1

BUSI 650 - Assignment 1



MOHD ADNAN JAHANGEER ANSARI (2344465)

University Canada West
BUSI 650: Business Analytics (HBD-FALL24-34) - BUSI 650 (HBD-FALL24-34)
Behnaz Bahmei
23-OCTOBER-2024
 2
Answer -1-

Hypothesis testing –
Η𝜊 = 500 𝑀𝑃𝑠
As we know, the company claims that the average tensile strength of metal compound produced is 500 MPs So,
 The null hypothesis for lines A & B will be equal to companies claim.
For line A = Η𝜊 ∶ 𝜇Λ = 500
For line B = Η𝜊 ∶ 𝜇𝑁 = 500
 For mean alternative hypothesis for lines A & B will differ from 500 MPa.
For line A = Η𝜊: 𝜇𝐴 ≠ 500
For line B = Η𝜊: 𝜇𝐵 ≠ 500
 This indicates a two-tailed test.

Answer -2-

Test statistics for z-value.

 For line A -
Sample mean = Χ𝐴 = 500
Standard deviation = 𝑆Α = 25
Mean = 𝜇𝜊 = 500 𝑀𝑃𝑎
Sample size = 𝜂𝐴 = 55

By using this formula for z-testing we got,

Z= 495-500 / 25/√55
 3
ZA = -1.483

 For line B –
 Sample mean = Χ𝐵 = 510
 Standard deviation = 𝑆B = 30
 Mean = 𝜇𝜊 = 500 𝑀𝑃𝑎
 Sample size = 𝜂𝐵 = 65

Using z-testing formula -

ZB = 510-500 /30/√65
ZB = 2.687

Answer -3-

The P-value for lines A & B using the Z-distribution table for a two-tailed test.
 For line A - since it’s a two-tailed test we multiply the p-value by 2 for increasing our probability or double
our probability.
 For z-value: -1.483
p-value for line A = 2*0.06944 = 0.13888
 Since 0.13888>0.05 we failed to reject the null hypothesis for line A.
 There is not enough evidence to suggest that the tensile strength on line A deviates from 500MPa.
 For line B – Since it has a two-tailed test we multiply the p-value by 2 for increasing our probability or
double our probability.
 For z-value = 2.687
p-value for line B = 2*0.99632 = 1.99264
 Since 1.99264<0.05 we reject the null hypothesis for line B

Answer -4-
For a two-tailed test with 𝛼 = 0.05 the total area under the normal curve is 0.05 that we will be testing.
 Splitting the tail for lines A & B.
0.05/2 = 0.025
 Using standard normal distribution table critical z-value for each tail.
 4
 For the lower tail critical z-value for 0.025 is -1.96.
 For the upper tail critical z-value for 0.025 is +1.96.
Z = ±1.96
Comparing z-value from tas2 2.
 For line A = the calculated z-value was -1.48 (-1.48>-1.96) It does not exceed the critical value of -1.96.
So, we fail to reject the null hypothesis which means there is no difference between the tensile strength of
line A and claimed 500 MPa.
 For line B – The calculated z-value was 2.69 (2.69>1.96) It exceeds the critical value of +1.96.
 So, we reject the null hypothesis which says the tensile strength of line B significantly differs from claimed
500 MPa.

Answer -5-

In hypothesis testing, sample size plays an important role in determining the accuracy and reliability of the
conclusion. So, the difference in sample size for line A and line B, which is (55 and 65 respectively) affects the
results by
 Line B has a larger sample size, 65, than the sample size of Line A, which is 55. A larger sample size is
going to be more representative of the true population mean and will provide a lower standard error.
𝜎
 So, standard Error =
√𝑛

25
For line A = = 3.370
√55

30
For line B = = 3.721
√65

The larger sample size will provide a more reliable conclusion.

 Line A (n = 55): With a smaller sample size, the hypothesis test has less precision, which is reflected in
lower statistical power. Perhaps this makes the detection of significant differences hard, and that’s why the
test did not reject the null hypothesis.
 Line B (n = 65): The estimates are more precise due to the larger sample size, and the statistical powers are
higher. That’s why we rejected the null hypothesis for Line B.
 5
 We can be more confident in that result since the sample size for Line B is larger. If the sample size for Line
A is smaller, then we will be less sure of the result for that production line. This reduced sample size
impacts the strength of trust for each line’s conclusions.

Answer -6-

If the company wants to assess the overall quality of components coming from both Line A and Line B together,
then, the objective is to check whether the combined average tensile strength from both production lines deviates
from the claimed 500 MPa.
Since we want to verify the overall mean tensile strength across the two lines, we will see if the combined average
is different from the claimed 500 MPa. Here are the hypotheses:

Null Hypothesis (H₀): The mean tensile strength of components for both lines is equal to the claimed 500 MPa.

H𝜊 : μ combined =500 MPa

Alternative Hypothesis (H₁): The mean tensile strength of components for both lines is not equal to 500 MPa.

HΑ : μ combined ≠ 500 MPa

It is a two-sided test since one is interested in whether the combined tensile strength differs noticeably from 500
MPa in either direction.

If the computed test statistic-the z-value-exceeds the critical value, or if the p-value is less than your chosen
significance level-usually 0.05-you reject H0 and can say the mean tensile strength is significantly different than
500 MPa for the combination of metals.
You will fail to reject the null hypothesis if the test statistic is less than the critical value and the p-value is greater
than the level of significance. You hence conclude that there is no significant difference from the claimed value.

Answer -7-

In hypothesis testing about the tensile strength of components coming off a production line, the consequences of
both Type I and Type II errors should be recognized: first, concerning compliance or non-compliance to quality
control standards, then also to customer satisfaction. Let's expound on what is involved with the Type I and Type II
 6
errors and their possible implication in this scenario:

Type I Error - Type I error rejection of a null hypothesis that is true. when the null hypothesis has assumed that
the average tensile strength for a lot of parts is 500 MPa, the Type I error will be the determination that the mean
tensile strength is significantly different from 500 MPa when, in fact, it is not.
Consequences - The company may needlessly alter or change the processes of manufacture upon mistaken belief
that the production lines are not producing the required tensile strength. This can lead to money being wasted, as
well as disruptions in production.
New equipment or process changes that are not warranted use resources, which can be spent better elsewhere for
improvement or maintenance.
For Example - Suppose that the quality control team mistakenly reaches the conclusion that the new
manufacturing process on Line B does not increase the tensile strength beyond the standard, and a costly redesign
or adjustment in the process is done based on this conclusion, which is not only unnecessary but also affects
operational efficiency.

Type II Error - A Type II error occurs when a false null hypothesis is failed to be rejected.
In this context, such a mistake would mean failing to identify a real deviation from the standard of 500 MPa when,
actually, this is true, the tensile strength significantly differs from the value.
Consequences - If indeed there is an issue with the tensile strength failing to meet the set standard of 500 MPa and
remains undetected, the company might continue manufacturing and selling such defects or non-compliant parts.
The use of components with specified tensile strength may lead to product failure. This may be viewed as a
potential loss of customer trust and satisfaction, possibly even exposure to safety risks.
For Example - if there are material quality problems and thus the tensile strength is always less than 500 MPa for
components produced from Line A, and such deviation has not been identified and rectified, weaker components
may be produced which could fail in use and give rise to product recalls or safety incidents.

Understanding and managing these errors effectively is essential for maintaining high standards of quality and
ensuring customer satisfaction while optimizing resource use and operational efficiency. Understanding and
managing these errors effectively is essential for maintaining high standards of quality and ensuring customer
satisfaction while optimizing resource use and operational efficiency.
 7
Answer -8-
In hypothesis testing, the choice between a one-tail or a two-tail test will depend on the research question or
business concern. Let's consider both alternatives in light of testing the tensile strength of components from
Production Line A versus Line B.
One-Tailed Test: It applies when one is concerned with the higher or lower deviations in one direction than the
value being claimed.
For Example - The one-tailed test is appropriate if the company is interested in testing only whether the tensile
strength is either too low or too high, but not both. More specifically, if the concern is from which direction tensile
strength is lower than 500 MPa meaning weaker components, then a one-tailed test about the left tail is appropriate.
When the firm is only interested in verifying if the tensile strength is greater than 500 MPa - meaning that their
components are stronger than necessary-you would use a one-tailed test on the right tail.

Two-Tailed Test: It is a test in which you are concerned with changes in both directions-the tensile strength is
either higher or lower than the value it is claimed to be (500 MPa).
A two-tailed test would be appropriate in a situation where the company is interested in the detection of any
significant deviation from the standard — whether too high or too low.
This applies to this scenario because the company wants to have the assurance of consistency in the components
meeting the exact
Which is More Appropriate in This Context?
In this case, a two-tailed test is more appropriate, and this is where the firm likely is concerned about the
underperformance component being weaker than 500 MPa and at the same time overperformance component being
stronger than 500 MPa, which may increase costs or introduce process inefficiency. A two-tailed test would pick up
either the negative or positive deviation and hence guarantee that any significant difference from the standard is
identified.
Since the company does not expect the tensile strength in only one direction to increase or decrease two-tailed test
is more appropriate because this examines both directions of deviation.
Deviations in either direction could affect customer satisfaction or safety; thus, both options of tensile strength
being too high or too low are equally significant to monitor. specification of 500 MPa and doesn't want to overlook
possible problems on either side.

In conclusion, the two-tailed test is more appropriate in this scenario because it provides a complete picture of
potential deviations from the desired tensile strength, ensuring that both quality control and customer satisfaction
standards are maintained.
 8

Answer -9-
Scenarios Where Quality Testing in Different Production Lines or Times of Day May Be Important Scenario.

1: Component Dimensions-Production Line A vs. Production Line B

Till now, in a manufacturing plant, it has been common to manufacture a given component on two different
production lines. But the dimensions-like length, width, or thickness of the manufactured components differ from
one line to the other due to machinery or expertise developed in operating those machines. However, the company
wants to ensure that the dimensions of components meet strict tolerance levels to avoid problems in assembly.

Why tests are to be conducted separately:

Difference between Production Lines: Differences in machinery calibration, maintenance schedules, or operator
competence would lead to minor, but crucial, variations in component dimension.
Hypothesis Testing Goal: Testing both lines separately will ensure that any deviation from the standard is
identified and corrected specifically on the production line where the issue occurs. For example, one line could be
producing parts that are a little larger than the specifications call for, whereas the other produces components
within the limits of tolerance.
Impact: Making sure both lines meet the exact specifications saves the company from costly assembly errors,
rework of the product, or even dissatisfaction with the customers.

Scenario 2: Day Shift versus Night Shift in Beverage Bottling Plant

In a bottling plant, the volume of liquid in each bottle is of critical importance for the uniformity of the product and
customer experience. A company operates such a facility around the clock, splitting the time between day and night
shifts. The management is concerned that the variation in performance either by the personnel or the wear and tear
of machines might lead to variation in fill level at different times of the day.

Why tests should be conducted separately:

Time of day Variation: The night shift may be different because of reduced staffing levels, worker fatigue, or
machine wear from having operated all day. Also, more stringent controls and maintenance might be in place
during the day shift.
Hypothesis Testing Goal: This will help the company to ascertain if one shift bottles at a different fill level from
 9
another. In such a case, if one shift is persistently under-filling bottles, the company can make sure that products
are of consistent quality.
Impact: Testing for each shift separately will let the company ensure that all its products meet standards of
volume; thus, helping the company maintain brand reputation and consumer satisfaction.

Conclusion: In both cases, running several hypothesis tests for different lines or shifts aids the firm in finding
exact points of failure; these can then be more precisely corrected than if there were one overall correction to meet
quality standards for all times and conditions of production. This is necessary to be able to maintain product
standards while minimizing waste and maximizing efficiency.

B. DATA ANALYSIS –
1) Trend analysis

Calculating the percentage change for each store

a) Formula Used
June Sales – January Sales / January Sales * 100
b) Percentage change for Store A =25%
c) Percentage change for Store B =20%
d) Percentage change for Store C =18.18%

Store A showed the highest sales, which can be due to a rise in consumer visits product was readily available to
customers when needed or improvement in inventory or operations.
2) Sampling Techniques

Therefore, SRS would be an effective way to analyze customer feedback data involving 5,000 reviews. In SRS,
every review gets an equal chance to get selected in a sample of 200 reviews which is representative and unbiased
from the overall feedback. The methodology herein is straightforward, with no previous categorization into groups
or categories of data required; hence, quite useful in a dataset where reviews are not divided into distinct groups or
categories. However, the limitation with SRS could be that it may not capture diverse or segmented feedback
efficiently if the opinion distribution varies wildly across review types.
 10
a) Quantitative Solution
 Population: 5,000 reviews
 Sample Size: 200 reviews
 Sampling Fraction: 200/5000=0.04 or 4%
Limitations:
 SRS might not capture diverse viewpoints effectively if the distribution of opinions is not uniform.
 Stratified sampling requires a known categorization of the reviews, which may not always be available.

3) Inventory Turnover Analysis

a) Store A= 4.5+4.8+5.0+4.6+5.1+5.3/6=4.883
b) Store B= 5.0+5.2+5.4+5.1+5.5+5.7/6=5.316
c) Store C= 3.8+3.9+4.0+3.9+4.2+4.3/6=4.016