posted 26 February 2001 11:14 AM               

Any suggestions out there on how I can improve my knowledge? Guidance from more experienced heads would be appreciated...
I'm willing to read books, study websites, attend courses, etc.

posted 26 February 2001 05:32 PM               

If you need something a little more advanced, look for "Process Quality Control, Troubleshooting and Interpretion of Data" by Ellis Ott, Ed Schilling, and Dean Neubauer. (ISBN 0-07-1350101-1) The book will cover how to analyze data given a problem situation. It assumes you have a basic understanding of statistics. About $75.00 a copy.

A final note, you can not go wrong with Juran's Quality Handbook although you might go broke paying for it at $135.00 a copy.

posted 27 February 2001 02:17 PM               

To gain a better understanding of topics listed in the book, the reader must go to other sources for the most part. It isn't the type of book one would read cover to cover.

Regards,

posted 27 February 2001 02:28 PM               

Good point! I prefer the other books but Juran is a good reference. Any suggestions for Casana on your favorites?

Regards,

posted 27 February 2001 11:08 PM               

Kind of a nice compact Juran - not to in depth though - just refreshers.

Otherwise the above references are good.

posted 28 February 2001 10:43 AM               

Quality Planning and Analysis, authored by Frank Gryna.

I don't have the new version but the other versions were very helpful.

posted 28 February 2001 12:48 PM               

Thinking.....thinking.....thinking.....

Al's suggestion of Juran/Gryna's "Quality Planning and Analysis" is a good place to start I think. It is easier to read than Juran Quality Handbook (IMO) and exposes the reader to things like Precontrol Planning that might help to develop a good foundation to understanding sampling plans (step by step approach). Either of these books will reference other reading. In my limited experience, Juran does a nice job of pointing the reader in the right direction. He often takes a lot of flack for that, but in reality, I feal that he is just directing folks to better authorities on subjects than himself.

My favorite SPC/Sampling book would undoubtably be "Statistical Quality Control" by Eugene Grant and Richard Leavenworth. The authors introduce the reader to SPC/SQC in the first half of the book and then into Sampling and Sampling Plans in the second half. Dry, but informative.

On statistics itself, I have read a few books, most pretty informative, but equally dry. One book that I read was pretty good on explaining statistics in ways that were practical yet informative. I do not recall the title or the author. Oddly, the cover color sticks in my mind. With a glance, I should be able to locate it in my library. I will post it here if anyone is interested.

Statistics is a interesting subject. My personal feeling is that we should make this a prerequisite in High Schools everywhere. Folks are prone to base decisions on chance occurrences to often. With a basic understanding of Statistics, I believe that decision making in business and in life itself would be improved. Just a feeling.

Regards,

posted 28 February 2001 02:20 PM               

You get my vote for school board czar. There is far to little emphasis placed on the use of statistics today. That may be due to the extremely poor instructional methods used. Statistics is extremely dry if not placed in a context that applies to the learners frame of reference.

Regards,

posted 28 February 2001 06:28 PM               

Back in High School, I remember them offering a class on Statistics as a math elective. When I read the paragraph used to sum up the course objectives, I remember thinking to myself that this couldn't be a serious course! I also recall thinking, "Boring!" I never signed up. I think you are right in your assessment. They need to make it real for the student and the teaching method must demonstrate how it is pratical.

While studying for the CQE, I reconsidered my position on Statistics. The more I researched, my interest grew. This lead me to search for a better guide to Statistics. The book I can't recall the title and author for had many examples about Gambling and rolling dice (were statistics has its origin), that many of us can relate with on a fun level. It helped me through the learning process.

These days, on my commute home, I listen to to the radio, sports talk mostly. Never a night where real statistics are driven to another level. Talk show hosts and the listeners do a nice job of oversimplifying sports figures. I don't know whether to laugh or cry at times.

Talk about sampling: how many times have you heard someone making a prediction based on last Sunday's Football results? Sample size: one.

Regards,

Kevin

p.s. Casana, congrats on your CQE by the way!

posted 01 March 2001 08:36 AM               

Practical Statistics, Simply Explained by Russell Langley. This book covers Probability, Sampling, Averages & Scatter, Design of Investigations, & Significance Tests. At the end of each chapter, the author has questions to solve to assist in the learning process. It is the best book I have read on Statistics. The author does a nice job of connecting principles to industry practice and gives a short historical connection for items which I found broke up this 'dry' topic. In my opinion, this book will probably have you thinking differently about news stories and the misuse of enumerative studies. He begins his book by briefly detailing the common errors that lead to misleading conclusions. This alone is pretty powerful. This is an older book in reprint, but worth the investment.

How To Think About Statistics, by John L. Phillips, Jr.. This book was written to assist his classes (he taught educational psychology at a university) as he noticed that his students lacked basic understanding of statistics. He wrote the book in such a way that anyone without basic knowledge could gain a good understanding of statistics. He had friends and students who did not understand any statistics read the book and comment on how much they grasped. Most thought it was a hit and other students in other classes in the university would use this guide (as it was orginally published) to assist their learning processes. I would say that this book was pretty decent and the author does a nice job of connecting each chapters topic into several industry examples at the end of each chapter. I did not find this as easy a read as with Practical Statistics, but a good book just the same.

Regards,

posted 01 March 2001 01:11 PM               

That is an excellent point! Deming recommended that any organization going through the Transformation should go out an hire a Master Statistician first. He stated very clearly the need to learn from a Master and not from "a hack". Doing otherwise, he would say, you deserve what you get(pretty harsh, but correct).

His business card labeled him to be a consultant on statistics though he is better known for his contributions to Management Theory (a Paradigm Pioneer). Drucker labels Deming as they only major contributor to management theory in the last century by the inclusion of statistics. Otherwise, Drucker feels that most everyone else has utilized or modified in some small way Fredrick Taylor's contributions.

Regards,

posted 01 March 2001 05:40 PM               

quote:

---

Originally posted by Rick Goodson:
Kevin,

You get my vote for school board czar. There is far to little emphasis placed on the use of statistics today. That may be due to the extremely poor instructional methods used. Statistics is extremely dry if not placed in a context that applies to the learners frame of reference.

Regards,

Rick

---

Maybe I can be the czar-ette then? I've been trying to get our district to add AP choices for math. One being statistics. (I'm on the board.) It's fun to use statistical theory when being presented with the "trend on increasing test scores" - (one data point) and then asking the Deming-question regarding if they "know exactly what caused the scores to increase." The old "without theory, there would be no knowledge." Unless we know why the scores went up, we can make no prediction that they will continue to go up. I would predict that every meeting there is mis-use of statistics.

posted 02 March 2001 09:16 AM               

Sam, I agree learning from a "master" is a great way, but some of us don't have easy access to masters so books can be an alternate source. (Besides, how easy is it to find a master? Aside from this BB, of course )

Laura, I haven't read Deming, partly because I've felt intimidated by all he's written and didn't know where to start. I'll give your book suggestion a try. BTW, I loved your school board story, you are SO right that people just don't understand stats - look at the debacle about the census, for example.

posted 02 March 2001 05:13 PM               

I guess you are the czar-ette then!

Folks are quick to make comparisons of data from one community to another without ever agreeing to Operational Definitions. Direct comparisons like these are mostly meaningless. Still, this doesnât deter us from trying. While intentions are often good, the results from decisions based on data and bad theory are too often not what we hoped they be. I am guilty of this crime, though I try to avoid it.

Your point regarding the moving Îaverageâ is well taken. I can see the blank stares you saw in my mindâs eye. People are so accustomed to using the prevailing paradigm in education that they cannot see anything else (paradigm paralysis). Teaching them basic statistics might improve their understanding, but more will be necessary. For this, I recommend that you read Myron Tribusâs papers on Deming in Education (if you havenât already). He and David Langford have done tremendous things in this area. Those following this thread may find his papers posted at the DEN (Deming Electronic Network) in pdf format for download. In many of the papers, Myron speaks about the costs of education as they could be and as they currently are. He offers a new paradigm for education. Spending more does not ensure better results! You are right that money is not the big issue here, education is.

Casana,

Learning is intrinsic. It is a life-long journey. Books are an alternate source for sure, but if at all possible, we should try to find mentors with knowledge. Books contain information, not knowledge. How will you know if what you read is correct? This is where two brains are better than one. I find that by discussing what you have read often confirms your understanding or causes you to reject it. There is nothing wrong about taking a position on things. It is important that anyone tests theory, refine it, and share it. I constantly learn from the folks here at the Cove. Being a bookworm, it is easy for me to believe my own B.S.! Discussing it here helps me to take the information that I read and process it further into knowledge. As Dr. Deming used to say, ãThere is no substitute for knowledge.ä I continue my pursuit.

I encourage you to read Dr. Demingâs books. I believe you might be pleasantly surprised.

Regards,

posted 08 March 2001 10:17 AM               

I have done the red bead demo using "faux" materials. Not as nice as a kit, but works. I have purchased dry white and red beans from the grocey store and used those. A plastic cup works in place of the paddle.

I don't know what might work with kids (use to be one, but don't remember back that far). I sometimes use height of people, time to drive to work, or size of oranges at the grocery when I am explaining normal distribution. Maybe perdicting how often they win at a video game might interest them. If your son is in to sports you might apply some statistics to figuring batting or rebounding averages. Hope this helps.

posted 12 March 2001 09:36 AM              

His paper says to run a 2^(7-3) fractional factorial design (16 runs, Res IV) using the following factors & levels: Paper Weight (P: Light, Heavy), Tail Width (W: 1.25", 2"), Tail Length (TL: 3", 4.75"), Wing Length (WL: 3", 4.75"), Fold (F: No, Yes), Taped Tail (TT: No, Yes), Taped Wing (TW: No, Yes).

He gives nice instructions in his paper, but here is a site with similar instructions: http://www.dot.state.mn.us/aero/aved/publications/skyslim3/helicopter.html

Note that Ron doesn't fold the bottom of the tail as shown, but instead puts a paperclip on the bottom of the tail. In this site E=W, C&D=TL, and A&B=WL. Clip off the ends of C&D and A&B to shorten the lengths. For F, he folds the upper right corner of the area between C &A, and the lower right corner of the area between D & A for better aerodynamics. For TT he tapes the vertical opening of the tail shut - where the top of D meets the bottom of C when folded. For TW he puts tape along the bent part of A & B - presumably to make the area a little stiffer.

The objective is to measure the time to fall from at least 10-15 feet with a stopwatch and try to understand what affects the time and then maybe maximize that time. You could premake these ahead of time if needed - it might also assure more helicopter-helicopter consistency.

Let's see, how can I do this:

Let AC = BD = area between A & C and between B & C = 1.5"

Width of paper longways = WL + AC + TL - the ends are trimmed for short/long tails/wings

Width of paper shortways = 4.25", the center region equals W. I'd trim outside of C & D to match W's width for easy folding.

Yes, there is confounding of 2-way interactions, but you could weed out insignificant effects and rerun with a smaller design.

posted 12 March 2001 06:17 PM               

Can't take credit for the paper plane presentation (Ken K). I didn't give it a close read, so I can't comment on it.

You might be suprised as to where statistics can be applied. Most folks figure statistics are to be used in Manufacturing(SPC mostly). The fact is, SPC can be applied to any process. Variable charts, Attribute charts, or run charts for instance. I use run charts to monitor inbound material flow, inspection efficiency, and inbound phone call answering. Run charts are terrific for helping to understand flow and determining if your processes (paper trails or stamping presses) are in control. They present Corrective and Preventive action opportunities.

Regards,