Hypothesis Testing

Q1.A F&B manager wants to determine whether there is any significant difference in the diameter of
the cutlet between two units. A randomly selected sample of cutlets was collected from both units
and measured? Analyze the data and draw inferences at 5% significance level. Please state the
assumptions and tests that you carried out to check validity of the assumptions.

Step1 Business Problem

Two check whether the diameter of two units are similar or not?

Step2 y and x

So here is y is continuous and x is discrete

Step3 Here we will use 2-sample t test

Step4 Find normality of this data

Normality Test

Normality test for unit A

null hypothesis(H0):Data are normal

alternate hypothesis(Ha):data are not normal

if p-value is > 0.05 => Accept null hypothesis

if p-value is < 0.05 =>Reject null hypothesis

Alternate hypothesis is used when we take action or p-value is less than 0.05

As p-value is 0.287 > 0.05 hence accept null hypothesis(H0)

Normality test for unit B

Here p-value is 0.687 > 0.05 => P high h0 fly

Hence accept null hypothesis(H0) which means data are normal

we can go for further test which is variance test

Variance Test

H0: variance of unitA = variance of unitB

Ha: variance of unitA NOT= variance of unitB

p-value is 1.00 > 0.05=>P high Ho fly => Accept Ho,Hence we prove variance of unitA = variance of
unitB.

2 Sample t test for compare mean

H0:Average of unit A = Average of unit B

P-value is 0.236 > 0.05=>P high Ho fly => Accept Ho, hence Average of unit A = Average of unit B

As per above results we can say that there is similarity between unitA and unitB i.e unitA = unitB

Q2.A hospital wants to determine whether there is any difference in the average Turn Around
Time (TAT) of reports of the laboratories on their preferred list. They collected a random sample
and recorded TAT for reports of 4 laboratories. TAT is defined as sample collected to report
dispatch.

Analyze the data and determine whether there is any difference in average TAT among the
different laboratories at 5% significance level.

Step1 Business Problem

Two check whether there is any difference in average TAT

Step2 y and x

So here is 4 labs are input TAT(Turn around time) is output

x is more than 2 discrete and y is continous

Step3 Here we will use ANOVA-One way

Find difference between 4 labrotaries with respect to time

X -> 4 labratory

y -> TAT(Turn around time)

Step4 Find normality of this data

Normality Test

Normality test for Laboratory 1

null hypothesis(h0):Data are normal

alternate hypothesis(ha):data are not normal

if p-value is > 0.05 => Accept null hypothesis

if p-value is < 0.05 =>Reject null hypothesis

alternate hypothesis is used when we take action or p-value is less than 0.05
As p-value is 0.532 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

Normality test for Laboratory 2

As p-value is 0.733 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

6)Normality test for Laboratory 3

As p-value is 0.577 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

7)Normality test for Laboratory 4

As p-value is 0.419 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

Variance Test

H0: All variance are equal

Ha: Atleast one variance is different

p-value is 0.070 > 0.05=>P high Ho fly => Accept Ho, hence we prove variance of all laboratory are
same

Anova Test-One way

H0:Average of all laboratory are same

Ha:Average of atleast 1 laboratory are different

P-value is 0.00 < 0.05= Accept Ha, hence Average of atleast 1 laboratory are different

As per results we can say that these are not equal i.e Average of atleast 1 laboratory are different
Q3.Sales of products in four different regions is tabulated for males and females. Find if male-
female buyer rations are similar across regions.

Step1 Business Problem

Two find buyer ratios are similar across region or not

Step2 y and x

x is more than 2 discrete and y is discrete

Step3 Here we will use Chi-square test

Step4 Find normality of this data

Normality Test

4)Normality test for East

null hypothesis(h0):Data are normal

alternate hypothesis(ha):data are not normal

if p-value is > 0.05 => Accept null hypothesis

if p-value is < 0.05 =>Reject null hypothesis

alternate hypothesis is used when we take action or p-value is less than 0.05

As p-value is 0.227 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

5)Normality test for West

As p-value is 0.227 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

6)Normality test for North

As p-value is 0.227 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

7)Normality test for South

As p-value is 0.227 > 0.05 P high Ho fly => Accept Ho,hence accept null hypothesis(H0) means data
are normal

Chi-Square Test

H0:All averages are same

Ha:atleast 1 are different

P-value is 0.674 > 0.05=>P high Ho fly => Accept Ho, hence Average are same

As per results we can say that there is proportion of male and female buying is similar
Q4.TeleCall uses 4 centers around the globe to process customer order forms. They audit a certain
% of the customer order forms. Any error in order form renders it defective and has to be
reworked before processing. The manager wants to check whether the defective % varies by
centre. Please analyze the data at 5% significance level and help the manager draw appropriate
inferences

Step1 Business Problem

To check whether the defective % varies by center or not

Step2 y and x

x is more than 2 discrete and y is discrete

Step3 Here we will use Chi-square test

Chi-Square Test

H0:All are same

Ha:atleast 1 are different

P-value is 0.227 > 0.05=>P high Ho fly => Accept Ho, hence Average are same

As per results we can say that all the canters are equal.

Q5.Fantaloons Sales managers commented that % of males versus females walking in to the store
differ based on day of the week. Analyze the data and determine whether there is evidence at 5 %
significance level to support this hypothesis.

Step1 Business Problem

To find proportion male vs female differ from weekdays or weekends are equal or not

Step2 y and x

x is discrete with 2 categories and y is discrete

Step3 Here we will use 2-Proportion test

2-Proprotion Test

H0:Proportion of male vs female in weekdays = Proportion of male vs female in weekends

Ha:Proportion of male vs female in weekdays NOT = Proportion of male vs female in weekends

Hence Proportion of male vs female in weekdays = Proportion of male vs female in weekends