DOCSLIB.ORG

Driven by Caring and Change

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Driven by Caring and Change INCE ITS FOUNDING, UNC’s school of public health S has encouraged innovators who developed new ways of collecting and analyzing information and solving problems. At the heart of that effort has been a synergistic collaboration of the School’s basic measurement science departments—biostatistics, epidemiology, and environmental sciences and engineering—along with other departments that have strong methods and measurement science foci. These include the School’s departments of health policy and management, health behavior, maternal and child health, and nutrition, as well as others outside the School. DATA DRIVEN BY CARING AND CHANGE 48 | CAROLINA PUBLIC HEALTH DATA DRIVEN BY CARING IF YOU ALWAYS DO WHAT YOU’VE ALWAYS DONE, YOU’LL ALWAYS AND GET WHAT YOU’VE ALWAYS GOT. CHANGE —HENRY FORD FALL 2014 | 49 [Above, left to right]: Drs. Jim Grizzle, Bernard Greenberg, Dale Williams and Michel Ibrahim discuss plans for renewal of their research grant on lipids; Drs. Gary Koch and Dennis Gillings conducted clinical trials work from a trailer on School grounds; Dr. Gillings, as a faculty member in 1971. DEVELOPMENT mathematical statistics department at drugs. Under Greenberg’s leadership, UNC-Chapel Hill. In 1949, less than the department developed a roadmap OF BIOSTATISTICS a decade after the School began, she for evaluating new drugs for efficacy. helped establish the Department of Without the clinical trials methodol- METHODS Biostatistics in the UNC–Chapel Hill ogy and infrastructure developed by School of Public Health, selecting her UNC’s biostatisticians, many advances former student, Bernard Greenberg, that benefit individuals might nev- In 1940, UNC system president Frank PhD, to lead it. er have been made. Members of the Porter Graham took a chance in hiring Greenberg, who later became the biostatistics faculty continue to make Gertrude Cox, MS, as the first woman department’s first Kenan Distinguished innovations in clinical trials. professor and department head at Professor, was awarded a National In 1967, NIH’s National Heart N.C. State University. Cox, professor of Institute of General Medical Sciences Institute commissioned Greenberg statistics and head of the Department grant, “Multivariate Analysis for the to develop procedures for conducting of Experimental Statistics, likely met Health Sciences,” with co-principal large, multicenter clinical trials. His Graham’s highest expectations; her investigator S.N. Roy, MSc. With Greenberg Report (see tinyurl.com/ hiring certainly set in motion a series of support from Hotelling and others Greenberg-report) is still a respected events that resulted in one of the finest in the UNC–Chapel Hill statistics document, having established methods biostatistics departments in the world. department, Greenberg also competed to evaluate new drugs and cancer As the UNC system’s graduate successfully for one of the first National treatments that have been used for program in statistics was housed at Institutes of Health (NIH) grants to nearly 50 years. N.C. State, it was incumbent upon Cox train biostatisticians and pioneered The tradition of statistical to develop any additional statistics efforts to establish the practice of innovation for the betterment of programs on other campuses. In 1946, cooperative multicenter trials. humankind continues in a program led she recruited Harold Hotelling, PhD, This groundbreaking, transforma- today by Michael Kosorok, PhD, W.R. formerly on the faculties of Stanford tional research changed the course Kenan Jr. Distinguished Professor and Columbia, to head the new of practice for the evaluation of new and chair of biostatistics, who 50 | CAROLINA PUBLIC HEALTH collaborates with colleagues at UNC, continuously funded NIH coordinating “With that,” Grizzle says, Duke and N.C. State universities, and center in the U.S., having coordinated “biostatisticians changed the world.” SAS (SAS.com), a business analytics studies with hundreds of clinical centers Williams, who succeeded Grizzle software and services company. in the U.S. and abroad. The late H.A. as principal investigator for the study, The National Cancer Institute Tyroler, MD, Alumni Distinguished says that trial “set the stage for the drug (NCI)-funded program, “Statistical Professor of epidemiology, was a vital development that led to statins and Methods for Cancer Clinical Trials,” is contributor to the study group and lipid-lowering compounds.” pioneering new methods that will help provided much insight into the tools Grizzle’s methodology for crossover researchers more effectively discover and methods used by the CSCC. Grizzle, trials is still referred to today, some 50 and evaluate cancer treatments, as well Clarence (Ed) Davis, PhD, retired years later. Another paper by Grizzle, on as treatments for other diseases, and former chair of biostatistics and director analyzing categorical data and whose to more quickly deliver effective new of the CSCC, and Dale O. Williams, co-authors included Frank Starmer, therapies to patients. PhD, took the lead in the first study, PhD, now with the Duke University- The biostatistics department was which found that lowering cholesterol National University of Singapore successful from the start in attracting reduces the risk of heart disease. Graduate Medical School, and financial support through contracts biostatistics professor Gary Koch, PhD, and grants, says Jim Grizzle, PhD, is among the most-cited biostatistics who chaired the department from papers in the world. 1972 to 1987. In 1958, Greenberg undertook In 1955, the NCI awarded the another landmark epidemiological department a contract to become the study—the first large study of statistical coordinating center for the cardiovascular disease that included Southeastern Cooperative Cancer African-Americans. All other large Chemotherapy Study Group (SCCC), trials in the U.S. at the time had only believed to be the first large-scale white participants. The Evans County, clinical trials group in the U.S. (The first Ga., study continued through the such study was published in England 1990s. (See tinyurl.com/evans-county.) in 1949.) Faculty members from 10 “All of these studies had a real medical schools and the coordinating impact on the health of society,” says center collaborated to investigate the Ed Davis. The studies led to insights effectiveness of chemotherapy agents. and treatment options for widespread The department developed some of health problems, such as high the new mathematics necessary to an- cholesterol and congestive heart failure. alyze data from these large-scale, mul- The impact continues with research ticenter trials. In 1957, Greenberg hired in progress now, including the Hispanic Grizzle as a graduate research assistant. Dr. Gary Koch, in the 1980s, stands Grizzle led the Lipid Research Clinics outside his office in ‘Trailer 39.’ Coordinating Center, established in 1971, which was among the first groups to perform distributive data entry. The unit was renamed the Collaborative Studies Coordinating Center (CSCC) in 1986, and is now the oldest, FALL 2014 | 51 Community Health Study/Study of and Drug Administration, both for- the CSCC in 1991, had cumulative Latinos, the largest, long-range study mer biostatistics professors and CSCC oversight of $35 million in grants. to date on this population. An original directors, and is currently overseen by UNC’s School of Public Health $22 million grant from the National Jianwen Cai, PhD, professor and vice extended its international reach, Heart, Lung and Blood Institute chair of biostatistics. Twelve faculty particularly in the 1960s after supported the multi-center study of members, 24 staff members and four Pranab K. Sen, PhD, DSc, Cary C. more than 16,000 Hispanic adults to students from UNC are involved in Boshamer Distinguished Professor, examine the role of acculturation in the study. joined the faculty and strengthened the prevalence of disease. A six-year, Grant funding reflected the ties to universities in India. Sen $21.7 million federal contract was biostatistics department’s success. has contributed significantly to also awarded to conduct a second Grizzle recalled that his department’s methodological studies related to examination of study participants, budget hovered close to $6 million diabetes, environmental health and beginning in October 2014. annually, but the state contributed bioinformatics, as well as clinical trials. The study was initiated by Lloyd only $90,000. The rest came from The lipids study also fostered a Chambless, PhD, now retired, and Lisa grants and contracts. Dale Williams, stream of global interactions, including LaVange, PhD, now at the U.S. Food who succeeded Grizzle as director of ones with eastern European countries, SPEEDING UP DRUG DEVELOPMENT, IMPROVING CLINICAL TRIALS Data pile up fast in large-scale, long-term, multisite clinical trials. As data collection methods evolve, so too must the means by which data are analyzed. Early on, the Gillings School’s biostatistics department developed a reputation for its program in Bayesian statistics, an analysis paradigm that involves determining probability of something based on particular pieces of evidence. The program is led by Alumni Distinguished Professor Joseph G. Ibrahim, PhD, who also directs the UNC Center for Innovative Clinical Trials (CICT) and the department’s graduate studies program. Ibrahim and colleagues in the CICT have conducted leading- edge research on innovative Bayesian clinical trials design, including development of new statistical
Load more

Recommended publications
  • April 2, 2015 Probabilistic Voting in Models of Electoral Competition By
    April 2, 2015 Probabilistic Voting in Models of Electoral Competition by Peter Coughlin Department of Economics University of Maryland College Park, MD 20742 Abstract The pioneering model of electoral competition was developed by Harold Hotelling and Anthony Downs. The model developed by Hotelling and Downs and many subsequent models in the literature about electoral competition have assumed that candidates embody policies and, if a voter is not indifferent between the policies embodied by two candidates, then the voter’s choices are fully determined by his preferences on possible polices. More specifically, those models have assumed that if a voter prefers the policies embodied by one candidate then the voter will definitely vote for that candidate. Various authors have argued that i) factors other than policy can affect a voter’s decision and ii) those other factors cause candidates to be uncertain about who a voter will vote for. These authors have modeled the candidates’ uncertainty by using a probabilistic description of the voters’ choice behavior. This paper provides a framework that is useful for discussing the model developed by Hotelling and Downs and for discussing other models of electoral competition. Using that framework, the paper discusses work that has been done on the implications of candidates being uncertain about whom the individual voters in the electorate will vote for. 1. An overview The initial step toward the development of the first model of electoral competition was taken by Hotelling (1929), who developed a model of duopolists in which each firm chooses a location for its store. Near the end of his paper, he briefly described how his duopoly model could be reinterpreted as a model of competition between two political parties.
    [Show full text]
  • Harold Hotelling 1895–1973
    NATIONAL ACADEMY OF SCIENCES HAROLD HOTELLING 1895–1973 A Biographical Memoir by K. J. ARROW AND E. L. LEHMANN Any opinions expressed in this memoir are those of the authors and do not necessarily reflect the views of the National Academy of Sciences. Biographical Memoirs, VOLUME 87 PUBLISHED 2005 BY THE NATIONAL ACADEMIES PRESS WASHINGTON, D.C. HAROLD HOTELLING September 29, 1895–December 26, 1973 BY K. J. ARROW AND E. L. LEHMANN AROLD HOTELLING WAS A man of many interests and talents. HAfter majoring in journalism at the University of Washington and obtaining his B.A in that field in 1919, he did his graduate work in mathematics at Princeton, where he received his Ph.D. in 1924 with a thesis on topology. Upon leaving Princeton, he took a position as research associate at the Food Research Institute of Stanford Univer- sity, from where he moved to the Stanford Mathematics Department as an associate professor in 1927. It was during his Stanford period that he began to focus on the two fields— statistics and economics—in which he would do his life’s work. He was one of the few Americans who in the 1920s realized the revolution that R. A. Fisher had brought about in statistics and he spent six months in 1929 at the Rothamstead (United Kingdom) agricultural research station to work with Fisher. In 1931 Hotelling accepted a professorship in the Eco- nomics Department of Columbia University. He taught a course in mathematical economics, but most of his energy during his 15 years there was spent developing the first program in the modern (Fisherian) theory of statistics.
    [Show full text]
  • An Experimental Investigation of the Hotelling Rule
    Price Formation Of Exhaustible Resources: An Experimental Investigation of the Hotelling Rule Christoph Neumann1, Energy Economics, Energy Research Center of Lower Saxony, 38640 Goslar, Germany, +49 (5321) 3816 8067, [email protected] Mathias Erlei, Institute of Management and Economics, Clausthal University of Technology, 38678 Clausthal-Zellerfeld, Germany +49 (5323) 72 7625, [email protected] Keywords Non-renewable resource, Hotelling rule, Experiment, Continuous Double Auction Overview Although Germany has an ambitious energy strategy with the aim of a low carbon economy, around 87% of the primary energy consumption was based on the fossil energies oil, coal, gas and uranium in 2012, which all belong to the category of exhaustible resources (BMWi 2013). Harold Hotelling described in 1931 the economic theory of exhaustible resources (Hotelling 1931). The Hotelling rule states that the prices of exhaustible resources - more specifically the scarcity rent - will rise at the rate of interest, and consumption will decline over time. The equilibrium implies social optimality. However, empirical analysis shows that market prices of exhaustible resources rarely follow theoretical price paths. Nevertheless, Hotelling´s rule “continues to be a central feature of models of non-renewable resource markets” (Livernois 2009). We believe that there are difficulties in the (traditional) empirical verification but this should not lead to the invalidity of the theory in general. Our research, by using experimental methods, shows that the Hotelling rule has its justification. Furthermore, we think that the logic according to the Hotelling rule is one influencing factor among others in the price formation of exhaustible resources, therefore deviations are possible in real markets.
    [Show full text]
  • IMS Presidential Address Teaching Statistics in the Age of Data Science
    IMS Presidential Address Teaching Statistics in the Age of Data Science Jon A. Wellner University of Washington, Seattle IMS Annual Meeting, & JSM, Baltimore, July 31, 2017 IMS Annual Meeting and JSM Baltimore STATISTICS and DATA SCIENCE • What has happened and is happening? B Changes in degree structures: many new MS degree programs in Data Science. B Changes in Program and Department Names; 2+ programs with the name \Statistics and Data Science" Yale and Univ Texas at Austin. B New pathways in Data Science and Machine Learning at the PhD level: UW, CMU, and ::: • Changes (needed?) in curricula / teaching? IMS Presidential Address, Baltimore, 31 July 2017 1.2 ? Full Disclosure: 1. Task from my department chair: a. Review the theory course offerings in the Ph.D. program in Statistics at the UW. b. Recommend changes in the curriculum, if needed. 2. I will be teaching Statistics 581 and 582, Advanced Statistical Theory during Fall and Winter quarters 2017-2018. What should I be teaching? ? IMS Presidential Address, Baltimore, 31 July 2017 1.3 Exciting times for Statistics and Data Science: • Increasing demand! • Challenges of \big data": B challenges for computation B challenges for theory • Changes needed in statistical education? IMS Presidential Address, Baltimore, 31 July 2017 1.4 Exciting times for Statistics and Data Science: • Increasing demand! Projections, Bureau of Labor Statistics, 2014-24: Job Description Increase % • Statisticians 34% • Mathematicians 21% • Software Developer 17% • Computer and Information Research Scientists
    [Show full text]
  • Asymptotic Theory for Canonical Correlation Analysis
    Journal of Multivariate Analysis 70, 129 (1999) Article ID jmva.1999.1810, available online at http:ÂÂwww.idealibrary.com on Asymptotic Theory for Canonical Correlation Analysis T. W. Anderson Department of Statistics, Stanford University, Stanford, California 94305-4065 Received November 30, 1998 The asymptotic distribution of the sample canonical correlations and coefficients of the canonical variates is obtained when the nonzero population canonical correlations are distinct and sampling is from the normal distribution. The asymptotic distributions are also obtained for reduced rank regression when one set of variables is treated as independent (stochastic or nonstochastic) and the other set View metadata,as dependent. citation Earlier and similar work papers is corrected. at core.ac.uk 1999 Academic Press brought to you by CORE AMS 1991 subject classifications: 62H10, 62H20. provided by Elsevier - Publisher Connector Key Words: canonical variates; reduced rank regression; maximum likelihood estimators; test of rank. 1. INTRODUCTION Hotelling (1936) proposed canonical correlations as invariant measures of relationships between two sets of variates. Suppose that the two random vectors Y and X of p and q components ( pq), respectively, have the covariance matrix 7 7 7= YY YX (1.1) \7XY 7XX+ with 7YY and 7XX nonsingular. The first canonical correlation, say \1 ,is the maximum correlation between linear combinations U=:$Y and V=#$X. The maximizing linear combinations normalized so var U= var V=1 and denoted U1=:$1 Y and V1=#$1 X are the first canonical variates. The second canonical correlation \2 is the maximum correlation between linear combinations U=:$Y and V=#$X uncorrelated with U1 and V1 .
    [Show full text]
  • Subsidies, Savings and Sustainable Technology Adoption: Field Experimental Evidence from Mozambique∗
    Subsidies, Savings and Sustainable Technology Adoption: Field Experimental Evidence from Mozambique∗ Michael R. Carter University of California, Davis, NBER, BREAD and the Giannini Foundation Rachid Laajaj Universidad de los Andes Dean Yang University of Michigan, NBER and BREAD July 6, 2016 Abstract We conducted a randomized experiment in Mozambique exploring the inter- action of technology adoption subsidies with savings interventions. Theoretically, combining technology subsidies with savings interventions could either promote technology adoption (dynamic enhancement), or reduce adoption by encouraging savings accumulation for self-insurance and other purposes (dynamic substitution). Empirically, the latter possibility dominates. Subsidy-only recipients raised fertil- izer use in the subsidized season and for two subsequent unsubsidized seasons. Consumption rose, but was accompanied by higher consumption risk. By con- trast, when paired with savings interventions, subsidy impacts on fertilizer use disappear. Instead, households accumulate bank savings, which appear to serve as self-insurance. Keywords: Savings, subsidies, technology adoption, fertilizer, risk, agriculture, Mozam- bique JEL classification: C93, D24, D91, G21, O12, O13, O16, Q12, Q14 ∗Contacts: [email protected]; [email protected]; [email protected]. Acknowlegements: An- iceto Matias and Ines Vilela provided outstanding field management. This research was conducted in collabora- tion with the International Fertilizer Development Corporation (IFDC), and in particular we thank Alexander Fernando, Robert Groot, Erik Schmidt, and Marcel Vandenberg. We appreciate the feedback we received from seminar participants at Brown, PACDEV 2015, the University of Washington, Erasmus, the Institute for the Study of Labor (IZA), Development Economics Network Berlin (DENeB), the \Improving Productivity in De- veloping Countries" conference (Goodenough College London), UC Berkeley, Stanford, and Yale.
    [Show full text]
  • Price Formation of Exhaustible Resources: an Experimental Investigation of the Hotelling Rule
    A Service of Leibniz-Informationszentrum econstor Wirtschaft Leibniz Information Centre Make Your Publications Visible. zbw for Economics Neumann, Christoph; Erlei, Mathias Working Paper Price Formation of Exhaustible Resources: An Experimental Investigation of the Hotelling Rule TUC Working Papers in Economics, No. 13 Provided in Cooperation with: Clausthal University of Technology, Department of Economics Suggested Citation: Neumann, Christoph; Erlei, Mathias (2014) : Price Formation of Exhaustible Resources: An Experimental Investigation of the Hotelling Rule, TUC Working Papers in Economics, No. 13, Technische Universität Clausthal, Abteilung für Volkswirtschaftslehre, Clausthal-Zellerfeld, http://dx.doi.org/10.21268/20161213-152658 This Version is available at: http://hdl.handle.net/10419/107454 Standard-Nutzungsbedingungen: Terms of use: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your Zwecken und zum Privatgebrauch gespeichert und kopiert werden. personal and scholarly purposes. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle You are not to copy documents for public or commercial Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich purposes, to exhibit the documents publicly, to make them machen, vertreiben oder anderweitig nutzen. publicly available on the internet, or to distribute or otherwise use the documents in public. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung
    [Show full text]
  • A Non-Normal Principal Components Model for Security Returns
    A Non-Normal Principal Components Model for Security Returns Sander Gerber Babak Javid Harry Markowitz Paul Sargen David Starer February 21, 2019 Abstract We introduce a principal components model for securities' returns. The components are non-normal, exhibiting significant skewness and kurtosis. The model can explain a large proportion of the variance of the securities' returns with only one or two components. Third and higher-order components individually contribute so little that they can be considered to be noise terms. 1 Introduction In this paper, we propose a non-normal principal component model of the stock market. We create the model from a statistical study of a broad cross-section of approximately 5,000 US equities daily for 20 years. In our analysis, we find that only a small number of components can explain a significant amount of the variance of the securities' return. Generally, third and higher-order compo- nents (essentially, the idiosyncratic terms) individually contribute so little to the variance that they can be considered to be noise terms. Importantly, we find that neither the significant components nor the idiosyncratic terms are normally distributed. Both sets exhibit significant skewness and kurtosis. Therefore, traditional models based on normal distributions are not fully descriptive of security returns. They can neither represent the extreme movements and comovements that security returns often exhibit, nor the low-level economically insignificant noise that security returns also often exhibit. However, these characteristics can be represented by the model presented in this paper. The finding of non-normality implies that meaningful security analysis requires statistical measures (1) that are insensitive to extreme moves, (2) that are also not influenced by small movements that may be noise, but (3) that still capture information in movements when such movements are meaningful.
    [Show full text]
  • Editorial Highlights Newsfront
    Editorial ¾ The 50th Anniversary of the Treaty of Rome and official statistics ......... 2 Highlights ¾ Euro-IND statistical news............................................................................ 4 ¾ Insight on: Celebrating Europe! A Statistical portrait of the European Union 2007 .................................................................................................. 10 Newsfront ¾ News from the Member States .................................................................. 11 ¾ Forthcoming events.................................................................................... 14 ¾ Cool tools and sites: Celebrating Europe! 50thAnniversary of the treaty of Rome........................................................................................................ 20 ¾ Webtrends................................................................................................... 22 ¾ Contact us.................................................................................................... 22 EUROINDICATORS Editorial ¾ The 50th Anniversary of the Treaty of Rome and official statistics The 50th anniversary of the Treaty of Rome is a good reason to look back at what official European statisticians have achieved over the last fifty years. In the early days they put together those national statistics that helped to shape specific EU policies, initially for coal and steel and later for agriculture. For the customs union they harmonised external trade statistics, and for the Common Market they evolved the basis for European
    [Show full text]
  • Economic Valuation of Cultural Heritage: Application of Travel Cost Method to the National Museum and Research Center of Altamira
    sustainability Article Economic Valuation of Cultural Heritage: Application of Travel Cost Method to the National Museum and Research Center of Altamira Saúl Torres-Ortega 1,* ID , Rubén Pérez-Álvarez 2 ID , Pedro Díaz-Simal 1 ID , Julio Manuel de Luis-Ruiz 2 and Felipe Piña-García 2 1 Environmental Hydraulics Institute, University of Cantabria, 39011 Santander, Spain; [email protected] 2 Cartographic Engineering and Mining Exploitation R+D+i Group, University of Cantabria, 39316 Torrelavega, Spain; [email protected] (R.P.-Á.); [email protected] (J.M.d.L.-R.); [email protected] (F.P.-G.) * Correspondence: [email protected]; Tel.: +34-942-201616 (ext. 1218) Received: 11 June 2018; Accepted: 19 July 2018; Published: 20 July 2018 Abstract: The economic assessment of non-marketed resources (i.e., cultural heritage) can be developed with stated or revealed preference methods. Travel cost method (TCM) is based on the demand theory and assumes that the demand for a recreational site is inversely related to the travel costs that a certain visitor must face to enjoy it. Its application requires data about the tourist’s origin. This work aims to analyze the economic value of the National Museum and Research Center of Altamira, which was created to research, conserve, and broadcast the Cave of Altamira (UNESCO World Heritage Site since 1985). It includes an accurate replica known as the “Neocave”. Two different TCM approaches have been applied to obtain the demand curve of the museum, which is a powerful tool that helps to assess past and future investments.
    [Show full text]
  • Computing Exact Power and Sample Size for Hotelling's T2-Test And
    Computing Exact Power and Sample Size for Hotelling’s T2-test and Related Multivariate Procedures Jimmy Thomas Efird, Ph.D., M.Sc.1,2 Kavitha Alimineti, B.Sc.1 1Biostatistics and Data Management Facility, John A. Burns School of Medicine, Honolulu, HI; 2Department of Biostatistics and Epidemiology, UCSF School of Medicine, San Francisco, CA ABSTRACT Multiple comparisons often arise in medical and epidemiologic research and several methods exist for adjusting statistical significance levels. In this paper, we discuss the multivariate method attributed to Harold Hotelling for handling multiplicity in the p- variate two-sample case. A generalized method is derived for computing the exact power for this test. The technique may be easily adapted to other multivariate procedures that utilize the T2-statistic. INTRODUCTION Given the case of p-variate observations from two multivariate normally-distributed populations with common covariance matrix, Hotelling’s T2-statistic may be used to test the equality of the vector of means associated with the above samples. If the null hypothesis of equal means is rejected, then simultaneous confidence intervals may be constructed to determine which components of the mean vector differ between the two groups. Below, we derive the exact power for Hotelling’s T2-test. METHODOLOGY Let d denote the vector of mean differences and S the common sample covariance matrix based on random samples of sizes n1 and n2 from two p–variate normally-distributed populations. A direct extension of the univariate pooled t-test to multivariate space (Kelsey, et al., 1996; Kramer, 1972) gives the exact power for Hotelling’s T-test as (n + n2 − 2)p = ( ) /( + )dS-1d'− 1 (1) T power n1n2 n1 n2 Fp,n1+ n2-p-1, (n1+ n2 − p −1) where is the 100 (1−α) centile of the F-distribution with p and n +n -p-1 Fp,n1+ n2-p-1 1 2 degrees of freedom (Johnson and Wichern, 1982).
    [Show full text]
  • THE EPIC STORY of MAXIMUM LIKELIHOOD 3 Error Probabilities Follow a Curve
    Statistical Science 2007, Vol. 22, No. 4, 598–620 DOI: 10.1214/07-STS249 c Institute of Mathematical Statistics, 2007 The Epic Story of Maximum Likelihood Stephen M. Stigler Abstract. At a superficial level, the idea of maximum likelihood must be prehistoric: early hunters and gatherers may not have used the words “method of maximum likelihood” to describe their choice of where and how to hunt and gather, but it is hard to believe they would have been surprised if their method had been described in those terms. It seems a simple, even unassailable idea: Who would rise to argue in favor of a method of minimum likelihood, or even mediocre likelihood? And yet the mathematical history of the topic shows this “simple idea” is really anything but simple. Joseph Louis Lagrange, Daniel Bernoulli, Leonard Euler, Pierre Simon Laplace and Carl Friedrich Gauss are only some of those who explored the topic, not always in ways we would sanction today. In this article, that history is reviewed from back well before Fisher to the time of Lucien Le Cam’s dissertation. In the process Fisher’s unpublished 1930 characterization of conditions for the consistency and efficiency of maximum likelihood estimates is presented, and the mathematical basis of his three proofs discussed. In particular, Fisher’s derivation of the information inequality is seen to be derived from his work on the analysis of variance, and his later approach via estimating functions was derived from Euler’s Relation for homogeneous functions. The reaction to Fisher’s work is reviewed, and some lessons drawn.
    [Show full text]