 |
Invited TalksTopic 5 : Statistics education in the disciplines and the workplaceStatistical practitioners facing the need to search for and/or collect, process, analyze, interpret and/or report on data as part of their job, work in many different subject areas such as humanities, social sciences, business and economics, government, law, politics and journalism. In this set of invited sessions we aim to facilitate their discussions with statistical educators as well as institutional data providers to explore ways in which effective statistical education and training can help improve statistical practice in any given subject area at any given workplace. Accountability conventions, along with values and ethical imperatives of a given discipline, resonate with this theme.Session 5A: Evidence-based policy making5A1: The use of official statistics in evidence based policy making in New ZealandSharleen Forbes Victoria University of Wellington, New ZealandTania Janssen Statistics New Zealand, New Zealand Statistics New Zealand has found in its education initiatives for raising statistical capability that one challenge is making sure data is used the right way. Lack of knowledge can lead to unintentional misuse of statistics. The power of official statistics is used to show policy analysts how to gain better value and make evidence based decisions. Statistics provide the quantitative evidence supporting Government decision making. While not the only form of evidence used in policy making, several examples will be given that demonstrate its critical importance. In the social area statistics have informed dramatic changes in smoking policy over the last fifty years and in the economic area the policy uses of the Consumers Price Index are discussed. Commuter data demonstrates local government level use of Census information. The high cost associated with making wrong policy decisions as a result of not using or misusing statistical information is also discussed. Paper 5A2: Challenges to evidence-based policy making in the decentralized U.S. statistical systemNancy Potok United States Census Bureau, United StatesAmy O’Hara U.S. Census Bureau, United States Ron Jarmin U.S. Census Bureau, United States One goal of the current U.S. presidential administration is to accomplish more evidence-based decision making to drive National policy. The FY 2014 President’s Budget contains funding for rigorous evaluations across government as well as evaluation capacity building. The stated purpose is to build knowledge so that spending decisions are based on strong evidence that investments yield the highest social returns. One promising approach to conducting data-driven evaluations cited in the budget documents is using administrative records to conduct low-cost evaluations. The US Census Bureau has extensive experience linking statistical survey data to administrative data to create powerful data sets. However, in a decentralized statistical system, many challenges exist to sharing records and data among federal statistical agencies. This paper discusses both advances and challenges in increasing evidence-based policy making. Paper 5A3: Statistics education, collaborative research, and LISA 2020: a view from NigeriaOlushina Olawale Awe Obafemi Awolowo University, NigeriaEric A Vance Virginia Tech, United States Reforms in pedagogical techniques, collaborative research, and the improved use of information technology are strategies in teaching research-based statistics courses that have not yet been implemented in Nigeria. Based on results from empirical survey, we review the current level of awareness and challenges facing statistics education in Nigeria. One survey revealed that 69.4% of the respondents did not know statistics was a separate discipline from mathematics while 44.4% had never taken a statistics course. We discuss the challenges of collaborative research in Nigeria and introduce LISA 2020, a program to bridge the gap in statistics education between developed and developing countries by training statisticians from developing countries to become effective collaborative statisticians who can train other collaborative statisticians in the future. Paper 5A4: International statistical standards as enabler for evidence-based policy making: the case of tourism statisticsOliver Herrmann United Nations World Tourism Organisation, SpainThe scale of tourism world-wide and the growing complexity of tourism-related data call for improved knowledge about the phenomenon. The United Nations World Tourism Organisation (UNWTO) is recognized as the appropriate organisation to collect, to analyze, to publish, to standardise and to improve the statistics of tourism. UNWTO facilitates an environment in which tourism statistics and analysis can take place with particular emphasis on the need for international comparability in tourism statistics. The paper describes UNWTO’s framework directed at developing tourism statistics and underlying training and capacity building methods. The capacity building initiative aims at producers but also users of tourism statistics and introduces them to the international recommendations on tourism statistics and how to implement these standards. Paper Session 5B: Evidence-based management5B2: The prevalence of statistics and data mining in management journalsShirley Coleman University of Newcastle, United KingdomAlex Douglas Liverpool John Moores University, United Kingdom Jiyu Anthony Heriot-Watt University, United Kingdom Postgraduate students tackling Business and Management Master’s degrees are told they will need to understand statistical terms when reading research journals. Some of the journals their lecturers recommend are philosophical or conceptual in nature but others are experimental and evidence-based. Many of the articles are purely qualitative and those that include quantitative analysis are either very basic or are rather sophisticated; it could be interpreted that there is a divide rather than a continuum between authors who are confident in statistics and those who avoid them. This paper reviews relevant journals and examines the prevalence of statistics and suggests some explanations for the findings. The quantitative methods lecturer can use this information when deciding what to teach and how to help students appreciate the importance of statistical analysis. Paper 5B3: Bringing the workplace into a National Certificate in Official StatisticsSharleen Forbes Victoria University of Wellington, New ZealandAlan Keegan Statistics New Zealand, New Zealand New Zealand’s National Certificate of Official Statistics was designed to raise the capability of public sector employees, in particular policy analysts giving advice to senior managers or Ministers. The competency based 40-credit certificate was first introduced in 2007. Four taught units (worth 24 credit) give students skills in basic official statistics and in critically evaluating statistical, research, policy or media publications for their quality (of data, survey design, analysis and conclusions) and appropriateness for some policy question (e.g. how to reduce problem gambling). Case studies are used to set the statistics learning into the real world context of the students. An evaluation of the first cohort of students led to a compulsory “umbrella” workplace-based statistics project worth 16 credits being introduced. Some of the challenges for students and their managers in achieving a project that is both useful to the organisation and meets the requirements of the Certificate are discussed. Paper 5B4: How do school principals understand and use the statistics in reports from national large-scale assessments?Iddo Gal University of Haifa, IsraelHani Shilton Ministry of Education, Israel This paper describes selected results from a project by the National Authority for Measurement & Evaluation in Education (RAMA) in Israel that examined how school principals understand the statistical information in national assessment reports they work with, and use the results for school improvement. The paper presents a new multidimensional conceptual model developed to guide the project and preliminary results based on interviews with school principals and supervisors. Based on the results, we reflect on the nature of the complex interpretation tasks faced by school principals who use reports on comparative large-scale assessments involving both achievement tests and school surveys, and on the link between principals’ statistical literacy and evidence-based management practices as school leaders. The study examines implications for statistics education and professional development of school principals, and for the design of statistical reports. Paper Session 5C: Statistics education beyond qualification (panel discussion)Ronald Wasserstein American Statistical Association, United States Gillian Lancaster Lancaster University, United Kingdom Judith-Anne Chapman Queens University, Canada Kazunori Yamaguchi Rikkyo University, Japan Summary Session 5D: Development of statistical thinking in the workplace5D1: Improving statistical literacy at universityJennifer Brown University of Canterbury, New ZealandHilary Seddon University of Canterbury, New Zealand Elena Moltchanova University of Canterbury, New Zealand Jenny Harlow University of Canterbury, New Zealand Irene David University of Canterbury, New Zealand At the University of Canterbury, New Zealand, the Department of Mathematics and Statistics offers a range of introductory statistics and mathematics courses. Nearly one third of the first year students enrol in our introductory statistics courses. Given this level of interest we could consider our work done: we don’t! We have an integrated programme to improve statistical literacy across all the campus. In this presentation we discuss our work in supporting students with very low levels of numeracy via learning skills support, through to supporting postgraduate students with their higher level statistical needs. Rather than running these as separate programmes we have pulled all these levels of support together and have integrated them into the existing university structures. We now offer a university-wide statistical service. Paper 5D2: The contributions of Six Sigma to the development of statistical thinking in the workplaceDoug Montgomery Arizona State University, United StatesArizona State University’s Ira A. Fulton Schools of Engineering offers a Master Black Belt certification program that enables Six Sigma Black Belt managers, directors, engineers and others to achieve the highest level of technical and organizational Six Sigma proficiency. Black Belt and Green Belt certifications are also offered through a variety of educational platforms. These programs have been running for a number of years. This presentation will review the range of people that participate in these programs, the projects they have undertaken and the program’s impact on the development of statistical thinking in the workplace. Paper 5D3: Development of training methods to accelerate the competencies in Weibull analysis: case study in the automotive industryHalimatou Ndiaye Renault SA, FrancePaul Schimmerling Renault SA, France Michael Huchette UMR Sciences Techniques Education Formation, France Zohra Cherfi Compiègne University of Technology, France Sébastien Castric Compiègne University of Technology, France In the automotive industry, statistical methods are well used in areas such as design, production or analysis of customer returns. However, real problems are often poorly formalized which creates special difficulties of implementation: choosing the appropriate method, selection of data for the model, physical and technical interpretation of results… Weibull analysis is a statistical method used to characterize the lifetime of a product during testing or in-service. It gives indication on failure mechanisms and the causes of product failures. It also predicts the expected number of failures based on real observed failure data. This paper focuses on the characterization of professional knowledge for Weibull analysis. It is based on analysis of technical notes written by engineers from RENAULT SA. The aim is to improve internal training. We present the context of the study, describe the implementation of the method and then explain how to use our results to improve training. Paper Session 5E: Mentoring young statisticians in the workplace5E1: The importance of inter-personal skills on statistical teamsSara Fett Mayo Clinic, United StatesUniversity students are often prepared to meet the technical demand of positions in their field of study. During my 20 year tenure as a manager of Statistical Programmer Analysts, it has become very evident to me that an employee’s interpersonal skills are equally important to their success in the workplace. To help mentor these skills in our young statisticians we have developed both formal and informal mentoring strategies. These strategies are centered on skills required for effectively working in a team directed environment. Some of these skills include communication, collaboration, innovation, initiative, inquiry, networking and knowledge sharing. In addition to sharing some of these strategies, background and anecdotal evidence will be used to help demonstrate the importance of these inter-personal skills. As you are preparing our next generation of Statisticians, it is my hope that you will find additional ways to weave more of these skills into their educational experiences. Paper 5E2: Best practices in mentoring and training young statisticiansLisa Grace Bersales University of the Philippines, PhilippinesMentoring statisticians in the university setting and updating practicing statisticians in the workplace of new statistical methods have produced highly motivated and trained statisticians when the following practices are used:
Paper 5E3: Mentoring advanced and newly graduated masters and Ph.D studentsLinda Young University of Florida, United StatesOften when reviewing the training of graduate students, the focus is on the technical skills needed in their future workplaces, whether in academia, business/industry, or the government. One of the challenges new graduates face as they transition is deciding how to approach a problem that is not anchored to the methods associated with a course. Some problems require innovative adaptations of diverse methods. Consulting classes can offer students an opportunity to begin solving real-world problems and to improve oral and written communication skills. Internships provide students with an opportunity to experience a workplace. Yet, even with the best of preparation, recent graduates need mentoring if they are to continue to develop professionally after graduation. Both employers and new hires can be proactive in ensuring this professional development occurs. Paper Session 5F: Bridging the gap between current statistical practice in the workplace and modern statistics5F1: Tradition should not supplant understanding and insightRichard Wilson University of Queensland, AustraliaJohn Maindonald Australian National University, Australia Technological changes and theoretical advances in the past several decades have created new demands and opportunities for cooperation between the statistical mainstream and application area specialists. Against this, traditions of statistical analysis have become embedded in some places that too often hinder understanding and insight. This paper will discuss: a) approaches that were never a good idea; b) common approaches which are outdated due to advances in modeling and computer technology; c) use of over-simplistic modeling assumptions; d) data pre-processing which removes key information; and e) discipline focus on one specific statistical paradigm, rather than choosing the paradigm that relates best to the research question. How can change best be effected through education and collaboration? What role may applied and mathematical statisticians have in such change? Paper 5F2: Once were warriors: the need of re-education in mathematics and computing for life “scientisticians”Jorge Navarro Alberto Independent University of Yucatán, MexicoThis paper discusses the pros and cons of the new computing era and the status of mathematics in Life Sciences in relation to the training in statistics a life scientist is expected to have. Reflections are made about how preferences in the use of software and the topics taught nowadays in mathematics at University level have mined the creative abilities of the life scientists whenever data analysis is an on due task. Contrasts of the Mathematics and Computing taught in the pre-GUI era and the paradoxically diminishing influence of these scientific subjects at present times are given. Examples of the powerful computing concepts and the ubiquitous computing resources useful for the data analyst and avenues of opportunity of applications of mathematical ideas are also presented, in order to make suggestions in the curricula that respond to the changing demands of what life scientists are expected to do. Paper 5F3: Training to develop modern statistics in the workplace using R and R Commander – experiences from the New Zealand government sectorIan Westbrooke Department of Conservation, New ZealandPeter Ellis Ministry of Business Innovation and Employment, New Zealand There is growing demand for robust, quantitative support for policy development and measurement of outcomes. In tandem, there are increasing challenges and opportunities from new ways of collecting data, often in massive quantities. Key modern statistical skills required in our workplaces include graphing and data exploration and visualisation; modelling, especially the linear model and its extensions, with emphasis on effect sizes rather than tests; designing and appropriate analysis of both small and “big” data; and application of statistical software/computing to support these. One strategy for dealing with these needs and challenges in the workplace is to increase skills of non-statisticians in the workplace. We describe our experiences in carrying training focussed on two contrasting government departments, and explain why and how we have used R and R Commander software as key elements. Paper Session 5H: In search of evidence: exploring the relationship between real workplace based data and statistics education5H1: Workplace and official statistics: how can higher education contribute to a better relationship?Mojca Bavdaž University of Ljubljana, SloveniaLejla Perviz University of Ljubljana, Slovenia Irena Ograjenšek University of Ljubljana, Slovenia Official statistics data are needed in the workplace (for benchmarking, market analysis, etc.) but have been largely under-utilized. A study among businesses of five countries in the BLUE-ETS project suggests it might be hard to find relevant data and use them adequately, which might also reflect the lack of search skills and knowledge about how to apply official statistics to the business situation and interpret the results. In this paper we link the origins of this state of affairs to higher education where an important part of future labor force obtains relevant knowledge about, attitudes towards, and skills to find and use, official statistics. Building on a survey among educators from the European EQUIS-accredited business schools, we aim to provide answers to two questions: (1) how can business school educators contribute to the broader use of official statistics, and (2) how can official statistics providers support business school educators. Paper 5H2: Experiences with real and accessible recent data in context to motivate student learning at higher levels in statisticsJohn Harraway University of Otago, New ZealandMotivational data and statistical analyses arise in research in the workplace and at university. Videos of researchers developing context are recorded. As well as designed experiments researching food authentication, the videos show data from surveys reviewing social issues. These include a postal survey of 2200 citizens investigating use of taxes for building a controversial sports stadium, a telephone survey of attitudes of 1200 women to alcohol consumption during pregnancy, and contrasting opinions of 5000 tourists from Japan, Australia and Germany to the attractions of New Zealand. The studies are practical, relevant and locally generated which capture student interest. Survey design is discussed. Statistical techniques used are regression modelling and multivariate procedures. A data set is given to students who have eight weeks to analyse it and write a report of 25 pages for the organization which commissioned the study. Paper 5H3: Supporting statistical consultant decision-making within a case-based learning environmentGlenn Johnson Penn State University, United StatesDurland Shumway Penn State University, United States Ensuring that students are provided, from the very beginning, with opportunities to engage in activities and decision-making that will be expected of them at the conclusion of their program of study is rarely practised in higher education. Instead, instructor directed, method specific examples predominate course instruction. One reason for this gap may be too few instructional systems that adequately support a structured presentation of, engagement in, or management of the necessary scenario oriented content. This session will share the development of a case-based learning environment that enables faculty to challenge students to demonstrate their ability to successfully engage in open-ended analysis, evaluation, and application of content-specific knowledge within real-world statistical consulting contexts. Providing students with access to indicators of how well they are developing as new professionals throughout their program of study will be discussed. Paper |