Science Journal generic logo
science-journal

Reinventing Statistics Education

1 December 2015
Image
Dennis Pear
Dennis pearl

Dennis Pearl has a CAUSE, and his goal is to improve statistics education and teaching. CAUSE, which is the Consortium for the Advancement of Undergraduate Statistics Education, is a national organization whose mission is to support and advance undergraduate statistics education through resources, professional development, outreach, and research. Pearl, who is a professor of statistics at Penn State, also serves as the director of CAUSE, where he has been working to develop a redesigned foundation for statistics education across the country.

“To effectively impact all students, there must be a way to personalize each instructor's teaching possibilities and each student's educational experience,” Pearl said. “Support for statistics education at Penn State and across the country is high right now, and technology is continually improving and becoming more accessible. These elements are making it possible for that type of personalized education to be feasible.”

Relevant Resources

Reinventing Statistics Education XKCD Comic
A CARTOON FROM THE CAUSEWEB.ORG RESOURCES. It is suitable for a course website that makes use of a boxplot to display an outlier and also uses the term "statistically significant" in its punch line. The cartoon is free to use in the classroom and on course web sites under a creative commons attribution-non-commercial 2.5 license. Credit: The cartoon is number 539 from the webcomic series at xkcd.com created by Randall Munroe

CAUSE has served as a catalyst for changing the way educators teach statistics by providing free resources to personalize pedagogy and content for different learners. Through the CAUSE website (causeweb.org), webinars, workshops, an electronic conference, as well as an in-person conference, the organization, and Pearl, have been making every effort to improve the way statistics is taught and learned.

The CAUSE website offers a multitude of resources to statistics educators worldwide. The resources are free to use and have everything from recordings of webinars on teaching methods and activities, to web app-based activities, data sets, and fun resources, such as videos, cartoons, and songs that can help facilitate learning.

Webinars have served as a way to help educate statistics educators without them needing to leave their desks. The webinars, which now come to dozens per year, are broken down into three areas: activities series, teaching and learning series, and research series. One from each series is offered each month and is recorded so that they can be viewed on demand any time. Industry partners, such as Minitab, Pearson Higher Education, W.H. Freeman Publishing, and SAS, help to cover the cost so that educators can access the webinars free of charge.  “Webinars offer a great way to present the newest innovations in teaching and learning statistics and CAUSE is grateful for the support from volunteers and sponsors who make those efforts a reality,” Pearl said.

CAUSE also hosts a biennial Electronic Conference On Teaching Statistics (eCOTS), which occurs on even number years. The next, to be held in May 2016, will focus on “Changing with Technology.” The conference, which is chaired by Kari Lock Morgan, assistant professor of statistics at Penn State, is intended to spark new ideas for how educators can change with technology and provide a virtual meeting space for educators to engage with and learn from each other. eCOTS 2016 will feature keynote speakers, virtual panel discussions, breakout sessions, workshops, virtual posters, and affiliated mini-conferences taking place around the country.

In addition, CAUSE hosts a biennial United States Conference On Teaching Statistics (USCOTS), which is held on odd number years, and is well attended by the statistics educator community. Last held in May 2015 at Penn State and attended by about 450 people, the conference featured plenary sessions, active breakout sessions, poster sessions, and a dozen associated workshops disseminating the work of NSF-funded projects in statistics education. Both the eCOTS and USCOTS serve as a mechanism to invigorate the statistics education community on a regular basis and also provide the infrastructure for the dissemination and sharing of information from educators on teaching experiences, research results, and hot topics, such as data science.

Making Statistics Fun

Pearl’s personal education research currently focuses on the latter aspect of the available resources: for teaching applied probability and for using fun resources in teaching statistics. “Project UPLIFT (Universal Portability of Learning Increased by Fun Teaching) questioned whether or not the use of cartoons and songs would improve student learning and decrease anxiety,” Pearl said. 

comic of Capture Recapture
A CARTOON TO TEACH ABOUT THE CAPTURE-RECAPTURE
METHOD. This is part of a three cartoon set from Dr.
Weaver that took first place in the cartoon category of
the 2007 A-Mu-sing competition. It is free to use in the
classroom and on course web sites. Credit: Cartoon by John Landers (www.landers.co.uk) based on an idea and sketch from Sheila O. Weaver (University of Vermont).

CAUSEweb.org hosts the largest collection of fun resources for college statistics teachers, which includes cartoons, jokes, quotes, songs, poems, word puzzles, magic tricks, and videos. Pearl and his colleagues at University of Texas at El Paso and Georgia Perimeter College assessed the materials throughout the three-year study, by observing students in three urban settings and interviewing them to gather their attitudes towards introductory statistics. In one experiment, all students took a pretest and post-test measuring their anxiety about statistics. Students were randomly assigned to either the experimental group, where they were exposed to cartoons or songs inserted into short content items, or the control group, where they only received the content. The researchers analyzed the data to see if students who received the extra fun content would perform better on related embedded multiple-choice exam questions, or experience a greater decrease in statistics anxiety over a semester.

The data showed that song items had a higher percent of correct answers among students who viewed the lesson in conjunction with the song compared with the control students who saw the lesson alone. The use of cartoons did not show any differences between groups on test item performance – but, along with songs, may have helped reduced student anxiety.

Building on their findings from project UPLIFT, Pearl and colleagues are beginning a new project, Project SMILES (Student-Made Interactive Learning with Educational Songs), which was just approved for funding by the National Science Foundation. This project will develop and experimentally test an innovation in online learning where students create a song by filling in key words (like Mad Libs) associated with a learning objective.  The songs are played back though synthetic voice technologies and can be shared between students. “I’m really excited about how these interactive songs are being designed to challenge students to construct examples and/or context thereby fostering statistical literacy and reasoning skills,” Pearl said.

While Pearl primarily focuses on cultivating the resources, professional development, outreach, and research available to statistics educators across the country, Kari Lock Morgan is specifically focusing on improving how Penn State students learn statistics by redesigning introductory courses and refining teaching methods to make them more effective.

Redesigning Statistics Education

Image
Kari Lock Morgan
Kari Lock Morgan

For many, statistics stirs up a mental picture of memorizing mathematical algorithms and computations but never really understanding them. Students have been taught through plug and chug type courses and retained information long enough for an exam, then disregarded it and moved on to the next topic. Although the content and teaching methods have not changed much over the last few decades, Lock Morgan is on a mission to change the perception, as well as the content and way that students learn statistics.

Traditionally, statistics has been taught such that students memorize mathematical equations, theory, and distributions. Although this method has worked for statistics majors and those who are very motivated and very interested in learning about the topic, it has been very difficult for students who need basic statistics knowledge on their way to another degree. Many non-statistics students develop only superficial understanding and do not retain much of the information learned in class. These students can repeat theories and solve basic problems, but most are doing it through conditioning and memorization. Students may not actually understand the numbers they are calculating.  Lock Morgan, who was named a 2015 Eberly College of Science Center for Excellence in Science Education Tombros Fellow, has dedicated her career thus far to teaching statistics in a more intuitive way so that students comprehend the information and can apply it. “I am grateful that the fellowship allows me to focus more of my time where my passion truly lies; on making statistics more about conceptual understanding and solving real-world problems,” Lock Morgan said.

For her fellowship project, Lock Morgan chose to overhaul STAT 250: Introduction to Biostatistics. Rather than relying on formulas and theoretical distributions, she is introducing inference via simulation methods, which include bootstrap confidence intervals and randomization hypothesis tests. With today’s technology, statistics students are able to utilize computing to get answers in a simpler and more intuitive way. This provides students with a more general way to approach statistical problem, focusing on the big picture and helping students apply the methods to real world problems. It also builds conceptual understanding so students better understand their results, and can better interpret statistics they may encounter in other situations. The changes she is working on are in both content and pedagogy. “Incorporating real-life scenarios into teaching encourages the students to relate to the material and therefore be more engaged in class and the learning process. Students who have a solid understanding of the foundational aspects of this method are able to effectively collect data, analyze data, and interpret conclusions drawn from data and see the real-world value of statistics,” Lock Morgan said.

She aims to help students develop a strong intuitive understanding of inference through randomization methods. Once this ground work is established and students have a conceptual understanding and appreciation for the results, they can then compute using the more traditional methods of statistics, including t-tests and chi-square tests.
 

A Family Affair 

Lock Morgan family image
A FAMILY OF STATISTICIANS. Pictured below are Patti Frazer Lock, Robin H. Lock, Dennis F. Lock, Kari Lock Morgan, and Eric F. Lock Credit: Photo from lock5stat.com

Lock Morgan is not alone in her quest to improve statistics education; her father, mother, and two brothers are also statisticians who have a passion for statistics and education. As a team, the five wrote Statistics: Unlocking the Power of Data, a textbook for introductory statistics. “It may be unusual to write a book with your entire family, but for us it has worked really well, probably because we all like each other a lot.  We generally agree on big picture things, but each bring our own unique perspectives and opinions on the finer details, which ultimately improves the final product,” Lock Morgan said. The book has been heavily adopted in academia; teachers across the county are utilizing the innovative approaches in the text to help teach more effectively and assist students in actually learning and using the material.

Additionally, the Locks have developed a website, lock5stat.com, that provides resources to support their textbook and statistics education as a whole. Their online software, StatKey, a collection of web-based statistics applications, accompanies the textbook. Their website also provides data sets in several formats that educators can use in their classrooms. These datasets help teachers provide real-world examples to help students see the practical applications of statistics. As of July, StatKey has had 8 million page views, with over 1,000 sessions per day, and has been used in 138 countries.

Lock Morgan herself uses the datasets in her classrooms and shows students how statistics applies in real life. As part of the Tombros Fellowship, Lock Morgan has organized each class and lab to focus on answering a relevant scientific question, hoping to emphasize that statistics is an important tool in science.  Several of these datasets and questions come from researchers in the biological sciences at Penn State, in an effort to give students an idea for how statistics may be used in their own departments.  Additionally, she embraces innovative teaching methods, such as active learning and the use of learning assistants, to get students more engaged in class and the learning process. In STAT 250, Lock Morgan requires students to use the i>clicker system. The use of clickers in the classroom encourages more class interaction and discussion, leading to better understanding and information retention.

Revising course content and changing pedagogical strategies not only helps with conceptual understanding, but improves student problem-solving abilities and their transfer of learning. Trying different approaches often leads to better understanding and helps students learn better.
 

Teamwork Approach

Lock Morgan, Pearl, and Matt Beckman, who will be joining the statistics faculty in January 2016, all recognize the need train high school teachers and better prepare students to use and understand statistics before they go to college. “Matt will provide a terrific boost to Penn State’s growing stature as a center for statistics education expertise,” said David Hunter, head of the Department of Statistics. “Even before officially starting at Penn State, he has already gotten involved as a co-principal investigator on a statistics education grant proposal with Kari and Dennis.”

Beckman is a Penn State graduate himself, having earned a bachelor’s degree in mathematics in 2006, along with a Pennsylvania certification in secondary mathematics teaching.  He later earned a master’s degree in statistics from the University of Minnesota, and since that time he has been working as a practicing biostatistician and statistical consultant, while simultaneously completing a doctorate in statistics education, working with two leading experts in that field. 

Together, the statistics trio is working on a NSF proposal to improve outreach and teacher training. Their goal is to help students be better prepared for statistics by improving how it is taught in high schools, while also refining the content and taking advantage of computing technology.  The Common Core State Standards in Mathematics recognize both the increased importance of statistics and the intuitive benefit of introducing inference via the simulation methods enabled by computers, so the team aims to prepare teachers to teach more statistics and teach it in this conceptual way, particularly because many teachers will not have seen this approach, even if they have taken a statistics courses.  “It’s fantastic that high schools are aiming to teach more statistics, and very exciting that the Common Core is promoting the use of simulation methods.  We want to do what we can to help make this happen!” Lock Morgan said.

It’s no surprise that with passionate and motivated faculty, and resourceful, dedicated staff, the Department of Statistics is becoming a well-known leader in statistics education. “Statistical literacy has always been a vital component of a broad scientific education, and it is all the more so in today's era of 'big data'. This fact makes the educational mission of our department more important than ever; and Penn State statistics has such a huge footprint not only in Pennsylvania but beyond the Commonwealth through our World Campus programs that we are well-positioned to play a strong leadership role in the practice of statistics education. In this context, I’m especially proud of the work we’re doing to continually improve the teaching we do in the statistics department. Adding experts like Kari, Dennis, and soon, Matt to our faculty not only enhances our day-to-day teaching but ensures that Penn State will remain at the cutting edge of research in statistics education,” said Hunter.

The outstanding research, resource, and educational improvements would not be possible without support and collaboration. Staff members have been essential in helping the faculty, department, and even CAUSE function on a daily basis.  Hunter, Pearl, Lock Morgan, and the rest of the statistics faculty recognize the important roles that the statistics staff perform in helping them to achieve their objectives and in elevating the department stature.

“Moving CAUSE from Dennis’s previous home of Ohio State here to Penn State in 2014 was a highly complex task.  We’ve hired two additional staff members, half of whose time is devoted to CAUSE-related tasks. Lorey Burghard joined our department in December 2014 and serves as CAUSE’s program coordinator, and Bob Carey came on board in May 2015 to support CAUSE’s IT presence. Meanwhile, Kathy Smith, who was already on our staff, took on a huge additional workload almost from the day Dennis joined Penn State to tackle the Herculean task of moving the web hosting for CAUSE from Ohio State to Penn State. There’s so much overlap now between the statistics department’s various missions and CAUSE’s mission that our newly enlarged staff has worked seamlessly into the life of the department. So we have CAUSE to thank not only for increased visibility in the statistics education community but for some excellent new staff hires as well!” Hunter noted.

Teamwork has a been a key compon­ent for the successes that the statistics department has achieved. By using an inclusive approach, the department has been able to make great strides in improving statistics education through collaboration in resources, teaching, and learning.