3
Statistical Engineering Division
INTRODUCTION
As observed in a National Research Council report,1 the Statistical Engineering Division (SED) has nearly a 70-year history of
Consistent and fundamental contributions to the central mission of NIST through the development and application of statistical methodology for metrology. This methodology supports the basic NIST activities of producing measurements and quantifying their associated uncertainties… The SED advances its mission of supporting research in measurement science, technology, and the production of standard reference materials (SRMs), through numerous scientific collaborations within NIST and externally. The SED conducts statistical research and provides important training and educational activities within and beyond NIST. It serves as a unique national and international resource for the metrology and standards communities and more broadly in high profile contexts where an acknowledged impartial broker of data analysis and interpretation is needed.
ASSESSMENT OF TECHNICAL PROGRAMS
Accomplishments
The SED is well-aligned and effective in its support of the main NIST missions. It currently conducts a wide variety of technical efforts, including classical statistical metrology projects such as mass calibration weighing designs, testing standards for cardiac leads, and laboratory experimentation for nanotoxicology. Other work supports national and international best statistical practices in metrology and the development of new methods of measurement and uncertainty quantification for important specialized technical applications. Current examples cover a broad range of scientific projects, including greenhouse gas emissions, U.S. extreme wind speeds, shape metrology, and quantitative imaging of materials, as well as several projects related to the current NIST priority in forensic science, including analysis for low-template deoxyribonucleic acid (DNA) traces, firearms and tool marks, and general modeling and inference for problems in forensic science.
The SED technical efforts are technically challenging and are well executed. Methodology that is developed for and applied to these challenges is sound and of high quality, equivalent to that of the best U.S. national laboratories and best industrial research facilities. In light of the unique mission of NIST and the role played by the SED in NIST, the division is a one-of-a-kind national resource.
_________________________
1 National Research Council, Assessment of the National Institute of Standards and Technology Information Technology Laboratory—Fiscal Year 2011, The National Academies Press, Washington, D.C., p. 30.
Opportunities and Challenges
The SED continues to have a low profile in the national and international technical statistics community. Some existing projects with concrete internal or subject-matter deliverables could be developed further or studied methodologically on a deeper level, which could lead to their publication in the mainstream statistical literature and thereby give the SED greater visibility.
Conclusions and Recommendations
SED could more fully benefit from and contribute to current statistical research.
Recommendation: The Statistical Engineering Division should take steps to raise its profile through technical publications in statistics venues and through educational efforts to improve the awareness of statistics educators of the vital role of measurement in data collection and the role of statistics in good measurement.
PORTFOLIO OF SCIENTIFIC EXPERTISE
Accomplishments
The division’s scientific expertise in statistics for metrology continues to be among the best in the world. The division strongly supports the larger NIST organization’s programs.
Opportunities and Challenges
It appears that the demands and pressures of important NIST project work presently militate against SED’s full engagement with the statistical research community.
Conclusions and Recommendations
Attention to both metrology and the basic statistics field needs to be a division priority in order to move both areas forward and guarantee ongoing excellence in supporting NIST priorities.
Recommendation: The Statistical Engineering Division should seek stronger ties with the statistical research community in terms of both publications and human resources.
ADEQUACY OF FACILITIES, EQUIPMENT, AND HUMAN RESOURCES
Accomplishments
Previous reviews have consistently recommended increases in technical staffing for the SED. Good progress has been made since the last review in 2011: The number of technical staff has increased from 23 to 25.2
_________________________
2 National Research Council, An Assessment of the National Institute of Standards and Technology Information Technology Laboratory: Fiscal Year 2011, The National Academies Press, Washington, D.C., 2011.
Opportunities and Challenges
The number of staff continues to be a primary concern for the SED. Sixty-five percent of the technical staff (16 of 25) will become eligible for retirement in the next 5 years. Technical staffing on important projects and in vital areas is very thin (often only one person per project), making the organization quite brittle. Continued hiring is essential into the foreseeable future, at least at the pace of the last 4 years, during which four permanent employees and one guest were hired.
Conclusions and Recommendations
Progress has been made on human resource issues since the 2011 review, but important challenges remain.
Recommendation: In hiring permanent employees, the Statistical Engineering Division (SED) should focus attention on (1) ensuring corporate memory of and expertise in the fundamental ongoing National Institute of Standards and Technology standard reference material (SRM), calibration, and the experimental design work that has always been part of the SED portfolio, and on developing new capabilities in priority areas such as shape metrology and greenhouse gas monitoring, which seem to be particularly thinly staffed; (2) maintaining high overall technical competence; (3) improving staff diversity; and (4) balancing emphases on the noncore and core areas—for example, while forensics is a current laboratory-wide priority and SED is a key participant, when hiring for forensics the SED should address general expertise, so that other more thinly staffed critical areas are not neglected.
Recommendation: The Statistical Engineering Division should invigorate its program of graduate student internships to include two to four Ph.D. students per summer, and it should begin to develop a much broader and stronger program of visiting faculty researchers.
DISSEMINATION OF OUTPUTS
Accomplishments
The SED research programs are clearly driven by stakeholder needs. The division has responded effectively to government priorities such as those related to greenhouse gas emissions and forensic science. Its publications are effective in the metrological literature and for stakeholders.
Opportunities and Challenges
Strong contributions to the statistical literature improve and maintain engagement with the statistical research community and will, over time, improve the methods being brought into SED work. Such efforts come at the expense of time on other projects and therefore require division- and ITL-level support.
Conclusions and Recommendation
Publications, presentations, interaction with visiting researchers, and sponsoring internships are important means for engaging with the statistical and metrology communities.
Recommendation: The Statistical Engineering Division should monitor its yearly division total publications in the three categories—subject-matter journals, metrology journals, and statistics journals—and external research presentations, research visitor days, and internships sponsored.