Deming 的 funnel test (漏斗试验) 是证明如果一个 system 是 "in control"，那麼我们就不应妄加修正，以免越改越糟。最简单的例子是在用 con trol charts 时，如果 the system is in control，千万不要以为是 "special cause" 而没法改良。当然，我们应该不断 kaizen 去改善 the "common causes" of the "system" 。
洛客 (2008-7-12 02:26:53)
上文有一 typo。是 "设法"，不是 "没法"。
Deming 的 funnel test (漏斗试验) 是证明如果一个 system 是 "in control"，那麼我们就不应妄加修正，以免越改越糟。最简单的例子是在用 con trol charts 时，如果 the system is in control，千万不要以为是 "special cause" 而设法改良。当然，我们应该不断 kaizen 去改善 the "common causes" of the "system" 。
洛客 (2008-7-12 13:44:03)
没看清 LZ 的问题，抱歉。
Deming Funnel Test，四个改变现状的方法 (Rules):
Rule 1, no adjustment: 当系统运作正常时 (in control) 不 tamper，没有调整。结果是 funnel 下的顯象相当集中 (as expected)。
Rule 2, modest adjustment (move the funnel in the opposite direction relative to the last position)，顯象的结果比 Rule 1 差，但并不太坏: 在当兵作实弹射击时，通常在每一发前都会根据上一发着弹點的差度作调整，(因为忽略了许多不可抗拒的因素如风速等) 结果是得不赏失，越调越糟。
Rule 3, extreme adjustment (move the funnel in the opposite direction relative to the target)，顯象的结果比 Rule 2 更差: 某公司订下规则在景气好时薪资要比市埸高10%。但某年遇到市场不景气，公司决定薪资调降到比市场低10%。结果是员工都跑了。
Rule 4, just like the last adjustment (move the funnel directly above the last drop)，这是最糟的改变: 最著名的例子是当年有家日本汽车公司从 DuPont 买漆。每次漆的样本就是上次剩下的。数月後该汽车公司发现最新一批的颜色和数月前进货的一批大不相同。
Ref.: Rafael Aguayo: Dr. Deming, the American Who Taught the Japanese about Quality
Strategy Description
1 Do not move the funnel, keep it aimed at the target
This is the best strategy, it is the one that will minimize the scatter
2 Move the funnel from its previous position a distance equal and opposite to the last error
This is akin to making a part, measuring it and then adjusting the machine to compensate. In the salinity example the sample mean was 52 and the machine would be adjusted to decrease the salinity by two units. This strategy will increase the variation by a factor of two
3 Move the funnel from the target a distance equal and opposite to the last error
The logic here is that the machine is badly calibrated and so the salinity level setting is reduced from 50 to 48. On the first drop this is no different to strategy 2, but subsequent adjustments will be larger than for strategy 2, and lead to increasingly large swings
4 Move the funnel over the last marble
A young man in his first job was told to cut pieces of rubber tube to a particular length. As he cut each piece he used it to measure for the next cut. After cutting a few dozen he realized that the tube lengths had gradually decreased and were now about half what they should be, and useless (the tube was taken from a coil and the error was caused by the curvature Another common example is when workers are given instructions that they pass on to others who join later and pass them on in turn
最新回复
398534378 (2008-7-11 16:54:19)
洛客 (2008-7-12 02:22:29)
洛客 (2008-7-12 02:26:53)
Deming 的 funnel test (漏斗试验) 是证明如果一个 system 是 "in control"，那麼我们就不应妄加修正，以免越改越糟。最简单的例子是在用 con trol charts 时，如果 the system is in control，千万不要以为是 "special cause" 而设法改良。当然，我们应该不断 kaizen 去改善 the "common causes" of the "system" 。
洛客 (2008-7-12 13:44:03)
Deming Funnel Test，四个改变现状的方法 (Rules):
Rule 1, no adjustment: 当系统运作正常时 (in control) 不 tamper，没有调整。结果是 funnel 下的顯象相当集中 (as expected)。
Rule 2, modest adjustment (move the funnel in the opposite direction relative to the last position)，顯象的结果比 Rule 1 差，但并不太坏: 在当兵作实弹射击时，通常在每一发前都会根据上一发着弹點的差度作调整，(因为忽略了许多不可抗拒的因素如风速等) 结果是得不赏失，越调越糟。
Rule 3, extreme adjustment (move the funnel in the opposite direction relative to the target)，顯象的结果比 Rule 2 更差: 某公司订下规则在景气好时薪资要比市埸高10%。但某年遇到市场不景气，公司决定薪资调降到比市场低10%。结果是员工都跑了。
Rule 4, just like the last adjustment (move the funnel directly above the last drop)，这是最糟的改变: 最著名的例子是当年有家日本汽车公司从 DuPont 买漆。每次漆的样本就是上次剩下的。数月後该汽车公司发现最新一批的颜色和数月前进货的一批大不相同。
Ref.: Rafael Aguayo: Dr. Deming, the American Who Taught the Japanese about Quality
转载请注明出自六西格玛品质论坛 http://bbs.6sq.net/,本贴地址:http://bbs.6sq.net/viewthread.php?tid=185173
georgecho (2008-7-12 18:44:31)
一个是以上次瞄准点为基准反方向调整，另一个是以目标值为基准反向调整。能不能在同一个例子里反应出来？
洛客 (2008-7-13 06:01:45)
Rule 1: the system is stable, 不作调整，再射。
Rule 2: 将炮位朝前推进 8 公尺 (射程仍设在 5000 公尺) 再射。
Rule 3: 炮位不动，将炮重新设在 5008 公尺，再射。
Makes sense?
georgecho (2008-7-13 13:38:43)
我一直在想找一个例子表达它们的不同，现在看来并不是例子的问题，而是只修正瞄准位置的一次的例子无法表达它们的不同，请看图：
漏斗实验.JPG
洛客 (2008-7-13 17:06:47)
wang168 (2008-7-13 20:41:36)
有兴趣者请连结下载
Demonstration of the Deming Funnel Experiment
http://www.minitab.com/support/macros/default.aspx?id=25
movir (2008-7-13 22:16:47)
wutao5415 (2008-9-23 11:01:14)
1 Do not move the funnel, keep it aimed at the target
This is the best strategy, it is the one that will minimize the scatter
2 Move the funnel from its previous position a distance equal and opposite to the last error
This is akin to making a part, measuring it and then adjusting the machine to compensate. In the salinity example the sample mean was 52 and the machine would be adjusted to decrease the salinity by two units. This strategy will increase the variation by a factor of two
3 Move the funnel from the target a distance equal and opposite to the last error
The logic here is that the machine is badly calibrated and so the salinity level setting is reduced from 50 to 48. On the first drop this is no different to strategy 2, but subsequent adjustments will be larger than for strategy 2, and lead to increasingly large swings
4 Move the funnel over the last marble
A young man in his first job was told to cut pieces of rubber tube to a particular length. As he cut each piece he used it to measure for the next cut. After cutting a few dozen he realized that the tube lengths had gradually decreased and were now about half what they should be, and useless (the tube was taken from a coil and the error was caused by the curvature
oxmhjf9975 (2008-9-23 11:22:59)
烟头灭了 (2008-9-26 10:14:59)