#### Focus on Posters

P1: Monday 9th, 17:30-18:30P2: Tuesday 10th, 17:30-18:30

Note: only the presenter is

shown below

## Posters

### Convenors

- Elizabeth Fry (United States)
- Jennifer Noll (United States)
- Richard Wilson (Australia)

### Poster session P2 (Tuesday 10th, 17:30-18:30, Foyer)

2 | Use of Interactive Apps in Teaching Bayesian Statistics | Haydar Demirhan |

Use of statistical software is essential to the teaching and learning of Bayesian statistics. Effective use of statistical technologies, which help transcend the static pages of a textbook, have a great potential to make Bayesian theory and concepts more accessible through effective, dynamic, and interactive visualisation. This poster will present the development of specialised apps for teaching Bayesian statistics using Shiny, an open source web application framework for R. The apps were designed to dynamically visualise key Bayesian concepts covered in a first course. The apps allowed the instructor to develop students’ understanding through experimentation, whereby the instructor or students could vary input parameters (e.g. alter a prior distribution) and visualise the resulting effect (e.g. posterior distribution). | ||

4 | Using concept questions to teach statistical terms | John Blake |

Statistics, like any academic discipline, comes with its own set of technical terms. A number of these terms, such as normal and significant, are polysemic resulting in learners of statistics assigning the inappropriate lay meanings rather than statistical word sense. This presentation shows how concept-check questions can be used to convey and check meaning in a concise and clear manner. The interactive nature of concept questions enables active learning and engages learners in critical semantic analysis. Given the centrality of the meaning of key statistical terminology, conveying the meanings without distortion to students is essential. The denotative and connotative meanings of a lexical set of statistical terminology will be used as a vehicle to demonstrate the effectiveness of this teaching technique. | ||

6 | Attitudes toward Statistics as moderator variables in the evaluation of activating methods in lectures on Statistics | Florian Berens |

Many studies report heterogeneous attitudes toward statistics among bachelor students of different subjects with a tendency to negative attitudes toward statistics. The format of a lecture cannot adequately counteract this heterogeneity. Recent studies offer some suggestions how to actively integrate students in lectures. Particularly in the field of mobile learning, instruments are provided that are suitable for involving high amounts of students. Although evaluations show positive first results about these instruments it is not clear whether these instruments are particularly suitable for students with negative domain-specific attitudes. For the study presented here, n=315 students were asked to evaluate a unit with a smartphone-based clicker system. As a comparison, the technology-free Think-Pair-Share method was examined. The results focus on the moderating effect of attitudes toward statistics. | ||

10 | A robust method to identify the statistical learning abilities of students | Hung Hung |

Identification of the statistical learning abilities of students in a certain population is useful for statistics educators. The success of this task can not only help improve the performance of statistical learning for students, but can also help reduce the loading of educators. While we have collected many characteristics for a student that are useful for the identification of his/her learning ability, the process of constructing the prediction model can suffer the problem of outliers that can make the identification results totally biased. In this work, we propose a robust procedure for the identification of students that are less affected by the presence of outliers. The newly developed identification procedure is shown to well reflect the learning ability of a student. | ||

12 | Molecular modeling and statistical software in modern organic chemistry teaching – Step forward to modernization of undergraduate studies at engineering faculties | Gordana Cetkovic |

The application of computers and software is becoming inevitable in modern study curriculum at majority of faculties in the World. The application of computers in organic chemistry teaching has significantly contributed to the solution of the problems related to the presentation of statistical approaches in chemistry in a simple and understandable way. It also has encouraged the creativity and innovation of both teachers and students. The present study describes the analysis of the possibilities of application of different statistical software in basic and advanced organic chemistry teaching on undergraduate level at engineering faculties, the easy-to-understand software approaches in modern organic chemistry and visualization methods of molecular structures. This study also presents the analysis and interpretation of possible outcomes of these approaches in teaching. | ||

14 | Teaching Graduate Students Sample Size Planning by Using R | Wei-ming Luh |

Instead of statistical test of equality of means, the test of equivalence of means has become more popular for clinical trials and pharmaceutical science, and it has recently been expanded into broader applications in many fields. However, the sample sizes planning using conventional techniques found in the literature on this topic have usually been under-valued with less statistical power than is required. For better preparing graduate students to design experiments and allocate required sample sizes for the experiments, the present study develops a new visualized method by using R to provide distinctive insights into effect sizes, the statistical power, and/or pre-specified equivalence boundaries to enhance comprehension of sample size planning. | ||

16 | Another Look at the Box Model | Dennis Sun |

We revisit the box model, an analogy introduced by Freedman et al. (1978) to teach sampling distributions and inference. The idea is to represent a random phenomenon in terms of random draws of tickets from a box. In this way, random sampling from a population can be modeled in the same way as familiar phenomena like coin-tossing and card-shuffling. However, Freedman et al. present box models only as a thought experiment; calculations are still done using normal approximations. We argue that a simulation-based approach to box models correctly places the emphasis on the modeling rather than the calculations. Furthermore, we demonstrate how the box model is useful beyond an introductory course by showing how it can clarify discrete distributions in a probability course. | ||

20 | Launching a statistical enquiry: Posing statistically worthwhile questions | Aisling Leavy |

A recent approach to statistics education is situating the teaching and learning of statistics within cycles of statistical inquiry. Learners pose questions, plan and collect data, represent, analyse and interpret data. We focus on the first step – the preparation of prospective teachers to pose statistical questions. Posing worthwhile statistical questions is a critical step as they inform the types of data collected, determine the representations used and influence the interpretations that can be made. We report on an investigation of prospective elementary teachers in Ireland (n=200) and Germany (n=50) as they design statistical questions. Support was provided through tutorials, peer-feedback and expert-feedback. We describe the features of statistical questions posed, identify obstacles and difficulties experienced and evaluate the effectives of both peer and expert feedback. | ||

22 | Technology in bi-dimensional statistics in High School textbooks | María M. Gea |

The technology is recommended in the teaching of statistics and a resource for teachers is usually the textbook. The aim of this paper is to analyze the directions on the use of technology included in the Spanish textbooks directed to high school students in the topic of correlation and regression. We analysed eight textbooks in the Mathematics modality using content analysis in different analyses: a) the use of technology in the problem proposed and related procedures; 2) the references to Internet resources; and 3) the CD included in most of the books. We found variability in the use of technology, generally related to the use of calculators or to the spreadsheet and not to data sets that can be used in projects or to simulators. | ||

24 | Student attitudes towards statistics at a South African university | Nombuso Zondo |

This study investigates the relationship between student attitudes towards Statistics and their performance in the Statistics course. We adopted the ‘SATS-36’ survey questionnaire to assess the attitudes of students towards Statistics. We used exploratory factor analysis to group the attitude responses according to factor loadings as was done in other studies using ‘SATS-36’. Moreover, we examined whether the attitudes to Statistics locally are related to demographic attributes, field of employment and academic exposure to Statistics. | ||

26 | Consideration on Students’ Statistics Report about Social Justice | Takashi Nakanishi |

I applied project teaching to graduation thesis (12th grade) on statistical literacy. The guideline was the following: (1) Report theme, (2) Motivation, (3) Method, (4) Arrangement of documents, (5) Consideration, (6) Reflections and comments. One student wrote the following report: Degrees of "cutting corners" of six television stations rebroadcasting programs in Japan. According to one viewer, the NHK educational broadcast had too many rebroadcasts about programs for children. The viewer pointed out that it looked like injustice. The student analyzed this objection based on ‘justice and structure (separated values)’. He also analyzed it by another viewpoint based on ‘caring and human connection (connected values)’. According to the public educators’view, it is interpreted that the integration of these two moral values was shown. | ||

30 | Datascientist education by e-Learning system | Kenichi Hirano |

In Febrary 14, 2017, i's Factory, Co. Ltd. offered datascientist education course by e-learning system (Japanese only). After that, over 100 students (businessperson, school students, etc.) already have learned our course. In Japan, there are the datascientist education course by schooling. But only we are providing the datascientist education course by e-learning system (Online). So, we make tuition fee reasonable (19,800 yen) and you can learn our course anytime, anywhere by accessing to the Internet with PC, smartphone, and tablet. | ||

34 | Implementing Inverted Instruction in Undergraduate Introductory Statistics Courses | Anushka Karkelanova |

This study was about an experimental design that compared different instructional styles in teaching introductory statistics to undergraduate students (N=270). Inverted classroom was the treatment, and the treatment effects were assessed against traditional classroom as control. Inverted classroom refers to the instructional practice where events that traditionally take place inside of the classroom now take place outside of the classroom and vice versa. Traditional classroom refers to statistics classes that are taught using the lecture method. This study aimed to provide insight into the effectiveness of inverted instruction and identify factors that facilitate or hinder this effectiveness. Given that inverted instruction is becoming popular, this study has a broader intellectual interest throughout higher education. This presentation would document this intervention and present preliminary results. | ||

36 | Young learners’ reasoning with informal statistical models and modelling | Michal Dvir |

There has been a growing interest within the statistics education research community in statistical models – an object – and statistical modeling – the complementary process. Research has generally focused on older learners, such as high school and post-secondary school students, rather than learners at the primary school level. Our goal is to provide a framework that both describes and provides a tool for analyzing the reasoning that accompanies young students’ informal statistical modeling. This poster provides a brief description of some fundamental definitions, followed by our framework for young learners’ reasoning with informal statistical models and modeling. Our framework identifies three separate, but not independent, modeling sub-processes: the conjecture, data and comparison modeling processes. An illustrative example of its dual usefulness is also provided. | ||

40 | Latent trait models highlight deficits in student understanding. | Robert Quinn |

Final exams are typically set in order to assess course content knowledge and to provide evidence that students have achieved at least some minimum level of competence in the learning outcomes. Exam papers are typically archived on completion but they contain abundant information that can highlight topics where there is either adequacy or a shortfall in understanding. Final exam papers from a first-year statistics unit were randomly selected. Using item response theory models, the probability of a correct response to each of 56 question items was obtained as a function of item difficult and student ability. Item difficulties were extracted to enable the ranking of question items from least to most difficult. Results of modelling and the impact on future teaching will be presented. | ||

42 | Rethinking Exams in Large-Lecture Statistics Courses | Melissa Pittard |

Exams are both time-consuming for instructors to grade and imperfect at assessing the depth of student knowledge. Exams that have to be completed in a short amount of time force students to quickly read the problems, determine the solution, and convey those solutions coherently. This artificial time constraint can be problematic for students. Why not think about the examination process differently? The author assigned a portion of some exams to groups and asked that these be completed outside of the classroom. Student often did the group portion by presenting solutions via video or as a result of using software to analyze their data. Discussion of these and other approaches along with some quantitative results will be presented. | ||

44 | StatHand: An application to support students’ statistical decision making | Peter Allen |

Quantitative research methods underpin professional competence across many disciplines. Despite this, many students struggle with the process of selecting appropriate statistics for common research questions, hypotheses and data types. StatHand (see https://stathand.net) is a free cross-platform application that aids this process by prompting students to focus systematically on each structural feature of their research problem. Student focus groups (N = 25) and instructor interviews (N = 9) support the subjective appeal and usability of StatHand. Furthermore, an experimental evaluation (N = 215) found that StatHand promotes decision making accuracy, reduces cognitive load, and is instructionally efficient relative to a range of alternative statistical decision making aids. StatHand can be readily integrated into a variety of classroom learning activities. | ||

46 | The Effectiveness of Using ICT Technology in Statistics Education | Makoto Handa |

In mathematics education, utilization of ICT should aim to teach students how to discover new evidence present in large amounts of data through organization, analysis, and interpretation and how to understand trends within data. In order to achieve these goals, statistics instruction should not just have students calculate statistical values, rather it is also necessary to introduce teaching materials that have students consider data’s meaning and value from its context. I have found that having students present the results of their research to the class is an effective way to increase their efficiency in organizing, analyzing, and interpreting data. | ||

48 | 100% satisfactioned medical statistics seminar that healthcare workers really need at present | Keiko Aoishi |

Healthcare workers feel difficulty in statistical analysis of data. Approximately 70% of healthcare workers were working statistically in the routine work, of which about 30% were doing analyzes other than simple counting. In addition, only one responded "I am able to analyze". Therefore, we held a monthly seminar about data analysis method necessary for medical staff in the business. The purposes of the participants were "Analysis method using Excel", "Visualization of data, such as table and graph creation", etc. We emphasize analytical methods that can be completed only with Excel without using statistical software, and teach how to create comprehensible materials simply by devising from everyday work. Satisfaction from participants is high, and we will report the details from the seminar. | ||

50 | An examination of computer versus tactile simulations for teaching sampling distributions in introductory Statistics | Wendy Rummerfield |

Sampling distributions are a notoriously difficult topic for students in post-secondary introductory statistics courses. Much of the literature suggests utilizing computer simulation methods (CSMs) and, despite few empirical studies, incorporating hands-on simulation activities prior to CSMs as a pedagogical tool in teaching sampling distributions. In our pseudo-experiment performed at a large research university, we randomly assigned discussion sections to sequences of sampling distribution activities either using CSMs preceded by hands-on simulation activities, or CSMs alone. We found moderate evidence of a positive effect of hands-on simulations on exam scores. However, the analysis of the sampling distribution-specific exam questions, which vary in type (e.g., multiple choice or free response) across exams, requires the development of a new statistical methodology for longitudinal data of differing response distributions. | ||

52 | Assessing High School Students' Statistical Literacy about the Measures of Central Tendency | Yanqing Ding |

Statistics is increasingly considered an important outcome of schooling. To explore Chinese high school students’ statistical literacy about the measures of central tendency, this study examined a sample of eighty-three students’ responses to a two-tier fifteen-item instrument. Rasch analysis demonstrated adequate reliability and validity of this instrument to measure student performance. Considering Rasch difficulty estimates, classical difficulty estimates, and the most popular distractor, this study identified the students’ strength in calculating median and weighted mean. However, they struggled with understanding the mean of a random variable, distinguishing between sample mean and population mean, and applying the measures of central tendency in diverse contexts. Making connections between the statistics concepts appeared a challenge for the students. Implications for statistics education are discussed. | ||

58 | Using Real Data on Beliefs about Maths to improve Primary Teachers Students' Capacities in Statistic Education | Carlos Carrión-Pérez |

The Survey for Mathematics Pedagogy and General Pedagogy Educators, developed by the International Association for the Evaluation of Educational Achievement - Teacher Education and Development Study in Mathematics in 2008, was initially passed to Mathematics, Mathematics Educators, but General Educators too. We have extended it to students also; we used only parts A and K and had got a sample of 135 tests with 10 questions about background and 34 questions on beliefs about Mathematics. Results of questionnaires were brought to students during several sessions on Statistic Education. Those students, who are being preparing to become Primary Teachers, did a complete statistic study from the original results, and gave special attention to making, reading and translating graphs. The whole process is presented in this work. | ||

62 | Views of Benguet State University Employees on the Role of Statistics in the Workplace | Marycel Sajise |

This study aimed to quantify the influence of subjects taught, educational background and length of service on viewing the importance and role of statistics in the workplace. Also this study aimed to measure the commonly used statistical tools/concepts among Benguet State University employees. Survey questionnaires were administered to randomly faculty and staff using stratified random sampling. Results showed that the respondents do not differ on their views of the importance and role of statistics when educational attainment was considered. However, respondents’ views differed across subjects taught and length of service. Also, the level of importance of statistics and its role in the workplace was above the moderate level. Among the topics/concepts that were considered widely useful and familiar among employees are mostly under descriptive statistics. | ||

68 | Statistical Literacy through Guided Block Play: An Exploratory Multiple Case Study | Robert Giebitz |

Misconceptions, anxiety, and negative attitudes impair adult learning of statistics. Courses in statistics often fail to impart conceptual understanding. Current guidelines suggest statistics learning begin with inquiry in primary school. Yet statistics learning might begin even earlier. Learning about distributions holistically through play might help convey the idea that a data set is an aggregate with emergent properties of shape, spread, and center, and help prevent misconceptions and anxiety in later years. What can a preliterate child learn about a frequency distribution? In this study, children as young as five manipulated blocks under the guidance of a tutor and created “embodied” frequency distributions. They found descriptive statistics, made X-plots and box plots. Only after sensorimotor experience with embodied statistical concepts did they perform statistical inquiry. | ||

70 | Using Google Docs and Sheets to Design and Collect Data for Classroom Experiments | Wendine Bolon |

The 2016 Guidelines for Assessment and Instruction in Statistics Education recommends the college professors use active learning pedagogies with real world data while using technology for data exploration and analysis. Pedagogies and technologies that actively engage students and use real world data have been shown to improve student engagement and achievement. Experiments offer an opportunity to actively engage students, generate real world data to analyze and organically integrate technology in the classroom. This poster describes some uses for Google Docs and Sheets that allow in-class collaboration in the design and implementation of experiments, as well as data collection and analysis. |