Statistical analysis is not always correct.

[Sam Summey]

We have all read numerous magazine articles that taut the benefits of this load or that load because of its low index of efficiency and deviation of velocity (standard deviation). I recently read an article about a new improved precision 9mm handgun that used about 4 super duper 9mm loads from various manufacturers. Low and behold, the most accurate ammunition had the largest standard deviation. I know, only 4 samples is too small of a sample size to make conclusive scientific or statistical decisions on what was the best ammunition but it did strike me as odd.

 

[GunnyGene]

We have all read numerous magazine articles that taut the benefits of this load or that load because of its low index of efficiency and deviation of velocity (standard deviation). I recently read an article about a new improved precision 9mm handgun that used about 4 super duper 9mm loads from various manufacturers. Low and behold, the most accurate ammunition had the largest standard deviation. I know, only 4 samples is too small of a sample size to make conclusive scientific or statistical decisions on what was the best ammunition but it did strike me as odd.

Speaking as recovering professional statistician, statistics very often only tell you how wrong you likely are, and/or how much you don't know.

 

[Sam Summey]

Speaking as recovering professional statistician, statistics very often only tell you how wrong you likely are, and/or how much you don't know.

I worked for GE, they forced everyone who was exempt to be trained as a "Six Sigma Blackbelt". I resisted this with a passion because I had been a product designer from 1985-1998 and had designed at least 2 products per year through the tooling cycle and implementation without any major product failures. During the "Black Belt Inquisition" which lasted until 2005, the General Electric Company squandered 2.5 billion dollars World Wide without significant profit improvement. When I retired, I took my "Green Belt certification plaque" to my local range and fired a palm size 8 round group in the middle. I took the plaque back to the plant and gave it to the head "Black Belt Guru" and presented it to him and told him I had finally found a use for the "effing" thing. I understand it was sent to GE headquarters.

 

[GunnyGene]

I worked for GE, they forced everyone who was exempt to be trained as a "Six Sigma Blackbelt". I resisted this with a passion because I had been a product designer from 1985-1998 and had designed at least 2 products per year through the tooling cycle and implementation without any major product failures. During the "Black Belt Inquisition" which lasted until 2005, the General Electric Company squandered 2.5 billion dollars World Wide without significant profit improvement. When I retired, I took my "Green Belt certification plaque" to my local range and fired a palm size 8 round group in the middle. I took the plaque back to the plant and gave it to the head "Black Belt Guru" and presented it to him and told him I had finally found a use for the "effing" thing. I understand it was sent to GE headquarters.

I remember that 6 sigma fad from my time with Boeing Commercial (1985-2000). All flash and no cash. And the FAA didn't think much of the process either. We damn near lost our production certificate because of it.

 

[contender]

Sometimes, when building test loads for a specific firearm,, the Standard Deviation of the most accurate load was higher than other loads. It's just ONE factor in helping determine what our reloads are doing. But each firearm is a machine unto itself,, and as such,, can surprise us.

 

[GunnyGene]

Here's where I spent most of my time with Boeing.

How to build a 747-8. Long video, so grab a pot of coffee.

 

[Johnnu2]

I remember those days of six-sigma folly and the "organizational development" fraud (at another equally large corporation). I was able to resist without significant punishment, by never openly criticizing what I knew was just another bag of tricks sold to senior executives by senior executives of other companies. I was fortunate enough to outlast them all (44 years) and be on the Corporate teams to "target" and "eliminate" those very 'strategy du jour' entities along with other "low hanging fruit". As far as STATISTICS are concerned, I have saved myself a lot of gun money by never reading anything with them espousing any gun related wiz-bang "stuff". I wouldn't have understood them anyway!!!
IMHO,
J.

 

[deac45]

No matter the numbers, tight and consistent holes in the target are what matter...with the exception of pressure...that one needs to be watched.

 

[Yaworski]

Here's where I spent most of my time with Boeing.

Did you happen to know Mark Feurstein (not the actor)?

 

[GunnyGene]

Did you happen to know Mark Feurstein (not the actor)?

Don't recall the name. There were 25,000 or so employees at just the Evt. plant, and I did a lot of business at all the Puget Sound facilities and regular trips to Kansas City and various suppliers around the country.

 

[Yaworski]

Don't recall the name.

He was a 747 test pilot. Knew him in college, about a billion years ago.

 

[sceva]

I spent 45 years + in Quality assurance including long stints as a Quality Engineer with several Aerospace manufacturers. I saw Quality circles, 6 sigma, SPC, and so on. They have their place but are not an end all. Do they still have the Baldwin award? I know of companies that practically went broke trying to qualify for one. SPC can be a good tool if used correctly, most companies don't. The company I retired from had customers that advised SPC to get a contract so the Company said YEAH we do that. They did a minimum of instruction, made charts not knowing what or why then filed them in a drawer without doing anything with them; just so they could say "We do SPC" I was able to get a program started in the external grinding dept, oversaw the charting and convinced them to work to the middle of the tolerance ( 0.0005 total tolerance was average with some requirements as low as 0.0003 or as big as 0.001) started getting great results, less turn back from final etc, Then my duties changed and someone else took over. Within a year it was as if I was never there and they were back to where they were a couple of years earlier.

 

[Dan in MI]

Ah yes, the old corporate flavor of the month.

Being in the Detroit area it was hilarious to see what company A was teaching this month, that company B was getting next month, and C just completed. A big stupid circle.

 

[sceva]

The consultants made a lot of money. There is also the ISO certification. A lot of companied tout that they are ISO certified. That only means that they have procedures in place and that they are following them. How good the procedures are is another story. Now AS9100 certification for aerospace is whole other ball of wax. While it includes ISO certification it adds a whole lot more including if the procedures are adequate, are you tracking discrepancies and reducing them etc. Auditors for ISO only don't seem to need to know manufacturing methods etc, only the rules and can you show them you are following them. AS9100 auditors on the other hand come out of the aerospace field, know manufacturing, metrology, Quality assurance methods and they actually talk to people on the floor asking them about their procedures and how they follow them. Much harder to get and keep AS9100 certification.

 

[peachhead]

Lies, damn lies and statistics.

 

[GunnyGene]

The consultants made a lot of money. There is also the ISO certification. A lot of companied tout that they are ISO certified. That only means that they have procedures in place and that they are following them. How good the procedures are is another story. Now AS9100 certification for aerospace is whole other ball of wax. While it includes ISO certification it adds a whole lot more including if the procedures are adequate, are you tracking discrepancies and reducing them etc. Auditors for ISO only don't seem to need to know manufacturing methods etc, only the rules and can you show them you are following them. AS9100 auditors on the other hand come out of the aerospace field, know manufacturing, metrology, Quality assurance methods and they actually talk to people on the floor asking them about their procedures and how they follow them. Much harder to get and keep AS9100 certification.

Aerospace indeed. Especially when the FAA must be convinced that your ISO cert, SPC, Inspection procedures, etc. is actually in line with what your production certificate requires. Makes for some very interesting conversations with the FAA and plant management.