Home › Forums › General Forums › New to Lean Six Sigma › Questions Regarding an Experimental Design Project
This topic contains 4 replies, has 5 voices, and was last updated by Robert Butler 1 week, 2 days ago.
guys,
I was ask do experimental design project with four factors not work related
please see all details below:
now
as I will do all at my own here is what I have in my mind
boiling egg
factors
pan size,
water temperature
time
cooker type
since it will have to be repeated 16 times is that mean each factor above has to be measured 16 at two levels or only eg temperature and time?
this is all new to me and asking on college forum did not get much clear respond and either from lecture.
other idea
strength of the ruler suspended between two glasses at distance 10 and 20 cm
at the middle cup hanging on eye
with use of pennies coins measure time and and number of coins required to bent ruler down 1 or 2 cm.
factors
space between glasses
ruler type
time
number of coins
if anyone can put me on right track I will be much appreciate
or point where im doing mistake
Objective:
Plan, conduct and analyse an experimental design (non-industrial or work related)
consisting of the following:
• Minimum of four factors
Use Minitab to analyse the results. Produce a report of the experiment that contains
details of the following:
• The specific objectives of your experiment. Include details of what you set out to
evaluate or optimise.
• How you planned and conducted the experiment.
• How you attempted to minimise systematic variation.
• How you analysed the design. Include results, tables, graphs etc. Explain what the
results mean.
• Effects
• ANOVA
• Response Graphs
• Analysis of Residuals
• Develop the empirical model and explain.
• Predict the optimum response of the product or process.
• Perform confirmatory runs.
• Write a conclusion.
• Perform a small number of hand calculations that concur with the Minitab results.
This could include:
• Effects
• Sum of Squares
• Residuals
• Response Graph
Peter: It looks like you are learning very basic two-level factorial design of experiments. Search this site (and the web for that matter) on those terms. I’m sure you’ll find plenty of information that will help you learn. If you still have questions, then come back here with specifics and we can clarify.
Good inquiry…
Peter ; basic level……But yes it still can be an DOE, if you are using minitab or similar software might help. Finalize the factor levels – high and low , based on the number of Runs you may go ahead.
Start doing it you will learn more than what any one can share with you.
The biggest issue I see with respect to your first design is that of the response. it appears that it would be egg hard boiled yes/no. That kind of response is going to put you in the realm of logistic regression. This also means your response will be expressed as a percentage of boiled and a percentage of not boiled for each experimental condition and in order to get estimates of percentages that will test as significant under these circumstances you are going to have to run through a LOT of eggs. This, of course, assumes you have chosen ranges of the variables so that there are percentages of boiled and not boiled associated with each of the variables of interest.
Your second experiment sounds possible but there are a few questions.
1. What does time have to do with anything? It would seem that you will have to identify a different variable in order to have a list of 4.
2. You will probably need to use meter sticks if you are going to get a deflection of 1-2 cm for weights that are “reasonable”.
© Copyright iSixSigma 2000-2017. User Agreement. Any reproduction or other use of content without the express written consent of iSixSigma is prohibited. More »