BARRIERS TO ADOPTION, AND THE FUTURE OF RESAMPLING Chapter 00 ended the saga of resampling by asking what will happen with resampling in the future. In 1976, I confidently predicted that most everyday statistics eventually would be done the resampling way. I still believe that, because sooner or later progress usually wins out. But now I also worry how long it will take. How many more bad analyses will be done in science and business because black-box formulae are misused? How many more students will suffer and be turned off from the extraordinarily valuable tool of thinking and acting that is statistics? RESISTANCES TO USING AND TEACHING RESAMPLING Early on, leading statisticians either did not accept the basic idea of resampling, or ignored it. Nowadays they say that the method is a great breakthrough but should not be taught to introductory students. For decades, numerous technical journals rejected articles on the method because it is too simple and lacks "real mathematics". For seventeen years publishers turned down a text of mine about resampling on the grounds that the ideas are sound but that there would be no market because instructors would not accept them - and the publishers might have been right. The National Science Foundation rejected applications for grants in several categories, on assorted grounds. School systems have simply been too preoccupied with their usual business to be willing to develop new curricula. Meanwhile, the American Statistical Association has invested large amounts of money and effort in developing a video series and printed materials to try to teach the old ways more effectively. There is no conspiracy against resampling. But individually, just about every channel has been and still is seriously narrowed even if not closed completely. This is despite the fact that no one - and I mean no one - any longer denies the basic validity or the practical usefulness of these ideas. It is not surprising that a discipline defends its turf; every guild does so <1> Commercial distribution of RESAMPLING STATS has been undertaken by Peter Bruce and myself as a way to reach potential users with a tool that greatly facilitates resampling, thereby disseminating knowledge of it. To date, however, sales have not nearly approached a self-sustaining endeavor. I have financed these efforts out of savings because there has seemed no other way to give these ideas a chance to be used. Efron remarked that "Good simple ideas...are our most precious intellectual commodity, so there is no need to apologize for the easy mathematical level." (1981, p. 1, italics in original). He has greater faith than I do in people's willingness to appreciate a "precious intellectual commodity." I expect the reactions of interested parties to resemble the following episode in the history of the British Parliament, as related by Charles Dickens (quoted by Dantzig, 1954, pp. 23-24): Ages ago a savage mode of keeping accounts on notched sticks was introduced into the Courts of Exchequer and the accounts were kept much as Robinson Crusoe kept his calendar on the desert island. A multitude of accountants, bookkeepers, and actuaries were born and died....Still official routine inclined to those notched sticks as if they were pillars of the Constitution, and still the Exchequer accounts continued to be kept on certain splints of elm-wood called tallies. In the reign of George III an inquiry was made by some revolutionary spirit whether, pens, ink and paper, slates and pencils being in existence, this obstinate adherence to an obsolete custom ought to be continued, and whether a change ought not be effected. All the red tape in the country grew redder at the bare mention of this bold and original conception, and it took until 1826 to get these sticks abolished. In 1834 it was found that there was a considerable accumulation of them; and the question then arose, what was to be done with such worn- out, worm-eaten, rotten old bits of wood? The sticks were housed in Westminster, and it would naturally occur to any intelligent person that nothing could be easier than to allow them to be carried away for firewood by the miserable people who lived in that neighborhood. However, they never had been useful, and official routine required that they should never be, and so the order went out that they were to be privately and confidentially burned. It came to pass that they were burned in a stove in the House of Lords. The stove, over-gorged with these preposterous sticks, set fire to the panelling; the panelling set fire to the House of Commons; the two houses were reduced to ashes; architects were called in to build others; and we are now in the second million of the cost thereof. Some statisticians are frank about viewing resampling as a frontal attack upon their own situations. One colleague put the matter bluntly to my colleague Peter Bruce: "You can give one class to my course", he said, "just as long as you don't tell the students that the rest of what they learned in the course is wrong or useless". To be safe, he offered only the last session of the semester, when it would be too late for the students to raise pointed questions to him, and it would be after the course evaluation had been administered. The simple fact is that resampling devalues the knowledge of conventional mathematical statisticians, and especially the less competent ones. By making it possible for each user to develop her/his own method to handle each particular problem, the priesthood with its secret formulaic methods is rendered unnecessary. No one - not handloom weavers in England two hundred years ago, not the physicians who used leeches - stands still for being rendered unnecessary. Instead, they employ every possible device fair and foul to repel the threat to the economic well-being and their self-esteem. At a seminar to bio-statisticians at the National Institutes of Health on December 14, 1993, one of the audience said at the end of the talk: "This is a very egalitarian method. But if users can understand everything that's being done, and can create their own methods, what do they need me for?" That, of course, is one of the core resistances to the resampling method. Machiavelli explained in The Prince about the self-interest resistance to innovation: [T]here is nothing more difficult to plan, more doubtful of success, nor more dangerous to manage than the creation of a new system. For the initiqtor has the enmity of all would would profit by the preservation of the old instituations and merely lukewarm defendes in those who would gain by the new ones (quoted by Gilder, p. 50). In commenting on Mandelbrot's work with fractals, and some other discoveries, Brenner says: ...when, after years of dedication, one's work is suddenly threatened with oblivion...the blindness [to new ideas] here seems calculated: the new approaches threatened reputations, status, and wealth. The reasons given for the rejection were either that the innovator-discoverers were `outsiders', or that they were working at low-status institutions" (Brenner, 1991, p. 536) I have no objection to mathematicians teaching mathematics to those who wish to learn it, for purposes practical or esthetic. I do object to their forcing mathematical methods upon people who need usable techniques and fundamental understanding of statistical inference, both of which are put out of people's reach with the formulaic approach. At a less-elevated level, it is not unfair to say that the teaching of statistical inference as mathematical manipulation is largely a racket of third-class minds to protect themselves from having to understand the "philosophy" and applications of their subject. For example, the instructor of the Department of Statistics' introductory course at one university explicitly says the course requires a high level of mathematical skill to "weed out" the "poorer" students on behalf of the engineering department. And I have heard some of the least wise and least productive of professors refer to their introductory statistics as "dumb" because they do not follow the formulas written on the blackboard. (Among others, that's me whom they're talking about - who doesn't easily absorb the formulas because I want to know what each symbol represents.) SO WHAT WILL HAPPEN? A lot depends upon chance in the diffusion of any new idea, a state of affairs that is particularly fitting for the subject of statistics. It was luck that Efron came along and - working through the very different route of the jackknife than my own route from first principles - found his way to resampling. Efron's rediscovery and promulgation of the bootstrap has had an enormously powerful impact; without him, the idea and topic surely would still be unknown to most or all professional statisticians. By 1993, it is apparent that resampling techniques have caught on like wildfire among statisticians. They are busily exploring the properties of the bootstrap, and applying it regularly to problems that are difficult with conventional analysis. The number of papers exploring and applying the bootstrap and (inevitably) related methods is growing by leaps and bounds (whether logistically or exponentially or whatever I have not measured). The bootstrap seems to be kindling interest in the other resampling procedures. But this explosive development has been concentrated among mathematical statisticians who are interested in the method's properties rather than in its instruction and daily workaday use. Almost no one has made it her/his business to take these ideas to teachers and students and introduce resampling into the regular curriculum in high schools and colleges. The American Statistical Association has recently moved a bit in this direction, as has the National Council of Teachers of Mathematics which urges that simulation be given as much attention as analytics in teaching probability. But neither of these groups has as yet picked up on function-oriented computer programs such as Resampling Stats which implement the simulation method in a fashion which is at once workable, understandable, and enjoyable. The gulf between the frontiers of mathematical statistics and the practices taught in introductory courses can continue to be wide for a long time. It was luck, too, that Peter Bruce came along when he did. Bruce, a former foreign service officer and recent MBA, who is now promoting the method full time, immediately perceived the promise of resampling and initiated the Resampling Project to promulgate the fundamental ideas on a full-time academic and commercial basis. And it was luck that he came along when it seemed to make sense for me to redirect (at least for a few years) a substantial portion of my time and energy away from the economics of population which had occupied me almost completely for two decades. THE FUTURE FROM A JOURNALIST'S POINT OF VIEW A reporter once asked me how I would write a news story about resampling that might have some dramatic appeal. I said I felt whipsawed between (on the one side) the negative impression I would leave if I seemed self-promoting, and (on the other side) the need to make every possible claim in order to promote resampling. With some embarrassment, I suggested this: Statistics Establishment Only Grudgingly Accepts Radical Method Which Boosts Learning and Productivity Outsider Beards Establishment For Third Time It took Julian L. Simon twelve years to get even a trial for the airline volunteer plan which is now in use at every American airport every time too many passengers with tickets show up for a flight. The industry and government officials and "experts" jeered at it as being impractical, impossible, and ridiculous. But when it was adopted by the Civil Aeronautics Board in 1978, it became an immediate success and has been hailed even by Ralph Nader -- who originally opposed it -- as the best consumer program ever. It took a decade for Simon even to get a hearing for his ideas about population and resources. And when the hearing came, the ideas were roundly rejected as preposterous by the environmental movement en masse, and were hardly adopted enthusiastically by professional colleagues. But by 1986 the National Research Council of the National Academy of Sciences had issued a report in response to Simon's ideas that largely validated his work, and turned almost wholly away from the previous National Academy report on the subject. Now Simon has gone public with his campaign to enable non-statisticians as well as statisticians to deal with all their problems in statistics and probability -- in business, sports, science, and technology -- with the RESAMPLING STATS technique, computer language, and computer program. This method gets rid of all the formulae, tables, and mathematical magic which have driven generations of confused and frustrated university students to curses and tears, and substitutes simulated experiments which are easy to understand by everyone from professors down to seventh graders. Simon has been developing and publishing these methods since the 1960's. Development of the bootstrap method by Bradley Efron in the late 1970's has given wide legitimacy to the use of resampling methods. But once again, the people whose careers depend upon the status quo - teachers, administrators, research statisticians - do not welcome Simon's contribution. Resampling ideas can help baseball managers make sensible decisions about whether a batter is in a true slump and should be benched, or instead is just suffering a bad run of chance...basketball coaches whether a player has a cold hand...manufacturing people whether a process is out of control and should be adjusted, or should be left alone...cancer researchers whether a new treatment shows true promise or has insufficient evidence in its favor...marketers whether a new promotion is really better than the old one, or is probably just doing better by luck...investors whether a money manager has better-than-average skill or has just had good breaks...gamblers whether the odds offered to them in complex wagers give them the edge or a skinning." To confront the opposition to these new ideas with a challenge that dramatizes the power of the resampling method, Simon offers to wager $5000 that after just six hours of instruction and practice, people will produce more correct answers to realistic problems than after 12 or 18 hours of conventional instruction. Other computer languages such as Basic, Pascal, and Minitab require complex and counter-intuitive programs in carry out the simplest resampling operations. RESAMPLING STATS is perfectly intuitive and transparently understandable even those who have never touched a computer and who think of themselves as without any mathematical ability. Controversy makes a news story interesting. And critics abound who take a different view of the matter than Simon. Some charge that Simon is a credit- grabber, saying: Simon didn't develop the method; Efron did. Indeed, there is no question that Efron originated the investigation of the properties of one sub-category of resampling, the bootstrap, in such convincing fashion that mathematical statisticians throughout the world have leaped onto his bandwagon. That is an extraordinary achievement. But that should not obscure Simon's earlier achievement of staking out the entire field and showing what the bootstrap and other methods could do, even if Simon did not also begin the study, as Efron did, of what the method could not do. Critics also say, more personally, that Simon loses his claim to discovery of the resampling method because he did not develop it, instead working on other topics. Simply having published the method is not of much value, they say, unless one affects the discipline and/or the public at large. In his glummer moments Simon tends to agree. "The ideas that flood in on me are the bane of my life", he says. "We'd all probably be better off if I never got the idea to use resampling methods in the 1960's. It may benefit no one, and my other work is the poorer for it. But then", he muses, "if I had stuck to the economics of advertising on which I worked in the mid-1960s..." and he stops there. Another explanation for Simon's lack of success in promoting resampling is that he is by profession an economist (and professor of business), and his central interest - the economics of population - has absorbed most of his energies over the years. Furthermore, he is not a card-carrying statistician, which makes it more difficult for him to relate to the statistical establishment. Over the years he has made a vast number of attempts, along a great number of lines, to interest people in the subject, but with little success. But all this leaves us far short of an understanding of the sociology of this innovation - as indeed, the sociology of all new ideas is as murky and little understood as is any aspect of human life. And this means that the crystal ball in which one may read forecasts of the future adoption of this innovation is most unclear as of now. [fn.[ But the turf claims need not be honored if they impose costs upon those who would apply the discipline to their work and lives. **ENDNOTES** <1>: As a general matter, the law and the bureaucrats prevent people from learning from the best teachers in the nation, hence preventing inellectual progress and productivity gains in education. Examples: 1) In twenty-seven states, high school students may not receive credit for courses taught by television. 2) In Maryland, the state university may not beam undergraduate programs to other institutions, including junior colleges. Even tougher are the informal barriers against presenting the best teaching by the best minds through high-tech media - video cassettes, computer tutorials, and "distance learning" via tv. For example, at George Mason University, Dr. X finds that the professors will not participate in "course-sharing" -- that is, bringing in television programs from other universities. All the talk, and all the commissions, intended to improve education are a waste of time if the most important steps that can be taken to improve education are stymied by organizational and individual self-interest on the part of the education establishment. No one should be surprised at the existence of these barriers to the production of the finest education. Put yourself in the place of teachers, and it's easy enough to imagine yourself not welcoming technical changes which will reduce the demand for your services. Teachers are no different than weavers in the eighteenth century, caboose-riding brakemen in the early twentieth century, and industrial workers right now.