Developments in the digital world dramatically influenced the statistics profession. Statisticians were faced with the problem of choosing a statistical package from among the many (100+) available. The “List of statistical packages” page in Wikipedia identified 51 statistical packages that were freeware, in the public domain, or open-source; 67 proprietary packages (however, not all were still available); and 6 add-ons to Microsoft Excel.3 This summary understated the spectrum of choices because many of the packages (e.g., R and SAS) included a multitude of specialized component packages. Here is a commentary about computational reproducibility and statistical software at the beginning of this decade:

“One of the consequences of the computer and internet revolution is that more and more scientists promote open source software and reproducible research. Science should be, per definition, both open and reproducible. In the context of statistics6 this means that the published article or report is not the complete scientific result. In order for the results to be reproducible, we should also have access to the data and to a copy of the computational environment in which the calculations were made.

Publishing is becoming more open, with e-journals, preprint servers, and open access. Electronic publishing makes both open source and reproducibility easy to realize. The Journal of Statistical Software, at, the only journal that publishes and reviews statistical software, insists on complete code and completely reproducible examples. Literate Programming systems such as Sweave,9  are becoming more popular ways to integrate text and computations in statistical publications.

 We started this overview of statistical software by indicating that the computer revolution has driven much of the recent development of statistics by increasing the size and availability of data. Replacement of mainframes by minis, and eventually by powerful personal computers, has determined the directions in the development of statistical software. In more recent times the internet revolution has accelerated these trends, and is changing the way scientific knowledge, of which statistical software is just one example, is disseminated.” 

- Jan de Leeuw 200912


The value of statistics became increasingly appreciated by the U.S. public. For example, in 2012, the statistician Nate Silver used the ‘wisdom of crowds’ approach and shrinkage estimators to predict successfully the outcomes of the US Presidential election in all 50 states. He became a media star, and the voice of the popular web blog ‘FiveThirtyEight.’

The 2014 London workshop on the future of the statistical sciences concluded that the numbers of bachelor’s and master’s degrees awarded in statistics both had roughly doubled over the previous 10 years. The representation of women in statistics programs was much better than it was in comparable disciplines such as mathematics, physics, and engineering. At the undergraduate level, enrollment in introductory statistics courses had gone up internationally by 90 percent from 1995 to 2010.

In the 2019 U.S. News and World Report article on the “100 Best Jobs,” Statistics was listed as #2 among the 100 Best Jobs. On a 1 to 10 scale, U.S. News gave the following rankings to Statistics: Salary - 7.2, Job Market - 10, Future Growth - 8, Stress - 8, Work Life Balance - 8, and Overall - 8.2. It rated Statistics as an Above Average profession for Upward Mobility (opportunities for advancements and salary), Stress Level (work environment and complexities of the job's responsibilities) and Flexibility (alternative working schedule and work life balance). The article said that the Bureau of Labor Statistics projected 33.8 percent employment growth for statisticians between 2016 and 2026. In that period, an estimated 12,600 jobs should open up. Statisticians made a median salary of $84,060 in 2017. The best-paid 25 percent made more than $108,500 that year, while the lowest-paid 25 percent made less than $64,230. U.S. News also ranked Statistics as #1 in Best Business Jobs and #2 in Best STEM Jobs, where STEM is the acronym for Science, Technology, Engineering, Mathematics.15  

Statisticians worked to develop more methods for analyzing unevenly spaced time series data, and to improve random forest, bagging and boosting methods. Statisticians made a concerted effort to enforce the “science” aspect of the hot field of data science. Statistical thinking is required for data analysts who use modern classification, cluster finding, and pattern recognition algorithms applied to large opportunity samples. Statisticians are inclined to look for convincing explanatory and predictive models. They won't be satisfied with simply finding an interesting pattern in a big data set.

By 2018, the American Statistical Association had authorized 5 new sections, the Statistics In Imaging Section (SI), the Mental Health Statistics Section (MHS), the Section on Medical Devices and Diagnostics (MDD), the Lifetime Data Science Section (LiDS), and the Section on Statistics in Genomics and Genetics (SGG).

Next Topic (Annals of MSU) during Era 10
Table of Contents

Last revised: 2021-04-19