Schedule/Abstracts

Students frequently arrive in university psychology statistics courses without the requisite skill set for success and report they are extremely fearful of math and statistics. In addition, research shows that young adults report higher stress levels and lower ability to manage stress than other age groups (APA, 2014; 2019). The ability to cope with stressful challenges in statistics requires resourcefulness. Resourcefulness is the ability to self-manage when faced with goal blocking challenges. When applied to academics, it is the ability to problem solve through planning and evaluating alternatives, think positively about challenges, draw on resources to assist with challenge, and structure learning through self-consequence (Kennett & Reed, 2009). Literature shows that students who are resourceful achieve higher grades and are more persistent in their studies (Kennett & Reed, 2009; Reed et al., 2019) and courses that promote resourcefulness benefit all students (Collis & Reed, 2017). Integrating resourcefulness into undergraduate statistics requires thought around course content and pedagogical practice. Here, we highlight practices in statistics courses that promote student resourcefulness. The expectation is that promoting resourcefulness practices within the structure of a statistics course will assist students’ in managing their math phobic reactions to course content and lead to better outcomes in their understanding of the material. A facilitated discussion will allow attendees to highlight practices that help students to cope with the stress of their statistics classes.

Introductory statistics courses are regarded as gatekeepers for psychology majors, as students must successfully pass these courses to obtain their degree. Teaching these courses is challenging as they tend to have a high student-to-teaching staff ratio, making giving and receiving feedback difficult. Further, students’ attitudes toward statistics are largely thought to be negative, and students often report low intrinsic motivation in these courses, creating a taxing environment to teach in. Given that motivation is a critical predictor of academic achievement and engagement, assessing motivation in introductory statistics courses could provide insight into student experiences, and provide real-time informative feedback for instructors. In the current presentation, we describe how motivation assessments were implemented within an undergraduate introductory statistics course. Additionally, we discuss how to use these assessments to improve pedagogy while teaching based on reflections from the teaching team.

How can we infuse passion into our students’ experience of introductory statistics? How do we show our students that statistics matter, not just in the realm of psychology but in their own lives? Our presentation will introduce Passion-Driven Statistics, a project-based, introductory statistics curriculum that supports students in conducting original research with real world data. Funded by the National Science Foundation, our curriculum engages introductory statistics students in data-driven research with large, real world data sets that provide context and meaning to their understanding of statistics in the field of psychology. Students pose exciting research questions that matter to them personally and/or professionally, then discover how to answer those questions using real data and statistics. In the pursuit of these answers, students also learn basic statistical programming in a software platform of their instructors’ choosing (e.g., SAS, R, SPSS, Stata, Python). Importantly, students learn how to communicate their methods and results using the universal language of statistics. This skill is highlighted in the culminating component of the curriculum – a research poster session where students present their original work and have the opportunity to share their statistical journey with their peers and campus community. Our presentation will briefly introduce the background, history, and goals of Passion-Driven Statistics, as well as describe the basics of the actual curriculum. For instructors wishing to pursue the curriculum, all instructional materials are open-educational resources and information about accessing those materials will be provided. For more information, please visit https://passiondrivenstatistics.com/.

Cramer and Jackson (2006) report a sizeable association between the home game winner of the most recent Washington Redskins professional football game and the presidential election outcome. With only two exceptions, the events were perfectly linked since the 1936 inception of the football franchise. This paper offers an update to those results, now incorporating the 2008, 2012, and 2016 election results. This paper will offer instructors a useful vehicle to understanding correlations, to further show that correlation is not causation, but more importantly to illustrate that robust phenomena in the world may have no underlying cause. Furthermore, we believe that the relation between these two events represents a shortcoming in post-hoc reasoning, by trying to explain events after they have been observed.

One tactic I use to help maintain interest/participation during lecture and also reduce anxiety during one-on-one sessions with students in office hours is what I refer to as “Stats Improv”. I begin by saying, “OK, give me a topic. What are your interests, a hobby, favourite sport, anything”. Then I create a question on that topic, make up some numbers and do the analysis. This has multiple benefits. 1) It puts the student(s) at ease because suddenly the question seems familiar. 2) It shows them how solve a problem in real time. 3) It shows them how to catch errors. 4) It helps maintain interest because they are involved in the both the creation and completion of the work. This technique can take practice and does have its pitfalls; however, with practice and a bit of quick mental math (estimating) this technique can be very effective (and, at least in my opinion, a lot of fun).

Recent work has highlighted the unrepresentativeness of samples typically used for psychological research. Samples for studies making claims about human thought and behavior come primarily from countries that are Western, Educated, Industrialized, Rich and Democratic (WEIRD). Furthermore, many published studies are under-powered, which poses problems with both Type I and Type II errors and replication. We will present two different activities that can be used to engage undergraduate students in statistics courses with these issues. The first engages students with the problem of WEIRD samples by assigning each student to examine a single issue of a journal to categorize the samples for each study, explore what demographic information is presented, and examine any constraints on generality statements (COG) that are made formally (or informally as limitations). The second engages students in the process of conducting their own power analyses for an independent-groups t test using the G*Power freeware, with an optional extension exercise in which they examine the sample sizes and power for published studies.

During the course of the talk we will present examples for how to foster inclusion in a statistics and methods classroom based on experiences and the recommendations of other faculty. We will cover why even when content is not directly related to diversity, equity, or inclusion it is still an important consideration for classroom choices. We will also provide examples of student responses regarding the value of this work both in course evaluations and in other settings. Finally, we will give suggestions for first steps if this is new to your pedagogical approach.

We’ve known for years that we can do better than the lecture-lecture-test format. Relying on lectures to convey course content may be particularly inadequate for quantitative subjects. One effective approach to delivering statistics content in a classroom is the flipped class. Rather than consuming the content in class and then struggling alone to compete homework and prepare for the exam, students acquire the content at their own pace using assigned textbook readings and prepared videos. Class time is used to complete homework sets with the support of the instructor and teaching assistants and then review the content that proved most challenging. Homework assignments are completed online, so I can watch answers in real time. Students get each homework set a few days ahead of a class meeting, and it accompanies a reading assignment. The homework questions are designed to lead them through the reading. During each 50-minute class period, the first 30 minutes are reserved for working individually or in small groups with students who are struggling with specific questions. I reserve time at the end of each class period to review the new concepts, with special attention to any that were particularly difficult. Since I can see which questions are difficult and which students are struggling, I prepare each class in response to students’ needs. There are real individual differences in learners that can result from either aptitude or attitude. You may be losing some students while others figured out your point minutes ago and are now bored. In the flipped class, students who are struggling can have as much individual time as they need, while students who understand the content in the first pass need not spend time in class waiting for peers to catch up.

The flipped classroom (FC) inverts the location where learning traditionally takes place. The lecture and reading components take place at home while the active learning components occur in class with the support of instructors. Given its challenges as a required course for students with varying abilities, Introductory Statistics (IS) for social science majors may be a great candidate for the FC format. IS students tend to have high statistics-related anxiety combined with low interest in the subject that may be overcome by the dynamic structure of the FC. Studies suggest the FC in IS increases performance related outcomes compared to traditional lecture-based classrooms (LC). A meta-analysis was conducted to compare performance in IS courses for nonmath majors between FC and LC classrooms. Students in the FC had statistically higher mean final performance (final exam scores) over the mean LC performance, with an average difference of 6.9% (Hedge's g = 0.43). However, there was evidence of moderation by the presence of weekly in-class quizzes, such that the advantage of the FC was attenuated when comparing FC and LC classes that administered weekly quizzes. These results suggest that implementing the FC within the IS class at the undergraduate level may improve learning achievement, but future research is needed to examine the role of class quizzes.

Research has shown that flipped classes can be effective in teaching statistics, so this presentation will explore one successfully flipped Statistics II course in the large psychology program at York University. Discussion of the application of the flipped classroom will allow participants to explore a concrete example of this pedagogical strategy in action. The course director will discuss the inspiration behind the flipped classroom set up, the planning process and the execution of the class throughout the semester. Student end of term reflections as well as the course director's reflections will be presented.

Unlike many course changes, which tend to be minor tweaks, this was a complete overhaul to how content is taught (shifted from the more traditional mathematical elements to the non-mathematical elements of statistics), the format of the class (incorporating online elements and in-class experiential learning), and the frequency of evaluations (weekly submissions). Changes were targeted to reduce procrastination and anxiety surrounding learning statistics, motivate mastery of material, improve resiliency, and to foster support systems including accessing TAs and developing connections with peers. Importantly, the changes did not alter the course objectives or course content. Implications of the changes include an increase in class average while also lowering the number of withdrawal and failures. More importantly, course format changes have resulted in students referring to the statistics course as ‘fun,’ ‘enjoyable’ and ‘their favourite class.’ The typical atmosphere within a stats classroom (quietness filled with apprehension) was replaced with casual banter and occasional laughter and smiles. Due to COVID-19, the course shifted to fully online (synchronous and asynchronous) and adjustments to grading were implemented. Changes that will be incorporated in future editions of the course will be highlighted

Learning by correspondence or distance has a nearly 200-year history in postsecondary education, but interest in learning online began to accelerate only in the past 15 years, with the advent of the massive open online course format (MOOC). Government funding initiatives to create online courses have emerged over the past 10 years with modest uptake. My colleagues and I created a graduate-level statistics course through one such initiative and have been offering it annually since 2016 alongside other typical in-person courses. In March of 2020, professors at Canadian universities were almost universally forced online in response to the COVID-19 pandemic, and we quickly became all too familiar with online teaching. In this talk, I will review some key strengths and shortcomings of teaching statistics online to psychology students that I’ve gleaned from five years of teaching such courses to graduate students in psychology. I will consider the unique challenges of teaching statistics in relation to asynchronous versus synchronous format, leveraging software and online tools, student engagement, and testing. Finally, I will offer recommendations to strengthen the quality of online statistics education in psychology.

It has been known for decades that the assumptions of classical test statistics are rarely met in psychological research. However, most instructors completely ignore this fact and teach traditional methods that behave very poorly under realistic conditions. This leaves students with a false sense of security regarding the performance of popular measures. Introducing assumptions encourages the exploration of data before any descriptive or inferential technique is applied. Further, teaching robust statistics to early researchers encourages responsible research practices and results in more valid conclusions from data analyses.

In this presentation I will offer some general advice for teachers who wish to add Bayesian content to their existing lecture series. I will then demonstrate how the open-source statistical software program JASP can be used to teach students the basics of Bayesian statistics in a way that is interactive and fun.

This panel comprises a mix of undergraduate and graduate students from several universities. Individually and collectively these students will share their statistics education experiences. Among other topics, they will discuss what they found to be useful/helpful, what they found to be challenging, and they will also share their suggestions for instructors and future students taking statistics courses.

This session will be divided into two parts. In the first half I will discuss previous and ongoing research into students’ experiences with and attitudes toward using R in their statistics courses. Specifically, I examined attitudes toward statistics and attitudes toward the statistical software package R in both undergraduate and postgraduate students across the duration of a statistics course. Participants’ responses were analyzed using both quantitative and qualitative techniques. Results demonstrated that, on average, students in introductory level courses held neutral (not negative) attitudes, but students at higher study levels held somewhat positive attitudes towards R and statistics. While not all students enjoyed learning R, our findings demonstrate that many students enjoyed statistics or R, most students found statistics/R valuable and they generally reported feeling competent using software by the end of their course. These results challenge the argument that R is not suitable for undergraduate psychology students. The second half will involve a question and discussion period for students and instructors to voice their thoughts and experiences with incorporating statistical software into their courses.

Despite reproducibility concerns in science, an increased awareness of the prevalence of statistical errors in life sciences research, and even a tightening up of quantitative reporting standards in journals, statistical errors continue to permeate life sciences literature. What can we, as educators, do to better prepare our students to be consumers, and perhaps even producers, of quantitative research? In this talk, I will share experiences developing and teaching a joint course offered by the University of Toronto Department of Statistical Sciences and Human Biology Program that integrates introductory Statistics and research design instruction. I will also highlight key findings of a study that was conducted to explore students’ readiness to engage with Statistics in life sciences research upon completion of this course and reflect on how these results can help inform future course offerings and training initiatives to better prepare our students for quantitative research.

Reproducibility "crises" have become common in many areas of science in recent years, for a variety of reasons ranging from technical implementation issues through to poor training in proper statistical methodologies. In this talk, we will discuss some ways to integrate reproducibility and replicability of statistical analyses into introductory statistics courses, in particular those aimed at complete novices. Developing well-founded work practices from the beginning allows students to build upon strong foundations, preventing a number of common analysis pathway-based issues from ever arising. The discussion will be centered around the use of RStudio's R Markdown document creation tools, as well as other utilities easily used from within the RStudio ecosystem.