第四十三 Drudgery to Strategy—A Statistical Metamorphosis
本帖最后由 小编D 于 2011-12-16 14:14 编辑
你好，我是小编H。请对以下文章有翻译兴趣的组员留下你的预计完成时间，并发短信息联系小编H，以便小编登记翻译者信息以及文章最终完成时的奖惩工作。感谢支持翻译组！
Drudgeryto Strategy—A Statistical MetamorphosisAstrategy of experimentation can point you toward success
by Lynne B. Hare and Mark Vandeven
Think back to your Stats 101 course. Youentered the first session laden with apprehension—induced by survivors’ horrorstories—and your worst fears were confirmed. Early on, the professor said,"OK, boys and girls, today we’re going to discuss t-tests and confidenceintervals." And you sat there thinking you’d rather visit the dentist.
Tools and occasional toy (artificial)problems characterize many introductory statistics courses. To be sure, theprofessors are gifted, enthusiastic lecturers; many have great humor and humankindness in their veins. But still, the class dealt with statistical toolssuitable for fixed occasions.
At semester’s end, you knew about a bunchof tools you could resurrect if the right occasion ever arose.
It never did.
There’s no getting around the fact thatstatistics is a difficult subject. The thinking is different from that of manyother disciplines: It acknowledges uncertainty, whereas others professdeterminism. It runs counter to everything we are taught in algebra: "Solvefor x, a fixed but unknown quantity."
Further, statistical applications requirefollowing mathematical formulas of which there are relatively few in politicalscience, English, history or philosophy. Even pure mathematics claims closure(not always true), which is absent in statistics. The semester usually endsbefore the professor scratches the surface. And because the first semesterlacks a happy ending, many students are reluctant to proceed to a second.
On the rise Over the last two decades, we’ve noticedstatistical stock is rising at some companies. They may be in the minority, butsome top executives have begun to see statistics’ strategic value.
We believe this occurred because thoseexecutives witnessed very positive (read: dollars to the bottom line) results,especially when scientists and engineers used statistical methods to guideprojects from highly uncertain project beginnings to solid, successful andsustainable products and processes. To their internal statisticians, executivesare saying, "I want more of that. Make it happen."
"Egad," says the internalstatistician. "Now what do I do? All that my colleagues learned from theirstatistics course was not to take another one. The only statistical methodsthey use come at the end of a project, when they compare the new prototypeagainst the current product. And now I have to change the organizationalculture so people not only use the methods, but also use them upstream in thedevelopment process for guidance. I’d rather visit the dentist and have root-canalsurgery!"
Well, this is not an appeal for long linesat dentists’ offices. Positive results can emerge from some lessons learned, ifonly by osmosis, in Stats 101. An appeal to intuition reduces resistance to thenotion that the t-test used to compare a sample mean to some hypothesized valueis analogous to signal-to-noise, for example. Build on this to show that areduction of noise makes it easier yet to hear the signal. Build on it furtherto show that the test can be expanded to compare two sample means as in Stats101, chapter 3, but it’s still signal to noise.
It is a bit of a leap from there to comparemore than two treatment means, but it can be shown that the guiding principleremains akin to signal to noise. Then, if you want to compare multipletreatment means, wouldn’t it be wise to economize by running half theexperimental treatment combinations on one material and the other half onanother? Doing that introduces a two-way classification painlessly—almost. Thediscussion with colleagues dredges up their unpleasant memories, true, but atleast they form a base for intuitive appeal.
If you can study multiple treatments andmultiple materials, why not add multiple speeds? Now you have a three-wayclassification. Conceivably, you could add other factors, such as ingredientlevels, machines and locations. Whoa! Wait a minute and those experiments willget big in a hurry. Big experiments cost big bucks, and they are logisticallyhard to control. Fugetaboutit!
To the rescue come the two-level designsand their fractions. Don’t do multiple factor levels. Examine only two levelsper factor, and make those levels the extremes. "Absurd," you say,"some of those levels could be best!"
"Right," we sayin appropriately humble rejoinder, "but we don’t even know which factorsare important right now. Let’s experiment to find which factors are most likelyto drive success. We can hone in on levels in subsequent experiments." This is the great leap or paradigm shift—wemove from comparing treatment or material A and B to estimating somethingcalled an "effect." The question has changed from "What factorlevel is best?" to "Which factors are most important?"
"You mean I have to do more than oneexperiment?" you ask. Yes. The first one or two identify which factors orcombinations of factors are important. The next experiments help identify bestlevels among the important factors. It is a strategy for experimentation, andit has a higher success rate than competitive approaches to experimentationsuch as spray and pray, try everything, and try your best hunch and hope youget lucky.
We see other advantages to the strategy ofexperimentation, especially as it is used to drive decision making upstream inthe research process. For one, the competitive method of giving it your bestshot and testing at the end simply tells you if you were successful. If youweren’t, you don’t know why.
The strategy of experimentation puts databehind your directional decisions so you know, early in the experimentalprocess, what path to take; there are no blind alleys. You gain product andprocess knowledge along the way.
The strategy of experimentation alsoprovides trade-offs. If practical constraints block the path to using onecombination of factors and their levels for success, there may be anothercombination that comes close. The data point you in the direction towardsuccess.
We’ve found that those who use the strategyare successful and they feel liberated. To go back to the old methods would belike having more oral surgery.
你好，我是小编H。请对以下文章有翻译兴趣的组员留下你的预计完成时间，并发短信息联系小编H，以便小编登记翻译者信息以及文章最终完成时的奖惩工作。感谢支持翻译组！
Drudgeryto Strategy—A Statistical MetamorphosisAstrategy of experimentation can point you toward success
by Lynne B. Hare and Mark Vandeven
Think back to your Stats 101 course. Youentered the first session laden with apprehension—induced by survivors’ horrorstories—and your worst fears were confirmed. Early on, the professor said,"OK, boys and girls, today we’re going to discuss t-tests and confidenceintervals." And you sat there thinking you’d rather visit the dentist.
Tools and occasional toy (artificial)problems characterize many introductory statistics courses. To be sure, theprofessors are gifted, enthusiastic lecturers; many have great humor and humankindness in their veins. But still, the class dealt with statistical toolssuitable for fixed occasions.
At semester’s end, you knew about a bunchof tools you could resurrect if the right occasion ever arose.
It never did.
There’s no getting around the fact thatstatistics is a difficult subject. The thinking is different from that of manyother disciplines: It acknowledges uncertainty, whereas others professdeterminism. It runs counter to everything we are taught in algebra: "Solvefor x, a fixed but unknown quantity."
Further, statistical applications requirefollowing mathematical formulas of which there are relatively few in politicalscience, English, history or philosophy. Even pure mathematics claims closure(not always true), which is absent in statistics. The semester usually endsbefore the professor scratches the surface. And because the first semesterlacks a happy ending, many students are reluctant to proceed to a second.
On the rise Over the last two decades, we’ve noticedstatistical stock is rising at some companies. They may be in the minority, butsome top executives have begun to see statistics’ strategic value.
We believe this occurred because thoseexecutives witnessed very positive (read: dollars to the bottom line) results,especially when scientists and engineers used statistical methods to guideprojects from highly uncertain project beginnings to solid, successful andsustainable products and processes. To their internal statisticians, executivesare saying, "I want more of that. Make it happen."
"Egad," says the internalstatistician. "Now what do I do? All that my colleagues learned from theirstatistics course was not to take another one. The only statistical methodsthey use come at the end of a project, when they compare the new prototypeagainst the current product. And now I have to change the organizationalculture so people not only use the methods, but also use them upstream in thedevelopment process for guidance. I’d rather visit the dentist and have root-canalsurgery!"
Well, this is not an appeal for long linesat dentists’ offices. Positive results can emerge from some lessons learned, ifonly by osmosis, in Stats 101. An appeal to intuition reduces resistance to thenotion that the t-test used to compare a sample mean to some hypothesized valueis analogous to signal-to-noise, for example. Build on this to show that areduction of noise makes it easier yet to hear the signal. Build on it furtherto show that the test can be expanded to compare two sample means as in Stats101, chapter 3, but it’s still signal to noise.
It is a bit of a leap from there to comparemore than two treatment means, but it can be shown that the guiding principleremains akin to signal to noise. Then, if you want to compare multipletreatment means, wouldn’t it be wise to economize by running half theexperimental treatment combinations on one material and the other half onanother? Doing that introduces a two-way classification painlessly—almost. Thediscussion with colleagues dredges up their unpleasant memories, true, but atleast they form a base for intuitive appeal.
If you can study multiple treatments andmultiple materials, why not add multiple speeds? Now you have a three-wayclassification. Conceivably, you could add other factors, such as ingredientlevels, machines and locations. Whoa! Wait a minute and those experiments willget big in a hurry. Big experiments cost big bucks, and they are logisticallyhard to control. Fugetaboutit!
To the rescue come the two-level designsand their fractions. Don’t do multiple factor levels. Examine only two levelsper factor, and make those levels the extremes. "Absurd," you say,"some of those levels could be best!"
"Right," we sayin appropriately humble rejoinder, "but we don’t even know which factorsare important right now. Let’s experiment to find which factors are most likelyto drive success. We can hone in on levels in subsequent experiments." This is the great leap or paradigm shift—wemove from comparing treatment or material A and B to estimating somethingcalled an "effect." The question has changed from "What factorlevel is best?" to "Which factors are most important?"
"You mean I have to do more than oneexperiment?" you ask. Yes. The first one or two identify which factors orcombinations of factors are important. The next experiments help identify bestlevels among the important factors. It is a strategy for experimentation, andit has a higher success rate than competitive approaches to experimentationsuch as spray and pray, try everything, and try your best hunch and hope youget lucky.
We see other advantages to the strategy ofexperimentation, especially as it is used to drive decision making upstream inthe research process. For one, the competitive method of giving it your bestshot and testing at the end simply tells you if you were successful. If youweren’t, you don’t know why.
The strategy of experimentation puts databehind your directional decisions so you know, early in the experimentalprocess, what path to take; there are no blind alleys. You gain product andprocess knowledge along the way.
The strategy of experimentation alsoprovides trade-offs. If practical constraints block the path to using onecombination of factors and their levels for success, there may be anothercombination that comes close. The data point you in the direction towardsuccess.
We’ve found that those who use the strategyare successful and they feel liberated. To go back to the old methods would belike having more oral surgery.
没有找到相关结果
已邀请:
RSS Feed
9 个回复
shulice (威望:0) (贵州 贵阳) 电子制造 技术员
赞同来自:
最近有点忙 暂不约稿