The Development and Enlightenment of the Department of Statistics of Columbia University

Founded in 1754, Columbia University (hereafter referred to as Columbia) is a strong and well-known figure, ranked third in the 2019 US News & World Report’s list of the best universities in the United States. Located in the heart of New York City, Columbia is a private research university and one of the Ivy League schools. Columbia is the annual Pulitzer Prize-winning institution. It has trained 101 Nobel Prize winners. Its alumni include 5 founding fathers, 9 US Supreme Court judges, and 29 heads of state (including 3 US presidents) [1]. On his official website, Columbia describes its mission as “God hopes that every department of the university will promote knowledge growth and learning at the highest level and spread the results to the world.” [2]

I. Department of Statistics and its contribution

The Department of Statistics at Columbia University (hereinafter referred to as the Department of Statistics) is an outstanding and dynamic organization and is in rapid development. The department currently has 30 faculty members, 50 doctoral students and more than 300 master students [3]. In addition to providing undergraduate majors in statistics, the department offers six master’s programs (including a master’s degree in mathematics finance with the Department of Mathematics) and a PhD program.

Since its establishment in 1946, the Department of Statistics has produced many talents who have made breakthrough contributions in the field of disciplines, which has greatly promoted the development of the discipline and produced tremendous social impact. The department’s contributions include the sequential analysis and statistical decision theory of Columbia University teacher Abraham Wald. His sequential analysis developed into an important branch of mathematical statistics after World War II, and its statistical decision theory also caused the innovation of post-war mathematical statistics. He views the mathematical statistics problem from the perspective of the game and defines the risk function of the statistical inference program to judge the quality of the program [4]. Howard Raiffa has made a breakthrough in statistical decision theory. An article co-authored by Burt Singer and others in econometrics has been cited more than 2,000 times. In addition to its theoretical contributions, the department has contributed to the creation and development of other university statistical departments in the United States. Among the early graduates of the Department of Statistics before 1970, Albert Bowker, Abraham Girshick, and Charles Stein established the Department of Statistics at Stanford University. It plays an important role in the development process. In particular, when Gershick was a Ph.D. student at Columbia University in 1948, he became the founding chair of the Stanford Department of Statistics. He later served as the president of the City University of New York and the University of California at Berkeley. [5]

Second, the creation and development

(1) Statistics and teaching activities before the establishment of the department

As early as the establishment of the Department of Statistics, Columbia had statistical teaching and research activities, mainly from other disciplines, especially the political science teachers are engaged in these activities [6]. The turning point occurred in 1931, the arrival of Harold Hotelling, which led to the world’s leading statistical department. He has contributed a lot to statistics and is the founder of many statistical methods. His most important contribution is multivariate analysis, including principal component analysis methods, and typical correlation analysis methods. [7]

In 1938, Hotling helped Wald receive funding from the Carnegie Corporation, who immediately went to Columbia to study modern statistics and become an ordinary faculty member of the department. In addition to scientific research, they also provided statistical teaching guidance for other universities in the United States, and prompted Columbia to establish a joint steering committee for the development of doctoral students in mathematical statistics in 1942 [8]. At this point, the conditions for establishing a new department have matured.

Shortly after the United States announced its participation in World War II, in order to meet the demand for statistical knowledge in war affairs, Columbia established a statistical research group and was funded by the National Defense Research Council. The research team has been in existence for 39 months. Under the supervision of W. Allen Wallis, there are 18 chief statisticians working here. Most of them are visiting scholars from other universities. . Wallis once commented: “In terms of quantity and quality, this should be the most outstanding team of statisticians ever organized.” This statistical research team played an important role in the statistical history of the United States. Members have made a number of important contributions to modern statistics, the most important of which is sequence analysis. At the same time, the experience of working in the statistical research team has made many statisticians develop in various fields, such as 8 of them become the chairman of the Mathematical Statistics Association, 4 became the chairman of the American Statistical Association, and 2 were Nobel. Two economics awards have become the president of the university. [9]

(2) The establishment of the Department of Statistics and the personnel crisis

After the end of World War II, statistical research and guidance continued to continue under the efforts of Hotlin and Wald. In 1946, the Department of Statistics of Columbia was formally established and was established in the School of Political Science. It was named the Department of Mathematical Statistics at the time, intended to show that it is a discipline that builds new methodologies based on complex mathematics [10]. But as more and more American universities established the Department of Statistics, it changed to the Department of Statistics in 1982.

The Department of Statistics has three founding members, namely Wald, Jacob Wolfowitz, and Theodore W. Anderson. In the three or four years after the establishment of the department, some employees joined, such as Joseph Leo Doob, Henry Scheffé, Howard Levene, etc. . In 1949, this small but dynamic department gradually formed. The faculty produced a series of important Xi Ming Na 1 papers. At the same time, teachers from many other disciplines were invited to teach and received many visiting scholars; a series of classic textbooks in this field were published, such as Joseph Leo Dub’s Random Process, Anderson’s Multivariate Statistical Analysis. Henry Schiffer’s “Analysis of Variance” and so on. [11]

The 1950s was an era of rapid development of statistics. A large number of first-rate universities established statistical departments or other statistical institutions, and teachers flowed frequently. In November 1950, Wald traveled to India to visit universities and research centers, and unfortunately died on December 3 in an air crash. Wolfowitz left Columbia University in 1951 to work at Cornell University. Xue Feifa went to the University of California at Berkeley in 1953. Fortunately, the Department of Statistics successfully recruited Howard Refa, Aryeh Dvoretzky, and Herbert Robbins (Herbert, formerly at the Institute of Advanced Studies at the University of North Carolina at Chapel Hill). Robbins). In 1953, the talent team of the Department of Statistics was relatively stable, and more than 10 assistant professors were recruited. Two people’s mobility crises broke out between 1967-1968 and 1980. Anderson went to Stanford University in 1967. Levin and Robbins retired in 1984 and 1985 respectively. In 1985 and 1986, two teachers went to Yale and Stanford respectively. However, during this period, the Department of Statistics was assisted by alumni, former employees, and other agencies and industry personnel, and maintained high quality academic output during the period between the two crises.

On February 27, 1992, Levin donated $1 million to the Chair of the Statistical Funding Professor. His wish is that this donation can “ensure that this small but important department survives.” Columbia officially announced the name of the chair by Levin to thank him for his contribution to the teaching research in Columbia. After visiting the Department of Statistics in 1991-1992, Paul Meier officially joined the department in 1992 and became the chairman. He was also the first recipient of the “Professor Howard Levine” chair. . Heyde CC also joined at the same time, and his arrival added new power to the application of probability theory. Since then, the Department of Statistics has become stable and has grown. [12]

(3) Further development after the stability of the Department of Statistics

The Department of Statistics was established as a postgraduate training unit. Since 1950, it has taught elementary statistics courses for students of the School of General Studies. By the mid-1960s, undergraduate teaching has become the primary function of the department, and as of 1966, undergraduate courses have been offered to as many as 330 students. Although undergraduate statistics was officially established as a major in 1982, the discussion on the inclusion of statistics in majors began several years ago. In order to meet the diverse needs of undergraduate students, the department has established minor programs and joint training majors (including economic statistics, political statistics, etc.). The department’s statistical majors have become an increasingly popular choice for college students. In 2012, more than 100 students chose major or minor programs in the department. Master’s and doctoral programs have also been expanded in recent years. At present, there are more than 1,000 students participating in the graduate program. The Department of Statistics is now recruiting teachers with the School of Political Science and the School of Sociology, and is expected to continue to recruit for this nature. The department has successfully consolidated its position as a key faculty at Columbia University, and with the evolution of statistics, it is expected to continue to make significant contributions to the front line of research. [13]

Third, the revelation

The Department of Statistics was formally established in 1946, but statistical teaching and research work has been carried out long before this. If it is calculated from the arrival of Hotlin in 1931, it has been nearly 90 years since. From the germination, growth, crisis, stability and further development, the Department of Statistics has experienced a process of ups and downs. This process can resonate with the development of many other research institutions and academic departments. Analysis of it can provide some inspiration for China’s “double first-class” construction.

(1) Establishing outstanding goals

The initial goal of the Department of Statistics was to do both first-rate statistical research and provide students with statistical education ahead of other US universities. Today, the goal of the Department of Statistics remains “to make a significant contribution to the front lines of statistical research.”

It can be said that the biggest goal and mission throughout the development phase of the Columbia Department of Statistics is to make the best research and provide the best education. This echoes Columbia’s own mission.

In the process of “double first-class” construction, especially in the construction of first-class disciplines, teachers and students of Chinese universities must set outstanding goals. Take the law, only get it; take the law, only the next. Only by setting a goal of excellence can we go to the next level and make corresponding efforts to achieve corresponding achievements.

(2) Allowing teachers to move moderately

The main reason for the several crises experienced by the Department of Statistics was the departure of the instructors. Frequent changes in personnel can lead to instability of the faculty, which in turn affects the work and development of the department. However, if the personnel structure is fixed for a long time, it is difficult to stimulate the team’s vitality. Moderate mobility of personnel can bring new ideas, new systems, and even play the “squid effect”, driving the rapid development of the entire organization.

Not only that, the moderate flow of people will prompt the university to provide a better working environment, conditions and systems for the instructors. The reason why Columbia was officially established in 1946 was an important reason. That year, Hotlin was ready to help the University of North Carolina to establish a statistical department and invited Wald to peer. Columbia worried that at the same time losing two famous scholars in the fields of economics and statistics, he decided to establish a statistical department. It is conceivable that the establishment of the Department of Statistics will be delayed for at least some time if it is not because of the invitation of other work by Hotlin and Wald. Establish a statistical department, the school will provide the corresponding venues, funds and students to smoothly carry out teaching and research work.

(3) Professional settings should keep pace with the times and be suitable for school

Professional settings are advancing with the times, and the school system should include both positive and negative aspects. When the social industrial structure changes, schools should actively set up or revoke corresponding majors according to their own conditions.

The development of modern statistics and the establishment of statistics majors in various universities are the processes that accompany each other and promote each other. The development of modern statistics has created a need for statistical education and research, so the statistics profession has been widely established; and the establishment of statistics, the establishment of statistics and statistics, and the promotion of statistical education and research, Promoted the development of modern statistics.

At present, China is in the stage of tackling difficulties in industrial transformation. The foundation of industrial transformation is the adjustment of talent structure. An important guarantee for the adjustment of talent structure is the setting and adjustment of university majors. Opening a corresponding profession according to the needs of the society, students and the objective conditions of the school is an important part of helping the country to develop and promote social progress.

It is worth noting that one of the most important aspects of the profession mentioned above is the interdisciplinary curriculum and professional settings. The division of the profession itself has its own limitations. Nowadays, the demand for interdisciplinary talents in the society is huge. The disciplines and interdisciplinary skills that meet the needs of society, students and disciplines will be the problems that universities need to think about continuously. In addition, when it comes to revoking or adjusting a profession that does not meet the needs of society or school, it needs to retreat. The contraction is for better expansion, and the resources saved can be used in more specialized professions.