Educating Professional Statisticians – Some thoughts from Brazil
Statistics as a profession is facing a variety of challenges:
• Big data and the associated revolution in data availability, accessibility, speed of production and pressure to ‘do something’ with the available data;
• Artificial intelligence, machine learning, data mining and many other trends and tools where the ‘art and science’ of data analysis, modeling and inference are being taken up by software which is evolving at a very fast pace, and which aims to depend less and less on the competence and skills of the users;
• Fast evolution of methodology and technology, demanding increasing investment of time and effort to keep up with the developments;
• Ever wider areas of application of Statistics with their own dialects, and their promotion of ‘do-it-yourself’ by practitioners who do not possess broad statistical education, but who master very complex and specialized statistical tools (models , methods and software) relevant to their fields.
Given these challenges, how can statistics education deliver Statisticians who can thrive in these challenging times and later help to keep them ‘fit’ throughout their careers? In this talk I consider some ideas to help facing these challenges.
Pedro is Principal Researcher at the National School of Statistical Sciences (ENCE) from the Brazilian Institute of Geography and Statistics (IBGE). He spent most of his professional career at the Methodology unit of IBGE (1981-2002). He then directed the National School of Statistical Sciences from 2003-2006, and worked as Principal Research Fellow in Sampling at the Southampton Statistical Sciences Research Institute (S3RI) 2006-2010, before returning to ENCE in March 2010. His main research interests are survey and sampling methodology applied to household and business surveys, as well as the analysis of survey data. He is President-Elect of the International Statistical Institute (2013-2015), presided the International Association of Survey Statisticians (IASS) during 2007-2009, presided IASI (the Inter-American Statistical Institute) during 2004-2005, is a member of the Brazilian Statistical Association, of the Royal Statistical Society, and the American Statistical Association. He was the chief editor of Revista Brasileira de Estatística from 1997-2002, and is currently associate editor to Survey Methodology, International Statistical Review, Revista Brasileira de Estatística, and Estadística. Pedro has extensive experience in teaching both at graduate and undergraduate levels, as well as short training courses both at IBGE, and other universities and statistics agencies in Latin America.
Communicating Statistics: Telling the stories about data
All statisticians are involved in telling the story from a data analysis. The effective practice of statistics involves an ability to work with clients from a wide variety of backgrounds. We need to be able to understand the research question of the client, to apply appropriate methods to address this question given the data available (or to design a method for collecting appropriate data) and then to present the results of this work to the client. The need to help students develop this skill is strongly recommended in recent reports from working groups of the American Statistical Association. In the first part of the talk, I will describe these skills and how the development of these skills can be built into client-focused classes. In the second part of my talk, is will focus on the question - how do we generate a broader appreciation for and understanding of the impact of statistics? I will describe the development of a webcast / podcasts 'Stats + Stories', a joint project that involved the Department of Statistics and the Department of Media, Journalism and Film at Miami University. This program tells the stories behind the statistics and the statistics behind the stories.
John Bailer is Vice President of the International Statistical Institute (ISI) and University Distinguished Professor and Chair of Department of Statistics, Miami University, Oxford, Ohio (USA)
Statistical Education in China
This talk first discusses the changes of statistical education in China in the last 35 years after the culture revolution, and then the current status of statistical teaching in primary school, middle school, high school to college and universities. With the rapid development of economics and change in society, the need of statistical graduates increases rapidly. Therefore the statistic program in colleges and universities have become a popular one and the number of students enrolled grows year by year. Consequently, China now has the world’s largest number of undergraduate and graduate students in statistics major. The main challenge for statistic educators in China is how to ensure the competence of our graduates against the challenges brought by the big data trend and competitions from majors like computer science.
Wei Yuan is a professor of statistics in Renmin University of China and the Director of National Survey Research Center., Former Executive Vice President of Renmin University of China and Chair of National Statistical Higher Education Committee (1993-2014). He is the elected ISI member (2004), IASS Council Member (2005-2009), ISBIS Council Member (2007-2009).
Brave New World of Data
Frequent data generation has promoted a fundamental change in various sectors of economy and personal consumption of goods and services. We are in a new era, surrounded by data on all sides. Huge masses of data resulting from automatic collection processes, electronic instrumentation, online transactions and historical data collected over many years. The moment is big, before the mining, fishing, snooping, dredging, and after the massive. A new consumer market, a new retail, emerges under new technologies and behaviors. In this new world, order of magnitude of the data sets to statistical science has experienced adaptive processes in order to continuing to provide useful knowledge effectively. Traditional analysis strategies have been reviewed in the light of the greatness of information, contaminating and missing data, variables that are not identically distributed, presence of non-stationary and non-numeric variables. It is imminent the need for developing alternative strategies based on methods of sequential estimation, adaptive segmentation structure and multiple combinations of models, directly associated to efficient computational strategies that provide real-time responses. This presentation discusses the opportunities that this brave new world offers us, in the light of the counterpoint between the ultra-modernity of data capture mechanisms and the statistical methodologies. And how we may contribute to the formation of a new professional required to act within it efficiently.
Francisco Louzada is a Professor of Statistics at the Institute for Mathematical Science and Computing, University of São Paulo (USP), Brazil, Research Productivity Fellow of the Brazilian founding agency CNPq (level 1B), He has served as President of the BSA and Secretary of the Brazilian Chapter of the International Society for Bayesian Analysis. Prof. Louzada has conducted consulting, training and technological transfer of statistical methodologies for more than 40 companies and institutions. More details
Hollylyne S. Lee
Stepping Outside Classroom Walls: Designing Experiences for Teachers in a Massive Open Online Course MOOC on Teaching Statistics
In many countries around the world, institutions and local schools struggle to help current and future mathematics teachers be prepared to teach statistics. My design challenge was to consider how an online massive format for professional teacher learning could possibly make any impact in the way teachers can teach their students statistics. In this plenary, I will share my design challenges, my guiding principles, and my potential solutions. We will also glimpse at preliminary data from the first implementation of the Teaching Statistics Through Data Investigations MOOC to re-consider the design of experiences for teachers.
Hollylynne S. Lee is a Professor of Mathematics Education in the department of Science, Technology, Engineering, and Mathematics Education at NC State University in Raleigh, North Carolina, USA. Her current work focuses on using innovative approaches and powerful technology tools to engage teachers and learners in statistics education. She has numerous publications, editorial positions, and has recently won several awards for her scholarship and teaching. More information