Scribd
Upload
35 views6 pages

AMS Email Interface Comparison Analysis

This document summarizes the results of an experiment conducted by AMS to determine the most efficient messaging service to reduce the time to fully update subscribers' messages. AMS tested two services on 15 random customers each and found that the second service was faster. Statistical analysis including F-tests and t-tests showed there was sufficient evidence to conclude that the second service's mean update time was less than the first, so AMS should choose the second service.

Uploaded by

clara lou
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
35 views6 pages

AMS Email Interface Comparison Analysis

This document summarizes the results of an experiment conducted by AMS to determine the most efficient messaging service to reduce the time to fully update subscribers' messages. AMS tested two services on 15 random customers each and found that the second service was faster. Statistical analysis including F-tests and t-tests showed there was sufficient evidence to conclude that the second service's mean update time was less than the first, so AMS should choose the second service.

Uploaded by

clara lou
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Lab assignment 3

Introduction to business Statistics and Data analysis

RAMI Ismael
ANTOMARCHI Clara
De LESTAPIS Dimitri
ZOUHAIRI Sophian
UZAN Mordehai

Case problem

First question :
In a search for recruiting the best messaging service, AMS decided
to conduct an experiment to see which of the two will be the most
ef cient to reduce the amount of time to fully update each subscriber’s
set of messages.
AMS decided to assign 15 random customers test on each of the 2
services in order to compare the result of the two services to make the
best choice.

fi

Here are the results:

To compare those two sets of data, a F-Test is needed rst to test


the difference of variability between the two samples (are the variance of
the 2 sets the same or not), hypothesis are:

0: 1 = 2
1: 1 ≠ 2

Since we have: 1 ≈ 0,9 and 2 ≈ 0, and   we cannot cannot reject


H0: there are insuf cient evidence to assume that =

𝐻
𝐻
𝜎
𝜎
𝜎
𝜎

𝜎
fi

𝝈
fi
𝟏
𝝈
𝟐

After having made this rst observation we must do a Pooled-


variance t-test.
Since the objective of AMS is to deal with the most competent and
fastest operator, in order to determine if the if the 2-population means
are equal, we must use the two-sample t-test (pooled t-test), hypothesis
are:

0: 1 = 2
1: 1 ≠ 2

*With: = 0,05

Since = 0,05 , with an interval of 95% we can state that we have


suf cient evidence to reject hypothesis H0, the null hypothesis could be
denied at a 95% level of con dence, also the T-Test appears to be
negative so that means that the E-Mail Interface 1 is slower than the E-
Mail Interface 2.

In order to choose the best e-mail interface, AMS did statistical


research comparing the 2 option and their set of data, through the result
we had in our appendix, we can state that AMS should chose the 2nd e-
mail interface to improve user experience and reduce the amount of time
to fully update each subscriber’s set of messages.

4
𝐻
𝐻

fi

𝜇
𝜇
𝛼
𝛼
𝜇
𝜇

fi
fi
Second question :

Giving the new rule that there are only 15 users instead of 30, and
according to the fact that the results won’t be independent, we must do a
“paired t-test”, hypothesis are:

0: =0

1: ≠0

With = 2− 1

We need to see if the mean difference of measurement between


the different e-mail interface are equal to 0 or not.
Since the p-value is below the signi cance level with 0,00292 < 0,05

&

Since the T-Stat is below the critical value with -3,59628 < 2,14478

As a result, there are suf cient evidence to reject the null


hypothesis.

In view of the results obtained by the new constraint and in view of


the conclusion of the old provision, we can certify that it is necessarily
more advantageous for the company AMS to select the e-mail interface
number 2. This one is faster than the

5
𝐻
𝐻
𝜇
𝜇
𝜇
𝐷
𝐷
𝐷

fi

fi

You might also like