• Population Parameter vs Sample statistic
• Examples
• Terms related to sampling
[email protected]
T56GZSRVAH • Multiple samples of a particular size from same population
• Sample Mean
• Sampling Distribution of the mean
• Standard Error
• Sampling Error
• Central Limit Theorem & how it helps
• Hypothesis Testing
• What is it & Why we make them
• What is null hypothesis & alternate hypothesis
• Points to ensure then formulating null & alternate hypothesis
• What is Type 1 error, Type 2 error, alpha & beta
• Alpha never calculated. It is a given or assumed value.
• This
How to identify, if the test to be file is meant
conducted is a for personal
1 tailed or 2use by test
tailed [email protected]
and if 1 tailed then rightonly.
tailed or left tailed
• Sharing or publishing
How to identify which test to be used to test hypothesis the contents in part or full is liable for legal action.
 Sampling Distributions
● Need for sampling
○ Given the limited resources and time, it
is not always possible to study the
population. That’s why we choose a
sample out of the population to make
inference about the population
[email protected]
T56GZSRVAH

● Sampling Distributions
○ It is a distribution of a particular sample
statistic obtained from all possible
samples drawn from a specific
population

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
Proprietary content. © Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited. 4
 Central Limit Theorem

The sampling distribution of the sample means will approach

normal distribution as the sample size gets bigger, no matter
what the shape of the population distribution is.

[email protected]
T56GZSRVAH
Assumptions

Data must be randomly sampled Sample values must be independent of each other

Samples should come from the same distribution Sample size must be sufficiently large (≥30)

Let’s see CLT in action by simulation - Link to external site

This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action.
Proprietary content. © Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited. 5
 Hypothesis Testing
• Why we hypothesize?
• What is null hypothesis & alternate hypothesis
[email protected]
T56GZSRVAH• Rules for correctly formulating numlli &
T
alrteernate hypothesis
• Only null hypothesis can have equality sign
• Alternate cannot have equality sign
• Null & alternate should be mutually exclusive
• Hypothesis testing process flow

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Hypothesis Testing – Process flow

[email protected]
T56GZSRVAH

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Hypothesis Testing
• α, β, Type I error, Type 2 error
• Significance level, Confidence level & power of test
• α + β is not equal to 1 !
[email protected]
T56GZSRVAH

• α is never calculated but is given or assumed !

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Hypothesis Testing Ho : μ = 60
Ha : μ != 60

• 1-tailed vs 2-tailed test

[email protected]
• In 1-tailed, left tailed or
T56GZSRVAH Ho : μ = 60
right tailed? Ha : μ < 60

• The answer depends on

Ha

Ho : μ = 60
Ha : μ > 60

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Hypothesis Testing – Which Test to use
Start

1
How many 2
[email protected] samples?
T56GZSRVAH
No
Population σ Paired
known? samples

Yes No

Yes

Independent 2 sample t-
One sample Z-test One sample t-test Paired t-test test

This file is meant for personal use by [email protected] only.

Sharing or publishing
the contents in part or full is liable for legal action.
Source: http://www.rebeccabarter.com/blog/2018-12-04_hypothesis_testing/
 Let’s discuss the case study!
[email protected]
T56GZSRVAH

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
Proprietary content. © Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited.
Can Uber double its revenue by changing a single
word ?
Uber keep running experiments to boost revenue
from its super users in India. Changing the word

from “buy” to “try” more than doubled the

[email protected]
conversions for Uber.
T56GZSRVAH

Uber wanted to discover how much extra its core users

are willing to pay and hence increase its revenue.

https://factordaily.com/opinions/uber-ab-test-boost-
revenue-india/

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
 Case Study-2 (Titan Insurance)
The Titan Insurance Company has just installed a new incentive payment scheme for its life policy sales-force. It wants to have
an early view of the success or failure of the new scheme. Indications are that the sales force is selling more policies but sales
always vary in an unpredictable pattern from month to month and it is not clear that the scheme has made a significant
difference. Life Insurance companies typically measure the monthly output of a salesperson as the total sum assured for the
policies sold by that person during the month.
[email protected]
T56GZSRVAH
Titan's new scheme is that the sales force receive low regular salaries but are paid large bonuses related to their output (i.e. to
the total sum assured of policies sold by them). The scheme is expensive for the company but they are looking for sales
increases to compensate for it. The scheme has now been in operation for four months. It has settled down after fluctuations in
the first two months due to the changeover.

To test the effectiveness of the scheme, Titan has taken a random sample of 30 salespeople measured their output in the
penultimate month prior to changeover and then measured it in the fourth month after the changeover (they have deliberately
chosen months not too close to the changeover).

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
[email protected]
ANY QUESTIONS
T56GZSRVAH

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
[email protected]
T56GZSRVAH

This file is meant for personal use by [email protected] only.

Sharing or publishing the contents in part or full is liable for legal action.
[email protected]
T56GZSRVAH

HAPPY LEARNING
This file is meant for personal use by [email protected] only.
Sharing or publishing the contents in part or full is liable for legal action.