Re: [R] OT: DOE - experiments for teaching

From: Thomas Kaliwe <>
Date: Wed 10 May 2006 - 02:44:08 EST

Thank you all, it helped me a lot.


-----Ursprüngliche Nachricht-----
Von: [] Gesendet: Samstag, 6. Mai 2006 02:17
An: Berton Gunter
Cc: 'Spencer Graves'; 'Thomas Kaliwe'; Betreff: Re: [R] OT: DOE - experiments for teaching

Since ECHIP no longer exists, I guess it is OK to reveal some of the stuff we did. We taught a course to engineers for more than 20 years and

our statistics indicated that 9 out of every 10 attendees ran at least one experiment after the course.

One of the things that was done was intended to break the mind set of the students, and we did this by posing a problem for them to solve before we attempted to teach them anything. We usually used the funnel, although in the early days it was a computer simulation. Many students had some idea of design to begin with, and tried to put what they knew to work with mixed success, while others rolled their own. Teams were used, and someone from each team reported the results. Criticism was always constructive. We followed this up by solving the problem using a design, and we explained the methodology as we went along with all students collecting and analyzing the data on their own computers.

Almost all students (even those reluctantly required to attend by management) bought into the problem; becoming so interested, that we often had to chase them from the classroom at the end of the day.

Berton Gunter wrote:

> I've had fun and luck with the apparatus described in my little paper:
> The American Statistician, 47, 4 p. 265-269 (1993)
> We continue to use this in our industrial training.
> I also would strongly second Spencer's remarks re the difficulty of
> students see the big picture. For some reason, viewing experimentation as
> part of an overall learning process/strategy does not seem to be part of
> most scientist's or engineer's formal education. I suppose if you look at
> typical science or engineering labs where the goal is to come to a > predetermined conclusion, it's not hard to see why. But we don't need to get
> into that imbroglio here.
> -- Bert Gunter
> Genentech Non-Clinical Statistics
> South San Francisco, CA
> "The business of the statistician is to catalyze the scientific
> process."  - George E. P. Box

>>-----Original Message-----
>>[] On Behalf Of Spencer Graves
>>Sent: Friday, May 05, 2006 3:59 PM
>>To: Thomas Kaliwe
>>Subject: Re: [R] OT: DOE - experiments for teaching
>> I fully endorse Richard Heiberger's recommendation of
>>the Bill Hunter
>>articles on teaching experimental design. For a
>>college-level semester
>>D0E class, I had students do experiments in groups. I found
>>it wise to
>>have them do a preliminary presentation with a discussion of the
>>experimental design plus their protocol for managing all the
>>details of
>>test materials, data collection, etc., then a final report with the
>>results. Many students did fine, but some were clearly
>>clueless about
>>the whole process, which indicated a need for some adjustment
>>in what I
>>taught or in some individual assistance.
>> If this is just a few hours or a 1-day thing, you
>>might consider
>> hope this helps.
>> Spencer Graves
>>Thomas Kaliwe wrote:
>>>I'm sorry for this not being related to R but I think this is a good
>>>place to ask. I'm looking for DOE examples(experiments)
>>that can be done
>>>at home or in class, such as Paper Helicopter, Paper Towel
>>etc.. I'm
>>>thankful for any comment.
>>> [[alternative HTML version deleted]]
>>> mailing list
>>>PLEASE do read the posting guide!
>> mailing list
>>PLEASE do read the posting guide!
> ______________________________________________
> mailing list
> PLEASE do read the posting guide! >
Bob Wheeler ---
    ECHIP, Inc. --- Randomness comes in bunches.

______________________________________________ mailing list
PLEASE do read the posting guide!
Received on Wed May 10 03:20:14 2006

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Wed 10 May 2006 - 04:09:59 EST.

Mailing list information is available at Please read the posting guide before posting to the list.