Invited Talks

Topic 9 : Technology in statistics education

Technology can play a central role in the education of our students at all levels. Access even to simple modern technical machinery can provide alternatives to mathematical manipulations for exploring fundamental statistics concepts. New visualization technologies can provide distinctive insights into the nature of data and data patterns which assist comprehension of information and inform judgement. Virtual environments offer access to authentic scientific inquiry through simulation and through interactive modes of data manipulation. The Internet provides access to data with rich structure that is of topical importance. Technology now enables educators to reinvent the statistics curriculum and to attain sustainable experience of engaging with uncertainty, in particular through acquisition of creative computing skills for evidence-based problem solving. Thus students can be prepared for collaborative cross-disciplinary partnerships and settings, and hence to advance in the modern technological workplace. These sessions explore how technology can radically impact what students can learn and how our teaching can be substantially enhanced.

Session 9A: The design of digital tools and technology-enhanced learning environments for teaching statistics

9A1: Expressions of uncertainty when variation is partially-determined

Janet Ainley   University of Leicester, United Kingdom
David Pratt   University of London, United Kingdom

Research into students’ knowledge about uncertainty has tended to focus on contexts where the variation observed might easily be attributed to randomness. Yet, the variation observable in many everyday phenomena might be seen as partially-determined, in the sense that one main factor might explain the signal in the variation but additional noise is often inexplicable and might be accounted for as random error. We set out to research how students (age 11) accounted for variation that is in this sense partially-determined. In this paper, we describe how students’ expressions of uncertainty were shaped by particular features within the activity structure in which children recorded and represented the results of their experiments and then modelled the variation.


9A2: A learning trajectory on hypothesis testing with TinkerPlots – design and exploratory evaluation

Susanne Podworny   University of Paderborn, Germany
Rolf Biehler   University of Paderborn, Germany

For a one semester course “Applied statistics – Understanding and solving complex problems via simulations with TinkerPlots” we designed and evaluated new learning material. The course follows an introductory course for statistics and probability and is meant as a focus on probability via simulation and inferential reasoning. At the end there is a four week learning trajectory on hypothesis testing and randomization tests with p-values. As new and innovative learning material for doing a simulation with TinkerPlots we developed a “simulation scheme”, which should help students either to plan, structure or document a simulation and to force them to explain their simulation. Some of the learning trajectories and the simulation scheme will be discussed in detail.


9A3: Students’ emergent roles in developing their reasoning about uncertainty and modeling

Keren Aridor-Berger   University of Haifa, Israel
Dani Ben-Zvi   University of Haifa, Israel

Roles that students take in solving problems can help in guiding and scaffolding their learning and meaning making. We present a case study – part of a UK-Israel research project – that focuses on the emerging roles spontaneously developed by Israeli eighth-grade students (14 years old) in solving a scientific-statistical inquiry task using TinkerPlots2. The task integrated four design approaches: Exploratory Data Analysis, Active Graphing, modeling, and gaming. We examine how this task design played a role in this emergence of students’ roles and how they respectively adopted perspectives on uncertainty and modeling. Implications of the findings are discussed.


Session 9B: Modeling, randomization and simulation tools for connecting data and chance

9B1: Hierarchical data visualization as a tool for developing student understanding of variation of data generated in simulations

William Finzer   Concord Consortium, United States

Data Games is a data exploration environment in which data generated by playing simple games is used by students to build models with which they can improve their game-playing strategies. The data is structured hierarchically, with games at the upper level containing game events at the lower level. Visualization of these two levels, typically with multiple graphs linked by dynamic selection, reveals patterns and variation among the games and illuminates the role that chance plays in the games. We explore the parallels of Data Games to repeated sampling simulations and in a NetLogo forest fire simulation. We look at uses of hierarchical data visualization techniques in each of these three situations.


9B2: StatKey: online tools for bootstrap intervals and randomization tests

Kari Lock Morgan   Duke University, United States
Dennis Lock   Iowa State University, United States
Eric Lock   Duke University, United States
Patti Frazer Lock   St. Lawrence University, United States
Robin Lock   St. Lawrence University, United States

StatKey ( is free online technology created by the Lock family, designed to help introductory students understand and easily implement bootstrap intervals and randomization tests. Randomization-based methods make the fundamental concepts of statistical inference more visual and intuitive, free professors to cover inference earlier in the course, and help students see connections that are otherwise lost with numerous different formulae. To make these methods accessible to all introductory students, our goal was to create technology that is free, widely available (StatKey works in any common web browser), very easy to use, and which helps build conceptual understanding. Although particularly designed for randomization-based methods, StatKey can also be used for illustrating descriptive statistics, sampling distributions, confidence intervals, simple linear regression, and as a replacement for the distributional tables found in the back of many textbooks.


9B3: Teaching resampling in an introductory statistics course

Webster West   Texas A&M University , United States

There is a movement underway among statistical educators to use resampling techniques in introductory statistics courses. For these techniques to be effective, students must have access to appropriate computer software. To help in this process, a number of resampling applets are now available within the StatCrunch package, which can be customized using specific data sets. These applets are discussed in detail in the context of classroom activities with an emphasis on how they may be used to enhance the learning process. Some practical issues related to introducing resampling methods into an introductory course are also discussed.


Session 9C: The emerging concepts of “data science” and “big data” for educational purposes

9C1: Exploring “white flight” via open data and big data

James Ridgway   University of Durham, United Kingdom
James Nicholson   University of Durham, United Kingdom
Sean McCusker   University of Durham, United Kingdom

Open Data and Big Data have important implications for statistics education (and beyond). We sketch some properties of each, and provide examples of the use of social media (an aspect of Big Data) in understanding some of the uses of statistics in argumentation about social change, and the use of Open Data in exploring complex social phenomena. We explore some implications for statistics education and for statistics educators.


9C2: Teaching data science to teenagers

Amelia McNamara   University of California, Los Angeles, United States
Mark Hansen   Columbia University, United States

Teaching data science to secondary school students might seem outlandish, given how few programs of instruction currently exist, even at the post-graduate level. However, teenagers are surrounded by data, and, in one form or another, view and manage data every day. Many teenagers carry smart phones, which already automatically store, retrieve and analyze data, but which can also be used to collect data. We will discuss challenges faced by the Mobilize project, an NSF-funded project that utilizes participatory sensing to teach data science to secondary school students (ages 15-17). In particular, we’ll talk about the challenges of including statistical computing as a natural and integrated part of the curriculum.


9C3: Integrating big data into the science curriculum

Daniel Kaplan   Macalester College, United States
Elizabeth Shoop   Macalester College, United States
Paul Overvoorde   Macalester College, United States

The contemporary practice of science widely involves “big data”: data with many cases and many variables and often with a distributed structure. The science curriculum, however, does not. Our approach to bringing big data to the science curriculum involves developing computational approaches that are powerful, engaging, and can be learned quickly with little background. We’re developing a 10-class-hour course (1 credit hour)—called Data and Computing Fundamentals (DCF)—to get students and faculty involved with big data, and able to tackle problems of realistic complexity.


Session 9D: E-learning, E-teaching and E-assessment in fully online, blended and open virtual web-based courses

9D1: Experiences teaching an introductory statistics MOOC

Alison Gibbs   University of Toronto, Canada

Can a Massive Open Online Course (MOOC) meet the needs of learners interested in introductory statistics? In the spring of 2013, Statistics: Making Sense of Data, an 8-week, non-calculus, introductory statistics course was taught on the Coursera platform, with over 60,000 learners enrolled while the course was taking place. The course was developed in response to a call for MOOCs focused on high-enrolment, low-success, general education courses. We describe the course, and summarize what we have learned so far from the data we collected about the learners, including what we know about who enrolled, their initial intentions, and how these are related to indicators of success in the course.


9D2: Statistical reasoning’s new look

John McGready   Johns Hopkins University, United States

“Statistical Reasoning in Public Health” is an introductory, two-quarter biostatistics class taught both online and in person to graduate students at the Johns Hopkins Bloomberg School of Public Health. The emphasis of the S sequence is on developing the statistical literacy of enrolled students. The course has been taught for 16 years, and for the past 10 years has had the same instructor for the online and on-campus section. After several partial makeovers, and a questionable choice of incorporating Stata code into the course materials, a complete makeover was undertaken in 2013. This makeover was not only designed to unify the course materials, update the data and literature examples, and to provide more feedback opportunities for students, but also to ready the course for its Coursera debut in January 2014.


9D3: Using Carnegie Mellon’s Open Learning Initiative (OLI) to support the teaching of introductory statistics: experiences, assessments and lesson learned

Oded Meyer   Georgetown University, United States
Marsha Lovette   Carnegie Mellon University, United States

As part of the Open Learning Initiative (OLI) project, Carnegie Mellon University was funded to develop a web-based introductory statistics course, designed to support students to effectively learn Statistics without an instructor. Numerous carefully designed studies have shown, however, that the course is most effective when used in a hybrid instructional model, i.e., face-to-face teaching plus online learning. We discuss the design features that make the OLI Statistics course “work,” present the most current assessment results, and describe different ways in which instructors have used the OLI materials to support their Statistics courses. In addition, we discuss how the OLI platform can be used by instructors to create their own materials.


9D4: A mobile web for enhancing statistics and mathematics education

Gunnar Stefansson   University of Iceland, Iceland
David Stern   Maseno University, Kenya
Victoria Mokua   Maseno University, Kenya
Jamie Lentin   Shuttle Thread, Manchester, United Kingdom
Anna Jonsdottir   University of Iceland, Iceland

A freely available educational application (a mobile website) will be presented. This provides access to educational material and drilling on selected topics within mathematics and statistics with an emphasis on tablets and mobile phones. The application adapts to the student’s performance, selecting from easy to difficult questions, or older material etc. These adaptations are based on statistical models and analyses of data from testing precursors of the system within several courses, from calculus and introductory statistics through multiple linear regression. The application can be used in both on-line and off-line modes. Results presented include analyses of how the internal algorithms relate to dropout rates, passing a course and general incremental improvement in knowledge during a semester.


Session 9E: Supporting teachers’ use of new statistics technology in their classrooms and development of their technological-pedagogical content knowledge

9E1: Data and Chance with FATHOM — teaching material for implementing computer-based stochastic courses

Andreas Prömmel   Gymnasium Ernestinum Gotha , Germany

Students often have inadequate primary intuitions when dealing with random phenomena at the beginning of stochastic courses. This experience gap between misleading intuitions and a sustainable perspective cannot be closed by physical random experiments alone, because the feasibility is limited in classrooms. In fact, the implementation of a statistical software tool like Fathom opens new opportunities for teaching and learning stochastics. According to our research, teachers need support by didactical materials for implementing computer-based stochastic courses. In our working group we have developed, evaluated and optimized our didactical concept through various empirical studies. Our textbook “Data and Chance with Fathom”, written for high school students in grades 6-12, is the result of this work. This paper gives an introduction to the outline and the structure of the textbook. Examples and teacher experiences are discussed.


9E2: Designing technology-rich learning environments for secondary teachers to explore and prepare to teach statistics

Sandra Madden   University of Massachusetts, Amherst, United States

This paper examines design characteristics refined over seven years through studies with pre- and in-service secondary mathematics teachers in the United States. Teachers simultaneously explore big statistical ideas while learning to use a dynamic statistical software tool that is new to them. Most study participants have taken at least one university statistics course, yet surveys, interviews, and content assessments indicate fragile understanding of statistical content, procedurally dominated learning environments, and little use of technology in courses. Important considerations when supporting these learners include respecting the statistical understanding that is present and connecting to it strategically. Productive design characteristics that have been developed and refined include: 1) focusing on big statistical ideas; 2) providing provocative tasks; 3) attending to dynamic technology scaffolding; and 4) implementing sustained intellectual press.


9E3: How a curriculum may develop technological statistical knowledge: a case of teachers examining relationships among variables using Fathom

Hollylynne Stohl Lee   North Carolina State University, United States
Jennifer Nickell   North Carolina State University, United States

We report results from 74 preservice secondary mathematics teachers’ [PSMTs] examination of a multivariate dataset using dynamic statistical software (Fathom™). The PSMTs were enrolled in similar courses at four US-based institutions, using the same curriculum materials. Adapting Lee & Hollebrands’ (2011) framework, we present an analysis of aspects of Statistical Knowledge [SK], Technological Statistical Knowledge [TSK], and Pedagogical Statistical Knowledge [PSK] the materials were designed to teach, and aspects of SK, TSK, and PSK evident in teachers’ work.


Session 9F: Technology for developing statistical thinking, reasoning, and literacy

9F1: Year six students’ reasoning about random “bunny hops” through the use of TinkerPlots and peer-to-peer dialogic interactions

Sibel Kazak   University of Exeter, United Kingdom
Rupert Wegerif   University of Exeter, United Kingdom
Taro Fujita   University of Exeter, United Kingdom

Students’ intuitions play a major role when making inference about uncertain events and they are often inconsistent with the accepted theoretical understanding of statistics and probability. These intuitions need to be challenged to develop more powerful, formal understandings of stochastic ideas. This paper describes how the reasoning of a pair of 11-year-old students develops as they conduct simulations of random “bunny hops” using TinkerPlots software for a large number of trials. We found that promoting dialogic talk, which involves questioning or challenging any claims and seeking reasons in response to challenges, facilitated students’ reasoning while they used TinkerPlots to test their conjectures and explain the outcomes. The study shows how the interaction of a software tool and talk between students can develop their understanding of statistical ideas and highlights some of the processes leading to conceptual change.


9F2: Technology for developing statistical thinking: a psychological perspective

Peter Sedlmeier   University of Chemnitz, Germany

Recent years have seen tremendous advancements and innovations in technology that also can be (and have been) used for teaching statistical thinking, reasoning, and literacy. However, these modern technological tools do not automatically yield better learning results than those achieved with traditional methods of instruction. More important than technology itself is a sound theoretical basis for building an effective technological tool. It is proposed that this theoretical basis should include aspects of usability, pedagogical aspects, and also content specific aspects. A brief description of the ACT tutoring systems program illustrates what a successful combination of these three aspects could look like. Then, the importance of the often neglected content specific aspects is demonstrated with examples from our own research. It is recommended that a special emphasis should be given to a systematic and sound evaluation of such technological tools.


9F3: Integrating technology in regular statistics courses and assessments of pre-service teachers

Andreas Eichler   University of Kassel, Germany

In 2011, along with a reform in the teacher education programme, the mathematics department of the University of Education, Freiburg, decided to integrate technology in all regular courses and assessments in mathematics and mathematics education for lower secondary pre-service teachers. In this report, an overview of the technology-integrated mathematics teacher education programme called IM2 is given. A further main focus is on the statistics courses as well as on assessments for these courses. Exemplarily, specific parts of the syllabus of statistics courses that are particularly connected with the use of technology are discussed.


Session 9G: Educational software for helping students learn statistics

9G1: An early look at rich learning analytics: statistics students playing “Markov”

Tim Erickson   Epistemological Engineering Oakland, United States
Mariel Triggs   Lick-Wilmerding High School, United States

In this paper we look at student work in a Data Game called “Markov,” designed to give students experience with conditional probability. The online game records all student moves and strategy-making. We look at these “rich learning analytics” to see what patterns of student behavior we can see and what types of conclusions we might draw. We will see that these data are an intriguing window into student thinking and behavior, with implications for both instruction and research.


9G2: Improving the attitudes of high school students towards statistics: an island-based approach

Minh Huynh   RMIT University, Australia
Anthony Bedford   RMIT University, Australia
James Baglin   RMIT University, Australia

This pilot study examined the use of an online virtual environment, known as the Island, at a secondary school level, and investigated its ability to help improve student attitudes towards statistics. The Island-based learning activity aimed to achieve this by focusing on the exploratory nature of the Island, allowing students to experience statistical problem-solving. Pilot data were obtained from 88 ninth and tenth grade students, who completed a learning activity that required them to conduct a statistical investigation on the Island. Before and after questionnaires were administered in conjunction with the Island activity. Results indicated that student attitudes towards statistics improved significantly after the Island activity. This paper summarizes these findings and discusses implications and strategies for active learning and improving student attitudes. In addition, the outcomes of this study will be utilized to plan follow-up studies.


9G3: Using on-line quizzes to help students learn probability and statistics

Mitchum Bock   University of Glasgow, United Kingdom
John McColl   University of Glasgow, United Kingdom

Online quizzes can be an effective and flexible means of helping learners develop key skills in probability and statistics. Quizzes give instant feedback, to help reinforce correct understanding and eliminate fundamental errors at an early stage in learning. We will describe our experience of designing and using quizzes with non-specialist and specialist students, on several different platforms including, most recently, Moodle. We describe Model Choice, a tool that helps students identify from a brief scenario the standard family of probability distributions they should work with to solve a problem. We will emphasize key design aspects of a successful quiz system, such as the importance of giving informative feedback to the learner. Using a standard platform, such as Moodle, is likely to require some compromise on design principles but building a stand-alone system to implement ideal design choices is very time-consuming.


Session 9H: Future trends for technology in statistics education (panel discussion)

Deborah Nolan   University of California at Berkeley, United States
Nicholas Horton   Amherst College, United States
William Finzer   Concord Consortium, United States
Webster West   Texas A&M University , United States