Goals: Where Are We and Where Should We Be Going?

Goals: Where Are We and Where Should We Be Going?Author(s): R. L. AndersonSource: Journal of the American Statistical Association, Vol. 79, No. 386 (Jun., 1984), pp. 253-258Published by: American Statistical AssociationStable URL: http://www.jstor.org/stable/2288256 .

Accessed: 15/06/2014 03:12

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

American Statistical Association is collaborating with JSTOR to digitize, preserve and extend access to Journalof the American Statistical Association.

http://www.jstor.org

This content downloaded from 185.2.32.106 on Sun, 15 Jun 2014 03:13:01 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=astata

http://www.jstor.org/stable/2288256?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Goals: Where Are We and

_Where Should We

Be Going?

R. L. ANDERSON*

1. SOME WORDS OF APPRECIATION

As a prelude, I would like to offer my profound grat- itude for the services rendered by the ASA staff, especially by Fred Leone, Ed Bisgyer, Jean Smith, Jill Stormer, Ede Denenberg, and Dorothy Zimmerman. The dedication of this staff to the ASA, its members, and the general public is outstanding. I would also like to express my heartfelt appreciation for the many amenities pro- vided by my most severe critic and constant companion, my wife, Mary.

I also wish to thank the ASA committees, and especially the chairs, for their service to the Association. I am sure that the staff and officers realize that the ASA could not function without them. We have adopted a new constitution developed by an industrious committee under the leadership of Kathleen Lamborn; the ASA- NCTM Committee under Dick Schaeffer is developing precollege teaching materials in statistics; the Future Goals Committee under Joan Rosenblatt is struggling to come up with a program that is both desirable and salable; Tom Jabine's Committee on Scientific Freedom and Human Rights has persuaded the board to take some strong action and continues to push for more definitive procedures.

One of the chief activities during the past two years has been the excellent ASA Management Review conducted by John Corson of the National Executive Service Corps. Mr. Corson's diligent, penetrating, and judicious report covered four areas: improving the functioning of the Board of Directors, reassessing ASA's programs, im-

* R. L. Anderson is Assistant for Statistical Services to the Dean of the College of Agriculture, University of Kentucky, Lexington, KY 40546. This article was presented as the Presidential Address at the 1983 Annual Meetings of the American Statistical Association in Toronto.

proving the functioning of the headquarters staff, and improving the structure and outreach of the association. The report and the board's response were published in the November 1983 The American Statistician.

An ASA Building and Development Committee, chaired by Ralph Bradley and Margaret Martin, is developing procedures to solicit donations to a Building and Development Fund. The generosity of many of you has resulted in pledges of over $250,000 to date. A consulting firm and a full-time coordinator have been employed to develop a "case statement" on the rationale for developing ASA programs and the need for a new building to support these programs. The committee has now shifted its operation into high gear with Fred Leone assuming a much more active role.

2. THE AMERICAN STATISTICAL ASSOCIATION

I now address my title: "Goals: Where Are We and Where Should We Be Going?" The "we" is both the ASA and the statistics profession. Although these two are in- trinsically intertwined, I will attempt to indicate the "wheres" for each. At the present time, the ASA has a membership of about 15,000; unfortunately, there has been little change in this total in the last four years. Stren- uous efforts are being made to find out why old members are leaving and why many statisticians have never joined.

To date we have 62 chapters, which should form the backbone of the association; unfortunately, most of them feel they have little impact on our policies, decisions, or operations. Efforts are under way to solicit and utilize recommendations from them. On the advice of Mr. Cor-

? Journal of the American Statistical Association June 1984, Volume 79, Number 386

Presidential Address

253



254 Journal of the American Statistical Association, June 1984

son, Fred Leone and I have initiated periodic visits to chapters; I hope that such efforts will be continued.

Another essential role in ASA performance is played by the sections. When I joined the ASA in 1942, it had one section: Biometrics. We now have 10 sections and subsections to cover specific fields of statistical methods, theory, or applications. An ASA member can choose to belong to any number of these-from 0 to all 10. A sample from the 1978 Directory of Statisticians indicated that about 30% of the ASA members belong to four or more sections or subsections. In order to ascertain from section members their opinions regarding ASA activities, it would be useful to know which sections a member con- siders to be of prime importance. I have asked the Com- mittee on Sections to find out from you ASA members whether it would be desirable and feasible to list the one or two sections that you consider to be of prime interest; then the business of the section might be transacted by only those members who designated that section as prime.

The 1978 directory is a list of members of the ASA, the Institute of Mathematical Statistics (IMS), and the two North American regions of the Biometric Society (ENAR and WNAR). I am disappointed and chagrined that the IMS has decided to issue a separate directory in 1984. I was amazed to find that approximately one-third of the American members of IMS were not members of the ASA in 1978. I am pleased that the IMS joined us in Toronto, but I feel that the two groups do not meet to- gether often enough. This apparently ever-widening gap between the applied and theoretical statisticians is of great concern to me. What can we do to narrow this gap? For example, should we have an ASA Section on The- oretical Statistics? Perhaps we should have an ASA Sec- tion on Applied Probability, which could serve as the focus for the activities of another group of applied mathematicians who are so important in the development of the statistics profession.

I am hoping that the Committee on Committees can develop a systematic procedure for deciding what continuing committees are needed and the distinct role that each committee should play in ASA operations. Two other problems with regard to the committees are

1. How can we evaluate their performance? 2. How can the President-Elect, who fills vacancies on

committees, be apprised of who would be willing and able to serve-especially the younger ASA members?

On the second question, perhaps departmental chairmen and division leaders in industry and government should be making recommendations.

As the ASA increases its number of journals, how do we best serve the membership in its journal selection process? An ad hoc ASA committee under Bob Hogg is now struggling with this so-called "unbundling" problem.

There have been complaints that the proportion of theoretical articles in our journals is increasing and many industrial statisticians state that the ASA is not catering

sufficiently to their needs. Are these impressions associated with the plateauing of the membership count? The direction we take on these matters can have a profound influence on our profession. Your advice is sincerely so- licited.

An important aspect of our profession in the future will be continuing education. A tremendous variety of short courses, workshops, tutorials, and films was presented at the Toronto meeting. There have been protests concerning overlap with the regularly scheduled sessions. Should we curtail this overlap? With the rapid changes in science and technology, there will be an ever-increasing need for retraining of research personnel. It would seem desirable to have continuing education sessions at the regional meetings and there is strong support to have more frequent regional meetings. It should be apparent that increased activity by the continuing education group will entail a need for increased space and personnel.

A most perplexing problem facing the U.S. statistical community at this time is the reduction of federal budget allocations and professional staff for statistical operations. The ASA is participating in the activities of two organized groups that are leading the opposition to these cuts:

1. COSSA, the Consortium of Social Science Asso- ciations, has an executive director who issues timely legislative reports on fiscal, legislative, executive, and other relevant federal operations as they affect the social sciences.

2. COPAFS, a Council of 12 Professional Associations on Federal Statistics, is now on the last year of a three- year pilot program with a full-time director. It reviews and monitors federal statistical programs, reports to its members, issues documents and letters of concern, and arranges for presentations to congressmen on specific action. The ASA Board of Directors has approved contin- uation of this program.

The ASA appreciates the support of the American As- sociation for, the Advancement of Science (AAAS)-at its 1983 Annual Meeting in Detroit-of a resolution that presented the need for the maintenance and improvement of federal statistical programs and concluded with the resolve "that the AAAS Council call upon the Administra- tion and Congress to recognize the centrality of quality statistical information systems and data bases of strategic national importance concerning the state of the economy and the state of technology, and to insure the existence of a viable, authoritative and independent unit to fulfill the needs for coordinating federal statistics and for maintaining their quality and integrity." It is essential that we continue and even expand cooperative efforts with the AAAS at all levels of statistical endeavor.

There has been considerable interaction with the U.S. Congress and the Executive Office of Management and Budget by individual members of the ASA board and other members. Senator Percy (of Illinois) convened a Conference on Federal Statistics and National Needs on



Anderson: Presidential Address 255

January 25, 1983, followed by a workshop on the "Data Needs of America in Transition" (with emphasis on the 1990 census) the following two days.

The ASA Committee on Law and Justice Statistics has just been awarded a grant to provide technical assistance to the Bureau of Justice Statistics. The ASA participates with the Committee on National Statistics of the National Academy of Sciences and a number of other professional organizations to sponsor a biennial Symposium on Sta- tistics and the Environment.

One of the most active committees in this respect has been the ASA Committee on Statistics and the Environ- ment, which has been working with the Environmental Protection Agency (EPA) to improve the statistical aspects of its program. A two-day short course, "Time Se- ries," was offered in 1981 and another, "Quality Assur- ance," in 1982. The Office of Radiation Programs supported five-day symposia in 1981 and 1982 entitled "Low Order Radiation and Environmental Sampling" and "Analysis of Sampling Data: Assessment of Human Exposures and Health Effects." One was held in July 1983, entitled "Radiation and Health," with financial help from the Occupational Safety and Health Administration (OSHA).

3. THE STATISTICS PROFESSION

But what about the statistics profession? I feel that the time has come to decide whether we should have some form of professional certification and/or a code of professional ethics.

An excellent paper, "A Professional Code for Statis- ticians? Some Ethical and Technical Conflicts," by Roger Jowell (1981) is followed by extensive discussion. Jowell argued for an educational code covering statistical practice and relationships with respondents, funders, peers, and society at large.

An ad hoc ASA Committee on Professional Ethics, which has been struggling with this problem since 1977, submitted a report containing a set of such guidelines in the February 1983 The American Statistician. Many statisticians strongly supported these guidelines; however, there were a number of caveats. Since I tend to oppose attempts to legalize morality, I was impressed by Herb Solomon's comments (1983):

There are already enough constraints-economic, social and legal-that operate to keep us responsible in our efforts. The abuses of such systems (of Ethical Guidelines) would far outweigh any gains. . . . I think the ASA should keep on doing what it does well: publish journals, organize meetings, develop chapters, develop educational activities, and render statistical advice to government agencies. The notion of having a court available to resolve matters of individual behavior in statistics is anathema to me.

On the other hand, I strongly support some form of certification that an individual has the necessary training to undertake the statistical activities for which he or she is employed. How much longer are we to tolerate the minuscule requirements for statisticians in the state and

federal civil services? The requirement for a statistician in the Kentucy Civil Service is a bachelor's degree with one course in statistics, except that the college education can be replaced by experience in gathering and analyzing data on a year-for-year basis. A senior statistician has the same requirement plus only two years of experience, whereas a Statistical Supervisor requires only four ad- ditional years of experience. I presume that these requirements are typical of many states.

For a GS-5 Federal Statistician, the requirements are only slightly higher: a bachelor's degree with two three- semester-hour courses in statistics and three more courses in mathematics or statistics; however, courses can count as statistics courses if as little as 50% of the course material is statistical methods. Experience can substitute for the other course work. Advance in grade is based on professional experience or graduate course work. To be designated as a Mathematical Statistician the candidate need have only two courses in statistics, four in mathematics, and two more in statistics or mathematics.

I will discuss two impediments to our efforts to have statistics recognized as a unique scientific discipline. The first is the public's perception of a statistician as only a number cruncher-a compiler of baseball, football, bas- ketball, track, and other sports records, Dow Jones stock averages, and the like. This was emphasized in two recent Peanuts cartoons in which Linus presented the horren- dous statistics for their baseball team and remarked that "Statistics don't lie." To which Charlie Brown replied, "No, but they sure shoot off their mouth a lot." Then Charlie asked what they should do next season, to which the unanimous reply was: "Get a new statistician."

When Janet Norwood presents the latest Consumer Price Index (CPI) on television, few members of the au- dience are aware of the complex statistical tasks involved in producing it-starting with the construction of the index formula, then data collection, tabulation, and analysis of results. Somehow we need to acquaint the public with the crucial role that statisticians play in this operation. ASA's Continuing Education Department is developing media presentations that will focus on this problem; they have been too long in coming.

A second impediment to our attempts to establish that we are a unique scientific discipline is the fact that statisticians are not identified in the Statistical Abstract of the United States (Bureau of the Census 1981); instead, they are included with the general classification of "Mathematicians or Mathematical Scientists." The Na- tional Science Foundation (NSF) is cited as the source for the abstract's material on characteristics of scientists and engineers, including employment, salaries, and doc- torates conferred. This problem was highlighted in some advice we received on information needed in submitting a proposal to the NSF to provide funds to initiate statistical training in the elementary and secondary schools; the main obstacle is the present loaded math curriculum. An article in the July 4, 1983 issue of Business Week is



256 Joumal of the American Statistical Association, June 1984

entitled "Industry's Hot New Find: The Mathemati- cian"; however, the examples indicate that most of the problems were resolved by statisticians.

In preparing statistical materials for precollege students and the general public, we should minimize our mathematical requirements by emphasizing our role in the assessment of chance and variability and their place in scientific, industrial, and governmental affairs. In order to accomplish this, do we need a National Society of Professional Statisticians? There is one for engineers and for many other disciplines.

Another problem of immediate concern is the difficulty of maintaining viable statistics programs in many of our universities. One is the matter of statistics faculties. In many universities it is very difficult to obtain promotions of faculty members for whom a substantial portion of time is devoted to consultation with graduate students aAnd other faculty members. As a result, much of their research output is that of a junior author (or simply an ac- knowledgment) on articles for which statistics is a tool rather than the raison d'etre. One of my sidelines these days is to write letters recommending the promotion of these applied statisticians; however, I should emphasize that this is no invitation for a sudden deluge of requests from department chairmen. If statistics is to remain a viable profession, we must recognize and continue to prod university administrators and other academicians to recognize that statistics is an integral part of the scientific method. The profession cannot afford to break away from the traditions established by such eminent leaders as Cochran, Cox, Fisher, Hartley, Neyman, and Snedecor, for whom statistics was essentially the servant of the sciences.

In June, when I returned to Iowa State to join in the celebration of the 50th anniversary of its statistical laboratory, I reflected on the tremendous contribution of its founder, George Snedecor, to statistics in this country and the world. In May I revisited my old stamping ground at North Carolina State, where six of the old guard have just retired, and pondered on the fact that most of us went there primarily as consultants. I beg our successors to continue that close connection between statistics and the real world of the natural and social sciences and technology.

Another and perhaps even more serious problem is the maintenance of viable graduate student programs. It is becoming more and more difficult to persuade local students to pursue graduate work in statistics, primarily for two reasons: a sharp decline in the number of undergraduate majors in mathematics and the burgeoning demand for graduates in computer science. There is a fur- ther problem of persuading students to continue to the Ph.D. after completing the M.S.; many industrial and federal salaries for M.S. graduates are almost as large as for Ph.D.'s.

In order to meet the demand for M.S. and Ph.D. graduates in statistics, several courses of action should be

considered:

1. Increased emphasis on developing undergraduate programs in statistics or the umbrella of mathematical sciences.

2. Recruit undergraduate majors in the physical and engineering sciences to pursue graduate work in statistics; although these students will have a good undergraduate mathematical background, it may be necessary to modify our advanced mathematics requirements in order to persuade them to come over to statistics.

3. Recruit undergraduate majors in the biological and social sciences; their possible deficiencies in even undergraduate mathematics would almost certainly require modification in our graduate requirements.

4. Develop more flexible course requirements so that statistics majors can select a variety of statistics courses and have time to elect courses in operations research, computer science, and the natural and social sciences.

5. Develop a consulting program that involves the graduate student.

6. Encourage statistical research projects that are devoted to the solution of real world problems.

The refusal of some mathematical statisticians to be concerned about the ultimate usefulness of their research and their disparagement of applied research reminds me of the old fable about the man who devoted three years learning to kill dragons and the rest of his life trying to find dragons to kill.

We must recognize that most of our graduates will become consultants in business and government and that our graduate programs should be structured to meet this need. Some statements by P. C. Mohalanobis (1965) are germane to this discussion.

Statistics must have a purpose. . . . The time has come to intro- duce educational programmes appropriate for statistics as a fully developed new technology which calls for the utilization of a wide range of scientific knowledge to help in solving scientific or prac- tical problems. As Fisher has pointed out, "a professional statistician, as a technologist, must talk the language of both theo- retician and practitioner." The education of a statistician, like that of other technologists, must have a broad base.

4. SOME IMPORTANT PROBLEMS

I close my remarks with a discussion of four important societal problems on which statisticians should be focus- ing increased attention.

4.1 Contamination of the Environment

Ever since I became a member of ASA's Statistics and Environment Committee, it has been apparent that the statistical profession should become much more involved in environmental problems. One of the major difficulties with our volunteer efforts is the lack of time to work on the really complicated problems such as the Love Canal and dioxin contaminations. Perhaps the best we can do in such cases is to recommend outside panel members or



Anderson: Presidential Address 257

individual consultants. Whether the ASA should do this for private for-profit concerns is another matter; the Board of Directors refused to allow us to become involved with the selection and monitoring of a panel to investigate the relationship between cotton dust and brown lung disease (byssinosis) for the American Textile Manufacturers Institute, a decision that I opposed. I have composed three Forum articles on this subject in Amstat News (Anderson 1983). In these I discussed the need for a substantial increase in statistical personnel in EPA and the deficiencies of its Love Canal investigation.

Some of the major problems in the investigation of the impact of pollutants (mainly chemical) on human health are as follows:

1. The minimal detection limits using existing meas- uring instruments are too large to estimate the cumulative effects of many chemical compounds.

2. The laboratory-to-laboratory variability in assessed contamination levels often is quite large.

3. Information on long-term health effects is usually nonexistent.

4. Numbers of animals required to detect response to low doses of a relatively low toxicity substance may be so large that experiments must be conducted at very high doses. How do you extrapolate the results based on these high doses to the low doses that occur in practice? Should we force more experiments to be conducted at low doses even if they are very expensive?

5. Can information obtained from animal experiments be used to make inferences on human subjects?

6. Is the incidence of cancer, birth defects, and other health-related ailments more frequent in contaminated areas than in noncontaminated areas?

A classical controversy is arising over the danger to people for dioxin contamination. P. H. Abelson, editor of Science, summarizes a number of dioxin incidents (1983):

Some of the accidents, notably one in 1949, occurred long enough ago that were cancer to be associated with them, it would now be evident. Some 121 workers were involved, and while iron- clad proof of a null effect is missing, so too is a basis for believing that TCDD (the most notorious of the dioxins) is a dangerous carcinogen in humans.... TCDD is highly toxic when adminis- tered orally but when bound to soil it does not pose much of a hazard. With time, natural degradation of the chemical will atten- uate its presence in the environment.

However, he admonishes that TCDD is all risk and no benefit; hence, we should exert every effort to curtail its unwanted production.

Another current problem concerns acid rain. Excellent reviews are presented in Science by Marshall (1983) and Kerr (1983) and by Beck, Hagen, et al. (1983). The ASA will be assisting the EPA in a multimillion dollar nation- wide research project on acid rain, the National Acid Pre- cipitation Assessment Program.

Recently, the executive director of the National Well

Water Association stated that the threat of our water supply is not the crisis headlines suggest (Lehr 1983). He believes that " we can eliminate new sources of pollution and put an end to ill-conceived toxic dumps within 10 years... . the odds are in our favor to end up with 98 percent of our ground water unpolluted at the end of this century." But this probably is not very comforting to many residents of New Jersey, which has 6 of the 20 dirtiest dumps as identified by the EPA in December 1982 (Beck, Junkin, et al. 1983).

I hope that everyone saw Bill Moyers's frightening 26 July 1983 telecast in which he disclosed the role of organized crime in the toxic waste disposal racket, with special emphasis on the New Jersey dumps.

Abelson's editorial on waste management (1983) con- cludes with this remark: "The hazards must and will be minimized, but there will be few quick fixes. Drastic action may be required at a few sites, but more harm is likely to come from hysteria than from chemicals." The same conclusion probably could be made concerning the Three Mile Island nuclear accident.

But what can we statisticians do to help separate fact and fantasy in this area? I do not have the answer, but somehow we must devise a more expeditious procedure to become involved.

4.2 Quality Assurances in Industry and Government

That the United States is in a struggle for its eminence as the leading industrial nation is now well known. One of the most publicized aspects of that struggle is the attempt to restore quality of product as essential to our success.

Speaking of American industry, I am reminded of the fictitious American auto dealer who went into bank- ruptcy-his wife divorced him and he became an alco- holic. As he stumbled in an alley one night, a genie popped out of his wine bottle and offered him one wish. The man said he wanted to be a foreign car dealer in a major city. When he awoke, he was a Chrysler dealer in Tokyo.

It is one of those historical anomalies that we who developed the use of statistical quality control (SQC) as an important part of the World War II effort should suffer from its successful adoption by one of the defeated na- tions in that war. American statisticians have been the leaders in advising the Japanese on these procedures and are once again being welcomed by American industry. Perhaps the ASA should renew its former close ties with the American Society for Quality Control (ASQC) in order to present a united front in this effort. We are working with representatives from industry, government, and universities in developing a National Award for Excel- lence in Quality Control and Productivity. I am hoping that such an award might go to federal agencies as well as to private industries. But this is only the beginning.



258 Joumal of the American Statistical Association, June 1984

Some of the other needed developments are as follows:

1. Promote the development of statistical quality control courses in our statistics departments and engineering and business colleges.

2. Develop SQC short courses (and, if possible, longer courses) for government agencies.

3. Although many of the major U.S. industries are now employing SQC experts, there may be a need to ascertain whether we should provide assistance to the smaller concerns.

4. Provide chapters with speakers on SQC. 5. Support the ASA Section on Physical and Engi-

neering Sciences with funds to develop forums and sem- inars on SQC and related topics. A typical example was a recent seminar, "Statistics, Product Design, and Qual- ity Control" held at Madison, Wisconsin.

4.3 Federal Statistics

As I indicated above, one of the tragedies of the recent past has been the diminution of the federal statistical sys- tem, especially at top administrative levels. Of perhaps even more importance has been the drastic reduction in force (better known as RIFing) in such agencies as the Energy Information Administration and the Census Bu- reau. The statistical profession must expedite procedures to diminish the chances in 1990 of a repetition of the legal tangles involving the 1980 census. An excellent review of those legal problems is presented by Barabba, Mason, and Mitroff (1983). We should urge the Census Bureau to update its quality control (QC) procedures, especially at the grass roots of local organization and data collection and compilation. A good example of what can be accom- plished is the QC program at the Bureau of Labor Sta- tistics.

One of our tasks should be to develop a mechanism to monitor the official statistics and pronouncements issued by our federal agencies on environmental pollution, un- employment and poverty, food stamps, aid to parents of dependent children, crime rates, Social Security, gross national product (GNP), producer and consumer price indices (PPI, CPI), money supply, and business and farm foreclosures, to name a few.

I commend to your attention the strong remarks of Bon- nen, Fienberg, and Slater in the August 1983 issue of The American Statistician and urge statisticians to unite in an effort to persuade the U.S. Congress to create a Central Statistical Agency in the Executive Office of the Presi- dent. In the meantime, we should support the efforts of Congress, COPAFS, COSSA, and our continuing education group to restore and enhance the prestige of statistical divisions in the federal agencies.

4.4 National Defense, Military Operations, and Human Rights

Where were the statisticians when the Vietnam War statistics were being manufactured and circulated? What are we doing about the statistics coming out of Central

America? What about the statistics on comparative military strengths? The ASA Committee on Scientific Free- dom and Human Rights has been concerned with the controversy over the monitoring of progress in human rights in El Salvador and has proposed that some ASA committee should review the statistical methods and evalu- ation of data sources used in monitoring freedom and human rights. Many of us believe that this is our respon- sibility, but others feel that it would be setting a dangerous precedent. What do you think?

5. CONCLUSION

Recently I watched a rerun of the telecast of the last episode of Carl Sagan's excellent Cosmos. He reminded us that history is replete with the destruction of scientific and artistic achievements and even civilizations because of fear, superstition, and the lust for power. I am sure that we all join him in the plea that this should not happen again, because this could be Armageddon. The future can be an exciting time for statisticians. In the words of Jo- hannes Kepler as quoted by Sagan, "The diversity of the phenomena of nature is so great and the treasures hidden in the heavens so rich that the human mind will never be lacking for fresh nourishment."

REFERENCES ABELSON, P.H. (1983a), "Chlorinated Dioxins," Science, 220, 1337.

(1983b), "Waste Management," Science, 220, 1103. ANDERSON, R.L. (1983), "Statistics and the Environment," Amstat

News, 94, 4-6, 98, 3-4, 99, 2-3. BARABBA, V.P., MASON, R.O., and MITROFF, 1.I. (1983), "Federal

Statistics in a Complex Environment: The Case of the 1980 Census," The American Statistician, 37, 203-212.

BECK, M., HAGEN, M., ACHIRON, M., and BRENNERS-STUL- BERG, L. (1983), "The Bitter Politics of Acid Rain," Newsweek, April 25, 36-37.

BECK, M., JUNKIN, D., TAYLOR, J., MANDEL, M., AGREST, S., and ACHIRON, M. (1983), "The Toxic Waste Crisis," Newsweek, March 7, 20-24.

BONNEN, J.T. (1983), "Federal Statistical Coordination Today: A Dis- aster or a Disgrace?" The American Statistician, 37, 179-192; Reply to Comments, The American Statistician, 37, 199-202.

BUREAU OF THE CENSUS (1981), Statistical Abstract of the United States (102nd ed.), Washington, D.C.: Author, 164, 603, 604.

FIENBERG, S.E. (1983), Comment on "Federal Statistical Coordi- nation Today: A Disaster or a Disgrace?" The American Statistician, 37, 192-196.

"Industry's Hot New Find: The Mathematician" (1983), Business Week, July 4, Research Section.

JOWELL, ROGER (1981), "A Professional Code for Statisticians? Some Ethical and Technical Conflicts," in Proceedings of the 43rd Session, International Statistical Institute, 1, 167-209.

KERR, R.A. (1983), "Emission Control Will Control Acid Rain," Sci- ence, 221, 254.

LEHR, J.H. (1983), "Threat to Our Water Supply Not Crisis Headlines Suggest," Herald-Leader (Lexington, Ky.), June 23.

MAHALANOBIS, P.C. (1965), "Statistics as a Key Technology," The American Statistician, 19, 43-46.

MARSHALL, E. (1983a), "Acid Rain Researchers Issue Joint Report," Science, 220, 1359.

(1983b), "Acid Rain, A Year Later," Science, 221, 241-242. SLATER, C. (1983), Comment on "Federal Statistical Coordination

Today: A Disaster or a Disgrace?" The American Statistician, 37, 197-198.

SOLOMON, HERBERT (1983), Comment on "Ethical Guidelines for Statistical Practice: Report of the Ad Hoc Committee on Professional Ethics," The American Statistician, 37, 15.



Documents

Goals: Where Are We and Where Should We Be Going?