Research support - Environmental Science & Technology (ACS...
0 downloads
77 Views
150KB Size
ES&T LETTERS Research support Dear Sirs: In his letter (ES&T, Vol. 15, No. 11, 1981, p. 1250), Dr. Nee dleman questions Dr. Jerome Cole's views on the grounds that they were based upon studies ". . . supported by lead industry funds." Since when in the U.S. is the accused denied the right to have his evidence presented and fairly evaluated rather than summarily dis missed as biased on the mere fact that the defendant presented them? Since, however, Dr. Needleman raises this ethical question, it would seem only fair that his motives, too, be examined to determine whether, in fact, he himself is free of any bias. As a university professor, his record of research and publication is part and parcel of his claim to enhancement of his position, whether in status, research funds, salary, or whatever. I would venture that Dr. Needleman's studies had financial support for which he may have felt grateful to the extent that the results could easily have re flected that gratitude as well as an expectation of continued support. While conscious or unconscious bias is a factor in all human views, scientists can reasonably be expected to over come them at least to the extent that only supportable claims are put forth over which reasonable differences of opinion can be argued on scientific grounds. If that argument is reduced to charges equivalent to a personal lack of integrity, I can only conclude that the one raising that question might well be just as much suspect as the one he has charged. I. W. Tucker, Ph.D. Professor of Environmental Engineering (Emeritus) University of Louisville 1810 Crossgate Lane Louisville, Ky. 40222 Trace analyses Dear Sir: The article of John Glaser et al. "Trace Analyses for Wastewa ters" (ES&T, Vol. 15, N o . 12, 1981, pp. 1426-1435) details a procedure for estimation of the method detection limit ( M D L ) . It appears that this procedure would yield low estimates of the M D L . Formulation of a procedure for es timation of M D L is dependent upon 262A
Environ. Sci. Technol., Vol. 16, No. 5, 1982
interpretation of the definition of M D L . One interpretation leads to the following requirements of the statis tical test of the null hypothesis (no analyte present) using the laboratory determination of concentration: • If the analyte is not present, then the statistical test of the determination should yield the correct conclusion of absence 99% of the time. • If the analyte is present at the M D L level, then the statistical test of the determination should yield the correct conclusion of presence 99% of the time. This in effect defines the Type I and Type II errors of the statistical test (1 ) and requires consideration of two error distributions: distribution A — " t h e analyte absent" error distribution, and distribution Β—"the analyte present at the M D L mean level" error distri bution. The critical value C of the statistical test would be the 99th percentile of distribution A and the first percentile of distribution B. The coincidence of these percentiles is obtained by ap propriately locating M D L (i.e., dis tribution B). This procedure would yield an estimate of M D L that would be greater than that obtained from the authors' procedure in Equation 8, p. 1427. Reference (1) Ostle, B. "Statistics in Research"; The Iowa State University Press: Ames, Iowa, 1963; p. 109. Ronald L. Jacobson Metropolitan Waste Control Commission 350 Metro Square Building St. Paul, Minn. 55101
1983 WHO fellowships Dear Sir: The World Health Orga nization ( W H O ) will make available in 1983 a limited number of short-term fellowships for travel/study abroad related to the improvement and ex pansion of health services in the U.S. This support is limited to U.S. citizens engaged in operational or educational aspects of public health employed by state or local governmental or educa tional agencies. Employees of the federal government are not eligible to apply. A selection committee will recom mend the awarding of fellowships based on the applicant's professional background; need, objectives, and lo cale of travel; and the utilization of the experience on return to the U.S. Ap plications will not be considered for basic research or attendance at inter national meetings, nor will they be
considered from undergraduate or graduate students. Medical interns and residents are considered to be in graduate training. The fellowships award will include per diem costs and transportation. Employers of successful applicants are expected to endorse the applications and continue the applicant's salary during the fellowship period. Except in unusual circumstances, the fellowships will be limited to short-term programs of one to two months. The number of fellowships awarded will be governed by the amount of funds available. Deadline for the submission of appli cations is Sept. 30, 1982. Additional information may be ob tained from the Chairman, W H O Selection Committee, Health Re sources Administration, FCB # 2 , Rm. 9-50, 3700 East-West Highway, Hyattsville, Md. 20782; (301) 4367770.
Cost of air pollution control Dear Sir: Dr. Lodge's letter (ES&T, Vol. 15, No. 10, 1981, p. 1113) adds weight to several important points made in the closing portion of our paper (p. 755). There are some serious practical problems that deter air pollution con trol agencies from using advanced engineering tools when designing air pollution control strategies. Careful design requires both time and money—certainly more time and often more money than air pollution control agencies have been provided in the past. Can these higher control agency costs be justified? We believe that the answer is yes, as the following exam ples will show. Our review of control strategy de sign procedures centered on techniques that have been demonstrated using data from case studies in at least 11 U.S. cities. The most costly study dis cussed involved oxidant control strat egy development in the San Francisco Bay area. One of the authors of that work places its cost at about $1.4 mil lion, excluding the initial cost of the LIRAQ model development effort that was funded by N S F . Of the remaining studies cited, most were performed within academic institutions or con sulting firms. Having done several such studies we know that most could have been executed within universities on budgets of about $150 000 or less in cases where no new field measurement programs are necessary. As Dr. Lodge notes, when new field measurement (Continued on page 264A)
