[Six Sigma]
INSTRUCTOR: [Sean Kiani]
CLASS: [QEM 3104]
• Read all questions carefully.
• Time allowed: 2.5 hours
DATE: 31/07/2025
STUDENT NAME: Amal Gangadharan, Arunya Buwaneswaran, Parvathy
Sudhakaran, Ruchi Shrivastava
STUDENT NUMBER: C0938097, C0941534, C0943746, C0933537
1. a. Hypothesis Test:
Null Hypothesis (H₀): σ² = 0.1, The population variance of the pin lengths is equal to 0.1
Alternative Hypothesis (H₁): σ² ≠ 0.1, The population variance of the pin lengths is not
equal to 0.1
Test statistic value:
Where,
Sample size, n = 50
Variance, σ02 = 0.1
Sample variance, s2=
Here, sample mean= 9.591
= each measurement value. Substituting the values in the equation,
We get s2= 0.00688
(50−1)∗0.00688
Therefore, Test statistic value, = = 3.371
0⋅1
The p-value = 4.62 x 10-20 , since the p-value is less than 0.05 we reject the null hypothesis
which indicates that there is strong evidence that the true variance is significantly different
and much smaller which is good for quality control.
The confidence interval for 95% is (0.0693, 0.1034)
b. Interpretation of results along with statistical proof using Minitab:
Figure 1.1 Test and Confidence interval
2.
Figure 2.1 Pareto Chart of Flaws
Figure 2.2 Pareto Chart of shift
A Pareto analysis was conducted to identify the most significant types of flaws leading to the
rejection of window guide rails in an automotive manufacturing process. The results revealed
that paint flaking was the most frequent defect, accounting for 37.5% of all cases. Scratch
flaws were the second most common at 32.5%. Together, these two types of defects represent
70% of the total issues identified, indicating that initial quality improvement efforts should
focus on addressing these two problem areas in line with the Pareto Principle (80/20 Rule).
Further analysis categorized the occurrence of these flaws by production shift. The Night
Shift emerged as the most problematic, responsible for 47.5% of the total defects. Notably,
this shift recorded the highest number of both paint flaking and scratch flaws. In contrast, the
Day, Afternoon, and Weekend shifts showed significantly lower defect totals, with each
contributing between 7 and 8 issues. This suggests a potential link between shift conditions
and quality outcomes, particularly during night operations.
Based on these insights, several key findings were drawn. Paint flaking is the leading cause
of rejection and should be prioritized for corrective action. Scratch defects are also prevalent
and require attention. Moreover, the Night Shift's elevated defect rate makes it a critical area
for targeted investigation and improvement.
3.
Figure 3.1 Xbar- R chart of diameter
In Xbar chart all subgroup means lie within control limits (LCL = 0.51785, UCL= 0.57533)
indicating process is stable and no unusual patterns or trends observed. In R chart all ranges
fall within control limits (LCL= 0 and UCL=0.0900) indicating process variability is under
control and consistent variation within subgroups.
Figure 3.2 Process Capability for Diameter
4. Conduct a GR&R study for the attached data and interpret the results accordingly
The purpose of this Gage Repeatability and Reproducibility (GR&R) study was to evaluate
the adequacy of the measurement system used for assessing part profiles. The study focused
on assessing both repeatability (variation due to the measurement device) and reproducibility.
Figure 4. 1 variance components and gauge evaluation
Figure 4. 2 Guage R&R Report for profile
The GR&R study confirms that the measurement system is acceptable, with minimal
variation due to the gage and operators. Most of the variation observed is due to actual
differences between parts, indicating that the system is capable of distinguishing product
variation. While a slight inconsistency was noted with one operator, overall repeatability and
reproducibility are within acceptable limits. The system can be reliably used for quality
control, with minor improvements recommended through operator training.
5. Complete a process map for the attached procedure.
Figure 5.1 process map
6. a
Check for Statistical Control (Control Chart)
Figure 6.1 I-MR Chart of Fat Original
Interpretation:
• The I chart shows that all points are within control limits, and there is no visible
trend or pattern.
• Therefore, the process is in statistical control.
• This means the fat content process was stable at the time of data collection.
Test for Normality
Figure 6.2 Probability plot of Fat Original
Interpretation:
• Since p > 0.05, we fail to reject the null hypothesis.
• This indicates the data follows a normal distribution, which validates the use of t-
tests and confidence intervals.
Confidence Interval for Mean Fat Content
Figure 6.3 One-Sample T: Fat Original
Interpretation:
• The CI does not include 15 grams, and the p-value < 0.05, so we reject the null
hypothesis.
• This means the mean fat content is significantly greater than 15 grams.
• The customers’ complaints about higher fat content appear valid.
Test if Standard Deviation < 1 gram
Figure 6.4 Test and CI for One Variance: Fat Original
Interpretation:
• Since p > 0.05, we fail to reject the null hypothesis.
• There is no evidence to support the claim that the standard deviation is less than 1
gram.
• In fact, the standard deviation is much higher than 1, which indicates high variation in
fat content.
b. Test if Proportion of Dissatisfied Customers is Acceptable
Figure 6.5 Test and CI for One Proportion
Interpretation:
• Since p < 0.05, we reject the null hypothesis.
• The dissatisfaction rate is significantly higher than 5%.
• This confirms that a notable portion of customers are unhappy, and quality
improvement is needed.
c. Compare Before and After (Two-Sample t-test)
Figure 4.6 Two-Sample T-Test and CI: Fat Original, Fat new
Interpretation:
• Since p < 0.05, we reject the null hypothesis.
• This means there is a significant reduction in mean fat content after process
improvement.
• The automated oil dispensing process appears to be effective.
Compare Standard Deviations (F-Test or Graphically)
Figure 6.7 Test and CI for Two Variances: Fat Original, Fat new
Interpretation:
• Since p < 0.05, we conclude that the variance has significantly reduced after the
improvement.
• This shows that the process is now more consistent and controlled.
Overall Interpretation
The I-MR chart analysis of the original procedure of fat content indicates that the process was
statistically in control since all the data points were within the I-MR control limits.
Additionally, normality test confirmed that the data were normally distributed and thus
appropriate for additional statistical analysis. However, the mean fat content was significantly
higher than the 15 grams stated because the 95% confidence interval was not 15 and the p-
value was less than 0.05. Also, the standard deviation was significantly greater than 1 gram,
indicating high variability of the process. Regarding customer complaints, the rate of
dissatisfaction was 8.3%, significantly higher than an optimal threshold of 5%, and reflecting
an actual concern on the part of customers. After the implementation of the automated oil
dispensing system, statistically significant reductions were observed both in the mean fat
content and in its variability, affirming that the process improvement was successful in terms
of offering more consistent and healthier hamburgers.
7. Town XYZ collects the following recyclable trash from its residents: glass, plastic,
paper, aluminum, yard waste, and iron. One of its primary aims for next year is
to encourage residents to reuse the recyclables as much as possible before
disposing of them.
Of late, the town has been receiving excessive recyclable trash from its
residents, and hence, the town hall aims to communicate to the resident’s
different ways to reuse the recyclables before the residents dispose of them. To
this end, the town hall collects data regarding how much waste of each category
was collected and recycled during the previous year.
The town hall first wishes to identify the categories of waste that contribute the
most to the total waste.
• Conduct the study to identify categories of waste. The employees of the town hall
brainstorm to identify the different ways to minimize the paper waste. They classify
paper waste into three types: office, kitchen, and other
To identify the major categories of waste, the percentage of each category contributing
towards the total waste is calculated in the below table.
Category Recycled (in hundreds of tons) Percent of total
Paper 600 39.5 %
Aluminium 260 17.1 %
Iron 210 13.8 %
Glass 200 13.2 %
Yard waste 200 13.2 %
Plastic 50 3.3 %
From the table, it can be said that major contributor to the waste is paper that makes up to
nearly 40% of the total waste.
• Conduct the study (Fishbone Diagram) and explain your finding
s.
The three major categories of paper waste identified in the brainstorming session by the
employers are : Office, Kitchen and Other (Newspapers, Telephone books and Junk mail)
From the provided image of categories of paper waste, a fishbone diagram can be constructed
to outline the main sources of excessive paper waste in Town XYZ and based on that, reuse
strategies before disposal will also be shown in the same diagram.
Figure 7.1 Fishbone diagram of Paper waste categories and reuse strategies
Summary of Fishbone diagram: From the above studies, it can be said that paper is the largest
contributor to the waste and the key reuse strategy from the fishbone diagram is to reduce
paper waste by promoting internal reuse (e.g., using office paper for notes), community reuse
(e.g., donating newspapers for crafts), and creative applications (e.g., using junk mail for
packing or art). These approaches target the main sources of paper waste: office, kitchen, and
other.
