Popcorn Experiment

Monitoring Popcorn Yield Using Control Charts
SIE 406: Quality Engineering
Spring 2016
Background
This simple (but fun) exercise makes practical use of SPC tools and concepts in one of our
favorite past times – preparing popcorn using a microwave oven. The primary objective is to
generate your own data by cooking popcorn and tracking how many kernels popped (or did
not pop). This project should demonstrate your broadening knowledge of SPC methods. This
project emphasizes three things:
1. Creativity in data generation. Exert some creativity in translating the experiment

into a real production scenario.
2. Preciseness in data collection. Based on the production scenario, generate data

correctly (e.g., rational subgrouping).
3. Correct use of SPC tools. Based on how data was collected, correctly use the most
appropriate SPC tools, such as control charts.
Materials
1. Popcorn kernels. A 32-oz bag costs around $1.50 (after taxes) in Walmart. How much
popcorn kernels you need depends on how many “samples” you plan to generate. So plan
it well! TIP: Try not to buy the buttered or flavored variety – counting buttery kernels is
nasty.
2. Microwave oven
3. Microwaveable container and lid. We are not responsible for any damage to property
as a result of this experiment, so please exercise caution.
Methods
Imagine that you are the product scientists or quality engineers in a manufacturing plant that
produces popcorn. Your main job is in ensuring that the plant produces enough popped popcorn
to satisfy demand. Anything un-popped is considered a defect. So what you do is to take n
samples from the production line every t minutes and count the number of “good” or “defective”
kernels. You keep doing this until the end of the shift.
In generating your data for this exercise, you can do the same thing. Cook n kernels every
t minutes and count the number of popped or un-popped kernels (depending on what you want
to track). Some guidelines:
1
• The number of kernels (n), cooking time, and time between cooking (samples) t are up
to you. However, we require at least 20 samples or subgroups, k. Of course, k can be
as many as you want. This implies that you need to cook at least 20 times, so plan your
time accordingly.
• In the real-world, engineers and scientists hate variation, so they do everything possible to
make the variables in the production line stay the same, run after run. Try to control for
all sources of variation that will affect the popcorn yield. In the case that you can’t control
a variable, take note of the time, sample number, and any changes in the experimental
setup.
• Make it interesting. You do not have to have equal sample sizes, n, from run to run. You
can just set a rule of thumb in determining the amount of popcorn to cook, e.g., a small
teaspoon of kernels. But count your kernels!
• I tried n = 24 kernels with a cooking time of 1 minute. About 8 popped. Do little

experiments to help you decide on your experimental parameters. If everything pops
each time, it won’t make for an interesting study. Your goal in this project is not
to maximize the yield, but to generate interesting data sets. So, do not feel
compelled to pop all the kernels.
• Note that all types of charts discussed in this class – variables and attributes, time-
weighted, or multivariate chars – are applicable in this experiment. It all depends on how
you generated the data. HINT: You can cook multiple small bowls at a time, where each
bowl contains n kernels.
Project Requirements
1. A draft of the project is due on 4/28/16, Thursday, in class. The draft should include
parts (3a), (3b), data generated, and the list of SPC tools or control charts you plan to
use in your analysis.
2. The due date for the final project report is 5/10/16, Tuesday, during the final exam.
3. The report (excluding the Appendix) should not exceed 5 pages and should be in
Word file format or PDF. At the minimum, the report should include:
(a) Introduction or Background of the Study

(b) Materials and Methods. You will be graded for the ingenuity of data generation,
so please document your methods as thoroughly as possible. You may include photos
of the set-up (no selfies please).
(c) Analysis. Prepare the control charts based on the data generated. Use software
such as Minitab. Prepare a write-up that addresses the following:
i. Explain why you chose this particular type of control chart(s) or SPC tools to
keep track of yield or defects.
ii. What assumptions are necessary for the control chart or SPC tool to be valid
for this data set? If possible, test these assumptions, e.g., normality.
iii. Is your process in control? Did you notice any systematic source of variation?
iv. In the future, what can you do to improve the yield based on the data?
v. You may also use other tools to derive further conclusions from your data or
to verify assumptions, such as hypothesis tests, tests for normality, sample size
calculations.
2
(d) Conclusion and Recommendations.
(e) Appendix. Include the raw data in tables.
(f) References. Only if you cited relevant work.

Popcorn Experiment

Uploaded by

Copyright:

Available Formats

Popcorn Experiment

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Popcorn Experiment

Uploaded by

Copyright:

Available Formats

Monitoring Popcorn Yield Using Control Charts

SIE 406: Quality Engineering

1. Creativity in data generation. Exert some creativity in translating the experiment

2. Preciseness in data collection. Based on the production scenario, generate data

• I tried n = 24 kernels with a cooking time of 1 minute. About 8 popped. Do little

(a) Introduction or Background of the Study

You might also like