 Proceedings
|
Invited TalksTopic 7 : Statistics education and the wider societyFor each of the last two ICOTS, this topic’s abstract has emphasized the ever-increasing need for society at large to be statistically educated. This is even more pertinent and true now. We, as statistical educators, continue to lead and extend debates in the media and elsewhere on issues of inequality, crime, effects of smoking, of alcohol, to name but a few. We are bombarded by results from surveys and polls. As those most concerned with educating society in the ways of statistical thinking and reasoning, how can we encourage people to want to engage in statistical learning? How can we contribute to subject-specific learning where statistical knowledge and understanding is necessary (nearly everywhere!!)? For topic 7, Statistics Education and the Wider Society, contributions from a wide range of subject perspectives are encouraged. Possible contributions relating to statistical education to the statistical profession itself, to issues in journalism, mass media, engineering, the sciences, medicine, psychology, law, commerce, business, economics, government, industry and sport are particularly welcome, but are not limited to this list. As at the last ICOTS we hope this topic will act as a forum for asking and answering current questions about how statistical education and teaching should develop to ensure sound statistical understanding and usage in other disciplines and society at large.Session 7A: Statistics and the media7A1: Association-causation problems in news storiesMilo Schield Keck Statistical Literacy Project, United StatesStatistical educators strongly emphasize the importance of distinguishing association and causation. Yet this semantic distinction is often obscured due to an inappropriate choice of words. This paper investigates related problems–inaccuracies, omissions and ambiguities–in numberbased news stories. It investigates association-related problems involving large numbers, confusion of the inverse, missing context, times-less and times-more comparisons, incomplete comparisons, slope comparisons and confusing “frequently” with “likely.” It investigates causation-related problems involving causation words and action verbs. These problems may create reader confusion and misunderstanding. Data is needed on how readers understand the presentation of association and causation in the media. Statistical educators, journalism faculty and quantitative journalists should join together in analyzing these problems in number-based news stories. Paper 7A2: Spinning heads and spinning news: the American media’s gap in quantitative reasoning skillsRebecca Goldin George Mason University, United StatesNews increasingly depends on a careful dissection of numbers. Statistics are everywhere, from how many people are not covered by health insurance to whether Vitamin E is good for you or not. Yet for being so prevalent, statistics are badly understood by journalists and the general public. Misguided representations of science can actually shape public policy, legislation, and individual choices. We describe why it is so important that media writers understand basic concepts from statistics, epidemiology and even toxicology using examples in current media coverage. We also discuss the gulf between the scientific and media cultures, which can lead to bad science coverage. We finish with constructive suggestions for improvements in communication of scientific progress by media writers. Paper 7A3: Statistics on national radio: some insights from working with professional broadcastersKevin McConway The Open University, United KingdomBBC Radio 4, the second most popular radio station in the UK, is a spoken word station run by the British Broadcasting Corporation (BBC). A programme on numbers, More or Less, has grown from a small beginning in 2001 to be a key regular part of the Radio 4 current affairs output, with around 1.2 million listeners to each programme and an international reach through two websites, podcasts and streamed audio. Its ‘numbers’ remit is interpreted very broadly, but economics and statistics form the bulk of the topics covered. Since 2005, the programme has been produced in partnership with the Open University. This paper describes the partnership from the point of view of the academic partners, outlining the differences in approach, purpose and timescales between academics and journalists, and proposing that these differences have contributed to the strength of the programme and its role in educating students, journalists and the wider public. Paper Session 7B: Statistics and sports7B2: Statistical models for student projects with sports themesRobin Lock St. Lawrence University, United StatesWe describe several types of student project assignments that involve applications of statistical models to address questions arising from sports data. Although we illustrate these ideas with examples from specific sports, our goal is to provide sufficiently general guidelines to allow instructors to adapt and extend the topics to different sports, teams, leagues or levels of play. Some of the projects are accessible to students at the introductory levels while others are more appropriate for a second course or even an undergraduate capstone/thesis. Topics include Bill James’ so-called “Pythagorean law” for estimating team winning percentages, investigations of home field advantage, logistic regressions on the chance of winning a match based on boxscore statistics, the use of empirical Bayesian Stein estimators to project player performance over a full season based on early season results, and methods for modeling outcomes in seeded tournaments. Paper 7B4: Using sports data to motivate statistical concepts: experiences from a freshman courseVittorio Addona Macalester College, United StatesWe discuss observations from teaching a freshman course: Statistical Analysis of Sports and Games. In many respects, this is a standard first college statistics course. Analyzing sports data, however, generates student interest in statistical ideas. Moreover, quantitative analysis in sports has become a serious research field, and many professional teams now employ statisticians. It is important for students to realize that academic and non-academic opportunities exist beyond the course. There also remain issues that need to be addressed before sports statistics courses become commonplace. They should: (1) appeal to both male and female students, (2) have a broad focus and not be too baseball-centric, (3) be primarily about statistics, not sports, and (4) have access to more appropriate textbooks. At the core of any statistics course is a desire to answer questions in meaningful ways. We offer ideas on how this can best be accomplished in this context. Paper Session 7C: Statistics in psychology and the social sciences7C1: Learning probability and statistics: cognitive and non-cognitive factors related to psychology students’ achievementCaterina Primi University of Florence, ItalyFrancesca Chiesi University of Florence, Italy The aim of the present is to ascertain the impact of both cognitive and non-cognitive factors on probabilistic and statistics reasoning in psychology students enrolled in introductory statistics courses. It was hypothesised that performance was related to the student’s general and mathematical background (cognitive factors), math self-efficacy and attitudes toward statistics (non-cognitive factors). A structural equation model was specified in which cognitive and noncognitive factors were considered as the exogenous latent variables having an impact on both probabilistic and statistics reasoning. Results stressed the role of both cognitive and non cognitive factors suggesting that competence as well as attitudes and self-efficacy should be the focus in planning interventions to help students in increasing performance. Paper 7C2: Human sciences student’s difficulties in parametric tests: a contribution to statistics educationNoëlle Zendrera Catholic Universty of l’Ouest, FranceThe purpose of this article is to present part of the results of our research in statistical education. In particular, we have studied difficulties faced by undergraduate students in human sciences when solving a concrete and complete problem within a parametric hypothesis test. Two methods of collecting data have been used: the written proof and the individual clinical task interview. Results obtained by written proof among 90 students have been quantified in order to give a global vision of the seven variables studied. The written test has also facilitated the process permitting us to select 10 students for the task interviews. The results obtained in the interviews show that these students displayed original conceptions regarding many concepts implied in these kind of tests, different to certain misconceptions commonly encountered in the literature. Paper 7C3: A cross-cultural psychometric evaluation of the attitude toward Estrada’s statistic scale in teachersAssumpta Estrada University of Lleida, SpainJorge Bazán Pontifical Catholic University of Peru, Peru Ana Aparicio Pontifical Catholic University of Peru, Peru Several attitudes toward Statistic scales have been proposed in the literature by considering samples of university students but specific scales with cross-cultural validity are not known for teachers. In this study we show the psychometric characteristic of the scale of attitudes toward the statistic proposed by Estrada (2002) and Estrada, Batanero and Fortuny (2003). The scale was applied to 288 in-service and prospective teachers, 140 from Spain and 148 from Peru. Item analysis with a classic perspective and using rating model (Andrich, 1978) was conducted. Results indicate that three items were not consistent with the scale. In addition, the final version of the scale was submitted to an evaluation of its dimensionality and reliability. Results indicate that scale is reliable and presents evidence of multidimensionality. Paper Session 7D: Statistics education for engineering7D1: Individualised learning for engineers — combining face-to-face teaching with non-linear web learningHelle Rootzén DTU Informatics, DenmarkHow do we meet the challenge of engineering students having different levels and different learning styles? How do we make mathematics and statistics more inspiring? Our approach is to combine face-to-face learning with e-learning. We have developed a prototype for a new kind of elearning called HEROS (Higher Education Re-usable Objects in Statistics): a non-linear learning object based learning system. Non-linear learning means that the students themselves can design– part of the course–by selecting different learning materials and use them in the order they like. Our learning materials include: an interactive “book”, assignments with interactive hints, videos, podcasts, games etc. So far we have developed a continuing education course in basic statistics and right now we are developing a university course in basic engineering mathematic –both based on the principle in HEROS combined with face-to-face learning. Paper 7D3: The impact of problem-based learning on statistical thinking of engineering and technical high school studentsAndreja Drobnič Vidic University of Ljubljana, SloveniaEngineering statistics teachers are often faced with two problems: time pressure and the lack of curriculum guidelines. This report presents two cases from Slovenia, illustrating how the problem-based learning approach can be applied in engineering education at university level and in high schools. This approach enables activities which encourage statistical thinking and improve students’ problem-solving skills, teamwork skills as well as skills for effective use of information technology (IT) even with a time-pressured curriculum. The report describes organization of engineering problems, which all trigger learning of statistics in a typical introductory statistics course as well as organization of engineering problems (cases) which connect statistical contents with some other subjects in technical high school curricula (such as sociology, IT, mathematics). Furthermore, the results of students’ success through problem-based learning are provided. Paper 7D4: Using directed online tutorials for teaching engineering statisticsRichard Wilson University of Queensland, AustraliaSince 2006, an internet based tutorial system for service courses in applied statistics has been under development. The motivation has been to provide students with relevant contexts according to discipline area (courses are multi-disciplinary), direction in modeling and analysis of data and interpretation of results, and individual data sets to assist them to see the random nature of data more clearly. Students obtain all information online, and complete their analyses and online reports in a tutorial setting with access to a tutor. As tutors need to correct different results and consequent discussion, they correct online and so cope with different numerical answers and interpretations of results. A description of the current version of the system will be given. A discussion of student responses for the tutorials will be discussed, highlighting the strengths and weaknesses of the system. Future directions will be covered briefly. Paper Session 7E: Statistics for biology and the health sciences7E1: Promoting autonomous learning in statistics among undergraduate medical studentsMargaret MacDougall University of Edinburgh, United KingdomFor many educators, the idea of autonomous learning in statistics among undergraduate medical students may seem too much of a utopia to be worth pursuing. This unhappy scenario can place undue pressure on teachers of statistics to assume the rôle of instant service provider. Furthermore, tomorrow’s doctors need to make informed judgments among competing sources of evidence, including statistical findings, in order to gain an accurate perspective on best practice. As teachers of statistics, we therefore have a responsibility to pursue strategies for turning the autonomy dream into a reality. Here, I seek to encourage educators through providing examples of such strategies based on successful teaching practise and findings from recent educational literature. In so doing, I take a multidisciplinary approach to statistics education and recommend ideas for integrating these strategies with teaching activities within existing medical programmes. Paper 7E2: A model to optimise statistical independence and critical thinking amongst researchers in a diverse disciplinary settingDiana Battistutta Queensland University of Technology, AustraliaHelen Johnson Queensland University of Technology, Australia Cameron Hurst Queensland University of Technology, Australia Dimitrios Vagenas Queensland University of Technology, Australia Ross Young Queensland University of Technology, Australia Reflective, open-minded, evidence-based decision making is a defining feature of critical thinking. The Institute of Health and Biomedical Innovation (IHBI) conducts research in diverse disciplines under qualitative and quantitative paradigms. Research methods capacity building activities must engage established as well as the majority population of postgraduate student researchers. A support model was designed utilizing a social constructivist framework and package of activities catering to various levels of statistical knowledge and phobias defined. Established and refined on the basis of three years experience, we describe the activities and relative uptake by researcher stakeholders. We suggest that this novel packaging promotes: (i) interdisciplinary exchange and intellectual curiosity, ii) more rigorous application of statistics, and (iii) professional development of team statisticians. Paper 7E3: Statistics for the biological and environmental sciences: improving service teaching for postgraduatesRuth Allen Lancaster University, United KingdomAndrew Folkard Lancaster University, United Kingdom Gillian Lancaster Lancaster University, United Kingdom Bev Abram Lancaster University, United Kingdom A challenge for statistics educators is to maximise the effectiveness of service teaching to nonstatisticians in order to create successful end-users of statistics, build research capability and to raise the profile of statistics within the wider community. In 2008 we undertook an evaluation of statistics service teaching within the Lancaster Environment Centre at Lancaster University. Collaborating with staff and students enabled a thorough study of the strengths and weaknesses of the course. Recommendations for how to make the course more suitable for occasional users in the biological sciences were made. An evaluation of the revised course was completed in 2009 and the results, presented here, are compared with feedback from 2008. We demonstrate that with close collaboration between all departments, making swift and tangible changes based on ‘real’ suggestions by the students who take the course is the only way to reach specific users. Paper Session 7F: Statistics in business7F2: High dimensional data – a growing businessBart De Ketelaere Catholic University of Leuven, BelgiumPaul Darius Catholic University of Leuven, Belgium When it comes down to understanding, predicting or optimizing business, the tools provided by statistics form an important corner stone. Training of statistics at university level is a crucial factor here and should be well aligned with the actual needs after graduation. Those needs depend on the actual business type involved, and are approached here from an engineer’s point of view. We will briefly touch some general trends and issues that might be considered when forming the new generation of engineers and statisticians. Paper 7F3: Some arguments for integration of qualitative methods into business statistics coursesIddo Gal University of Haifa, IsraelIrena Ograjenšek University of Ljubljana, Slovenia The processes which facilitate informed decision-making in business contexts have been gaining in importance in many areas such as business, healthcare, education and government. Consequently, quantitative methods which form the basis of curricula for typical courses on business statistics are finding recognition in the wider business community. In contrast, qualitative methods, although used by management researchers and applied statisticians alike to solve specific research problems, suffer the unjustified exclusion. This paper outlines some of the core assumptions of qualitative research methods and provides three examples illustrating selected types of qualitative methods that are useful in the business context. We argue for the introduction of a balanced mixed method approach early in the process of studying business statistics, as a preferred basis for developing business students’ ability to respond to diverse types of real-life managerial challenges. Paper Session 7G: Statistics for non-quantitative majors7G1: Using media reports to promote statistical literacy for non-quantitative majorsStephanie Budgett University of Auckland, New ZealandMaxine Pfannkuch University of Auckland, New Zealand At The University of Auckland we teach an undergraduate course entitled Lies, Damned Lies and Statistics, the purpose of which is to facilitate students to “think statistically” when confronted with evidence-based arguments. In this paper we first describe how we use media reports in teaching to enhance students’ ability to understand and evaluate statistically based information. Second, we report our observations on interviews we conducted with three non-quantitative and three quantitative majors seven months after they completed the course. A comparison of their responses to two media reports does not suggest any meaningful difference between the two groups in terms of their understanding of statistically-based information. There does however appear to be a difference in the way the two groups explain their understanding. Possible reasons for these observations are discussed. Paper 7G2: Luring non-quantitative majors into advanced statistical reasoning (and luring statistics educators into real statistics)Sean McCusker University of Durham, United KingdomJim Ridgway University of Durham, United Kingdom James Nicholson University of Durham, United Kingdom Introductory courses in statistics often progress from simple analysis to anova and interaction. The epistemology is close to elementary physics–simplify, decompose, analyse, then add the bits together again, and you will understand the phenomena. Most interesting problems in social sciences are multivariate, and variables interact in complex ways. Biological systems provide a better analogy–below a certain level of complexity, you destroy the problem by decomposition. The big statistical ideas in social policy are things like the shape of multidimensional data surfaces, effect sizes, interactions, limiting factors, extrapolation, and confidence intervals. Statistical significance is usually irrelevant; assumptions about distributions are risky. Non-quantitative majors are quite right to be wary of standard statistical models. We offer a plan to lure nonquantitative students into statistical reasoning by proving rich, multivariate data from large scale studies, along with media accounts (which are often simply wrong). Some ‘proof of concept’ will be provided. Paper 7G3: A five step framework for interpreting tables and graphs in their contextsMarian Kemp Murdoch University, AustraliaBarry Kissane Murdoch University, Australia High school, college, and university students as well as citizens encounter quantitative information in a wide variety of media and contexts, such as in books, journals, newspapers, magazines, advertising, the workplace and on the Internet. While it is often assumed that people are able interpret published data in context, this assumption is open to question. In this paper, a Five Step Framework to help both teachers and students to interpret data in the form of tables or graphs is described and exemplified. Development of the Framework by Kemp was based on the SOLO taxonomy devised by Biggs and Collis. The Framework offers a progression from simple numerical reading of a table, to more complex interpretations of tables and graphs needed for a better understanding of data in their context. The Framework has been used successfully in primary, secondary and tertiary mathematics education to support both students and their teachers. Paper 7G4: How we can all learn to think critically about dataIan Gordon University of Melbourne, AustraliaSue Finch University of Melbourne, Australia “Critical thinking with data” aims to teach undergraduate students to review and evaluate critically statistical information and arguments, to convince them of the relevance of statistical literacy, and to imbue them with the disposition of an enquiring statistical consultant. We use diverse strategies for engagement and broad content. Our materials are relevant, accessible, rich, interesting and varied; many are taken from the general media and are media rich. Presentation of statistical content is enriched through the use of eminent guest lecturers, case studies, and many real world examples. We do not rely on mathematics but present multiple perspectives on ideas, concepts and principles. Our abstract building block concepts are first presented through appealing and concrete case studies. Once students grasped the ideas in one context, they generalise these concepts through repeated exposure to manageable open-ended interpretative tasks using on-line and standard assessment tools. We provide illustrations of our approach. Paper Session 7H: Official statistics in statistics education:links between IASE and IAOS7H2: Improving statististical literacy by national and international cooperationReija Helenius Statistics Finland, FinlandThe advancement of statistical literacy is a challenge for each statistical office. However the goals and willingness do not always connect with reality, especially at times of scarce resources and cut budgets. At such time new ways to improve statistical literacy must be adopted. Sharing experiences, learning by best practices, networking and cooperation are increasingly important forms of activities. This paper presents examples on activities carried on to this end. These examples are discussed form the point of view of different user groups, user friendliness and various forms of cooperation. Paper 7H3: Some case studies on the links between National Statistical Offices and statistical educators: what are the main developments?Dennis Trewin Dennis Trewin Statistical Consulting, AustraliaIt is important for National Statistical Offices to work with statistical educators. Improved statistical literacy will mean that their outputs are used more extensively and with more wisdom. The paper will be based on a series of case studies of what is being done in seven countries around the world to improve the link between National Statistical Offices and statistical educators (or teachers), especially with supporting resources. The focus will be on schools. As well identifying some the most important initiatives, I will try to draw some lessons on the way forward. Paper |