PROFICIENCE, CENTRE FOR CONTINUING EDUCATION,

INDIAN INSTITUTE OF SCIENCE, BANGALORE

JAN-MAY 2007
Course: INTRODUCTION TO SIX SIGMA

Date: 12th May 2007. Time: 2-5pm.

1. Answer all questions. Different sections carry different marks.

2. Use calculators wherever required.

SECTION ONE. Objective questions (20 X 1 marks)

Note: This section may have multiple answers to a given question. Answer carefully. No
partial marks will be given.

Q1. For a normal distribution, six standard deviations on each side of the mean would
cover
a. Approximately 99.99998% of total population

b. Approximately 99.23739% of total population

c. Approximately 99.73124% of total population

d. None of the above

Q2. Input for QFD comes from

a. Voice of Customer
b. Surveys
c. Internal Data
d. Customer Satisfaction Form

Q3. Process Mapping helps in

a. Visualizing the activities
b. Understanding the overall picture
c. Identifying the bottle necks
d. Identifying the waste in the process

Q4. When referring to a curve that tails off to the left end, you would call it
a. Symmetrical
b. Skewed right
c. Skewed left
d. All of these
e. None of these

Q5. The Central Limit theorem

a. Requires some knowledge of the frequency distribution
b. Permits us to use sample statistics to make inferences about population
parameters
c. Relates the shape of a sampling distribution of the mean to the mean of the
sample

1
 d. Requires a sample to contain fewer than 30 observations

Q6. A major Automobile manufacturer has had to recall several models from its 1993 line
due to quality controlproblems that were not discovered with its random final inspection
procedures. This is an example of
a. Type I error
b. Type II error
c. Both Type I and Type II error
d. Neither type of error

Q7. If we say that Alpha = 0.10 for a particular hypothesis test, we are saying that
a.10% is our minimum standard for acceptable probability
b.10%is the risk we take of rejecting a hypothesis that is true
c.10% is the risk we take of accepting a hypothesis that is false
d. (a) and (b) only
e. (a) and (c) only

Q8. Which one of the following terms does not belong with the others?
a. Ishikawa Diagram
b. Pareto Diagram
c. Fishbone diagram
d. Cause and Effect Diagram

Q9. Which Kind of variation can be seen in a control chart

a. Inherent variation
b. Special Cause variation
c. Random variation
d. Patterns in variation

Q10. In the equation Y = A + BX +e, the e represents

a. The X intercept of the observed data
b. The value of Y to which others are compared to determine the best
fit
c. Random disturbances from the population regression line
d. None of these

Q 11. A statistical technique to assess, monitor and maintain the stability of a process.
a. Pareto Chart
b. Control Chart
c. Run Chart
d. Cause and Effect Diagram
e. Histogram

Q.12. If a bar chart is seen arranged in the descending order of frequencies, then you are
seeing a
a. Pareto Chart
b. Bar Chart
c. Histogram
d. Control Chart
e. Flow Chart
f. Run Chart

2
Q13. In Statistics, ‘Mean’, ‘Median’ and ‘Mode’ are measures of
a. Central Tendency
b. Variance
c. Deviation (Standard Deviation)
d. Dispersion

Q14. What are the characteristics of discrete data?

a. Location
b. Spread
c. Shape
d. Opportunity for Error

Q15. Out of the following, what can be used for transforming the non-normal data?
a. Y2
b. Y3
c. 1/Y
d. Box-Cox transformation

Q16. The following are the different types of benchmarking

a. Internal
b. Functional
c. Competitive
d. Hierarchical

Q17. Concluding a manager A is better than manager B when they are at the same level of
efficiency, the following error is made.
a. Type I
b. Type II
c. Type III
d. Type IV

Q18. The following are the prioritization tools for identifying the right contributors to an
out put.
a. Pareto
b. Regression
c. Correlation
d. FMEA

Q19. F-test is used to

a. test if two variances are equal
b. test if two means are equal
c. test if two ratios are equal
d. all of the above
e. none of the above.

Q20. If an opportunity for error is 2, the defect per opportunities is calculated as

a. Defects Per Unit * 2
b. Defects Per Unit / 2
c. (Defects Per Unit)2

3
 d. Defects Per Unit + 2

SECTION 2 TRUE/FALSE (10 X 1 mark)

1. Do not apply SPC tools to processes that are known to be out of control.
2. A specification is a customer-defined tolerance for the output unit characteristics
and it can be one sided or two sided.
3. Upper and lower control limits for control charts usually represent 4.5 standard
deviations from the mean value of the variable measured.
4. Assuming a linear relationship between X and Y, there is a strong negative
relationship between them if the coefficient of correlation (r) equals -0.10.
5. The Six Sigma Management approach is based on the concept that a process should
not create more than 3.4 defects per million products produced or services rendered
which is equivalent to six standard deviations from the mean of the monitored
variable.
6. Central Tendency, Spread and Shape are the characteristics of discrete data.
7. σ total = σ process + σ measurement where σ is measurement system
variation.
8. In order to measure process capability, the process to be measured must only have
common causes of variation.
9. Discrete data does not allow understanding the process variation and needs larger
samples for analysis.
10. If the population distribution is sufficiently skewed or the sample size is
sufficiently small (<30) then the population parameter cannot be tested.

SECTION 3: Fill in the Blanks ( 10 X 1 Mark)

1. Multiple regression analysis is for ______ response and _________ contributors.
2. When P value is more than alpha, null hypothesis is _____________.
3. A good P value, but poor R2 value in regression is due to ______________.
4. The six sigma project sponsorship is provided by _______________.
5. ___________ is a precise description that tells how to get a value for the characteristic you are
trying to measure.
6. Looking at the data set 40,20,10,20,50,90,80,30,20,40, the mean is _________, mode is
_______ and median is _________

4
7. Data that is derived from counting is ________ data and data that is derived from measuring is
________ data.
8. Flow chart is for process analysis as Scatter Diagram is for ________________.
9. The best tool for analyzing the relationship, over time, between quantitative data from two
variables suspected to influence one another is _____________.
10. All possible combinations of the levels of factors are studied in ____________ designs.

SECTION 4: Match the following (10X1 Mark)

COLUMN 1 COLUMN 2
1. A measurement system that is consistent in its A. Affinity Diagram
variability
2. A measurement system that is consistent when B.Quality Function Deployment
different appraisers produce consistent results
3. Organizing ideas into meaningful categories by C. Poka-Yoke
recognizing their underlying similarity
4. Plots of data arranged in a time sequence D. Zone rules
5. Running a trial combination twice without E. CTQ Drill Down
changing the setting in an experiment
6. A technique to make errors difficult to happen F. Repetition
7. Out of control process G. Affinity Diagram
8. Single Customer CTQ H. Reproducibility
9. Green Belt I. Six Sigma Project Manager
10. Two sample t-test J. Replication
K. Six Sigma Project Sponsor
L. Repeatability
M. Elevator Speech
N. Benchmarking of two sets of data

SECTION 5: Numerical Questions (4X5 Marks)

1. An accounts firm knows that homogeneity in work experience affects effectiveness of account
executives. It collects the following data on years of experience for its executives in office 1 & 2.
Which office is more effective?

Office 1 Office 2
4 7
3 2
6 3
5 4
5 7
2 4
5 2
4 4
6

5
 4

Ans: The mean for Office 1 is 4.40 and Std deviation is 1.26.
Office 2 is 4.125 and Std deviation is 1.95. The range also varies from 2.49 to 5.76.
Hence Office 1 is more effective.
2. A candy bar manufacturer is interested in estimating how sales are influenced by the price of
their product. They randomly choose 6 small cities of similar characteristics and offer a candy bar
at different prices in each of them, then collect the following data as described in the table. If
candy bar sales are the dependent variable (Y) and the company conducts a simple linear
regression answer the following questions.

City Price Number Sold

Chennai 1.30 100
Bangalore 1.60 90
Hyderabad 1.80 90
Calcutta 2.00 40
Delhi 2.50 38
Mumbai 2.90 32

a) What is the estimated slope parameter for the candy bar price and sales data? 47.3
b) What is the percentage of the total variation in candy bar sales explained by the
regression model? 79.5%
c) If the price of the candy bar is set at $2, what is the value for predicted sales? 65/66

Answer:

The regression equation is

Number Sold = 160 - 47.3 Price

R-Sq value is 79.5% and P value is 0.003.

Predictor Coef SE Coef T P

Constant 160.39 25.04 6.41 0.003
Price -47.30 11.99 -3.94 0.017

S = 15.8596 R-Sq = 79.5% R-Sq(adj) = 74.4%

3. A quality control analyst for a light bulb manufacturer is concerned that the time it takes to
produce a batch of light bulbs is too erratic. Accordingly, the analyst randomly surveys 10
production periods each day for 14 days and records the sample mean and range for each day:

Day Mean Range

1 58.5 5.1
2 47.6 7.8
3 64.3 6.1
4 60.6 5.7

6
 5 63.7 6.2
6 57.5 6.0
7 55.0 5.4
8 54.9 6.1
9 55.0 5.9
10 62.7 5.0
11 61.9 7.1
12 60.0 6.5
13 58.3 5.9
14 52.0 5.2

a) What is the center line and Lower Control Limit if the analyst constructs an R chart of this
data to see if the variability in production times is in-control?

LCL – 3.565

CL – 6

b) Suppose the analyst constructs a control chart for the mean to see if the production process is
in-control, what will be your final conclusions for process stability and the deviations?

LCL – 45.38
CL – 58
UCL -70.62

Though the process is within control limits, the trend shows that the process is not very stable and
consistent.

4. Kailash is the Pizza Hut dispatcher for Jayanagar area. He wants pizzas delivered to the
customers within 30 minutes of leaving the kitchens. Pizzas with longer delivery times tend to be
too cold when they arrive. Each of his 10 volunteer drivers is responsible for delivering 15 Pizzas
daily. Over the past 1 month , Kailash has recorded the percentage of each day's 150 Pizzas that
were delivered on time and the data is shown below.

a) Which control chart Kailash will use to see whether the delivery process is in control?
Simple X bar chart
b) Mention the no of days when the process is out of control and which is worrying to Kailash
None.
c) What action do you recommend for Kailash?

% on % on
Day time Day time
1 89.33 16 89.33
2 81.33 17 89.33
3 95.33 18 78.67
4 88.67 19 94
5 96 20 94
6 86.67 21 99.33
7 98 22 95.33
8 84 23 94.67

7
 9 90.67 24 92.67
10 80.67 25 81.33
11 88 26 89.33
12 86.67 27 99.33
13 96.67 28 90.67
14 85.33 29 92
15 78.67 30 88

SECTION 6: Case Analysis ( 1 X 20 Marks)

Refer to the attached Case Study. This talks about an account growth six sigma project executed
in a particular organization. Criticize the approach they have taken in executing this DMAIC
project based on your knowledge learnt from this course.

DON’T write stories. Provide the bullet points for positives and negatives individually. The
case analysis should not go beyond 2 pages of the answer paper. Long answers will be penalized.
Mentioning only the tools will NOT yield any marks. Any tool usage should be justified for the
right use in the right phase.

--------------------------------- END OF QUESTION PAPER-----------------------------------------