Customer Story

Exercising the body and mind 

At Nanjing Sport Institute, a culture of statistical thinking permeates research and learning in epidemiology and exercise science

Nanjing Sport Institute

ChallengeIn a world where scientific scholarship is ever evolving, students must develop the skills to solve problems, not just understand established facts. “The most important thing,” says Professor Kai Xu, “is for students to develop a curiosity to explore knowledge and keep learning throughout their careers. Statistical thinking is a huge part of that.”
SolutionStudents at Nanjing Sport Institute now have the opportunity to use JMP® and JMP Pro in the classroom to analyze and evaluate survey and experimental data sets. Because the software encourages trial and error, faculty believe it helps to promote statistical thinking and cultivate an appreciation for the scientific method. 
ResultsWith JMP, students and researchers alike can quickly acquire skills in data exploration, cleaning, visualization, modeling and more. But beyond the technical skills, students gain experience using data to solve problems – a life skill advantageous to many and diverse career paths.

Nestled on the Yangtze River not far from Shanghai, Nanjing Sports Institute (NSI) is known as “the cradle of world champions.” Founded in 1956, the institute has produced more than 100 Olympic gold medalists, including swimmer Li Lin, gymnast Huang Xu and badminton players Ge Fei and Gu Jun.

Beyond its reputation as an elite athletic training ground, NSI is also one of China’s leading research institutions in the fields of sports medicine and exercise science. A recent institutional focus on “rejuvenating science and technology” has resulted in research and educational partnerships with prominent institutions in the United States, Japan and Germany, and in exchange programs with counterparts in more than 40 nations worldwide.

Just as NSI’s athletes learn to push the boundaries of physical fitness, the institute’s academic program challenges students to hone a skillset that will enable them to keep up with the march of scientific innovation.

“Modern scientific progress is moving very fast,” explains Xu Kai, Professor of Exercise Physiology at NSI. “Students may feel that any rote knowledge they learn in school won’t be useful in their later working careers. But statistical problem solving is different.”

Critical thinking and data literacy, he continues, are transferable competencies that will help graduates succeed, regardless of whether they become physician to one of China’s elite Olympic teams – or start a career outside of exercise science. “The most important thing is for students to develop a curiosity to explore knowledge and keep learning throughout their careers. Statistical thinking is a huge part of that.” 

Bringing scientific research into the classroom

Xu is an expert in kinesiology, or the study of the mechanics of human body movement. Among other things, his research investigates the epidemiological connection between physical activity and health status among adolescents. The goal of this work is to determine the kind of daily exercise regimen young people need to better their physical health and fitness.

“We can see that not only in China but around the world, young people aren’t getting enough physical activity to maintain healthy levels of muscular strength and cardiopulmonary function,” Xu explains.

To better understand this phenomenon, Xu recruits middle school students to participate in various studies of physical activity. Participants wear acceleration sensors which capture information about their movement and heart rate as they walk through their daily routines of school, exercise and home life. The resulting data set forms the basis for much of the work that Xu and his graduate students spend their time on, and it has also found its way into the undergraduate classroom.

For Xu, there is little barrier between research and teaching. That is, of course, by design as he believes students should have hands-on exposure to research concepts like the scientific method and access to the data that informs experimentation and discovery. More than learning the material, Xu hopes that his students will come to master the key tenets of scientific research and understand how data helps test theories, unlock critical insights and motivate new research questions.

“We cover how to conduct basic research and how to collect, organize and analyze data,” he explains. Students gain hands-on experience with both survey and experimental data and learn to deploy a range of statistical modeling and visualization techniques. Working with a data set from Xu’s own research has helped familiarize students with challenges like outliers, missing values and most importantly, large data sets.

In the case of Xu’s adolescent activity study, sensors can record around 10,000 data points per week per participant including values for age, gender, weight, sitting, standing, walking, running, step count, speed, intensity of exercise, duration of exercise, duration of rest and duration of sleep. Even after consolidation, such a data set can run into the millions of megabytes.

This kind of data can be daunting for researchers, much less students. That’s why Xu has not only introduced his students to the data set, he’s also introduced them to the tool he uses to analyze it.

That tool is JMP®.

“By familiarizing students with software like JMP, it makes statistical thinking less intimidating – they start to think of statistics as a routine, professional thing,” says Xu. “With JMP, it feels like the threshold for statistical problem solving is actually very low – and that’s incredibly valuable.”

JMP® encourages trial and error, ‘the key to the learning process’

Most of the undergraduates in Xu’s classes start the semester without much background in statistics and, he says, it’s important to give them a tool that will also provide them with a hands-on way to learn statistical concepts. With JMP, students can learn everything from the basics like calculating means and standard deviations to more advanced concepts like data exploration.

“JMP encourages trial and error, and that’s key to the learning process,” Xu says.

Furthermore, that students are using real research data exposes them to the practical and logistical challenges that face researchers today. Xu encourages them to think critically about the best way to address missing values and outliers. And, he says, they begin to learn that exploratory data analysis and data cleaning can often be just as important, if not more so, than the actual analysis.

“What my students like about JMP is that it’s very easy for them to learn,” Xu says. “They like that after each step, there are graphics to present the results. So it’s intuitive, and they get feedback. If they go in the wrong direction, they can easily erase the vector and add a new method to try again. This ease of use motivates them to study further. They find studying with JMP to be very interesting.”

An approachable tool with powerful statistical capabilities

In the days before faculty at NSI had access to JMP, as at many universities around the world, Microsoft Excel was the institutional standard for both teaching and research. Despite its uses, Xu says, Excel is too limited to manage complex data sets replete with outliers and missing values, adding, “the amount of data I’m dealing with is also very large, and in Excel the processing speed is very low.”

Xu has met similar constraints with SPSS, which he says is more professional than Excel, but with similarly slow processing speeds. “I once helped a colleague using SPSS for research. Using my own data set of 700,000 samples, the slower processing speed was very obvious!

“SPSS can also be very difficult to use. If you don’t know [how to run a] stats check at a professional level, you could end up at a dead end very quickly because you don’t know how to go to the next step.”

More complex statistical methodologies like logistics analysis and partitioning, which are central to Xu’s research, are fully absent in Excel but one of the strongest features of JMP – and JMP Pro, which Xu and his graduate students use for its more advanced statistical capabilities. “I’m looking to understand the intersection between exercise and physical health,” he explains. “And with logistics analysis and partitioning [in JMP and JMP Pro], I can really add a richness to our understanding of how factors interact.”

Furthermore, the software’s graphical features enable dynamic visual representations that promote exploration. “With Graph Builder, you can change the X and Y axes at any time. There are so many special features that Excel or SPSS can’t do.”

Daily exercise makes for better health outcomes

Xu’s research has already yielded meaningful results, with noteworthy implications for public health. Most importantly, Xu says the data shows that young people should aim for a moderate level of activity for at least 60 minutes per day. “After reaching this amount of exercise, it will have an impact on your physical health,” he explains, noting that such a regimen is linked to improvements both in cardiopulmonary function and muscular strength. Furthermore, health in adolescence – a time in which the body is still developing – can have consequences for an individual’s lifelong health and fitness.

“This is one of the biggest problems today; the level of physical activity among young people is insufficient,” Xu says. “And this problem is not just facing China, but also the United States and Europe.”

That’s a public service announcement that everyone can get behind. It’s a relatable message and, if you’re an undergraduate student at NSI, a great starting point to begin learning about what statistics can tell us about exercise habits and health. “You can use statistical methods to solve many of today’s most important questions,” Xu concludes. “Knowing how to use statistics to solve problems is one of the most important life skills any student can develop.”

The results illustrated in this article are specific to the particular situations, business models, data input and computing environments described herein. Each SAS customer’s experience is unique, based on business and technical variables, and all statements must be considered nontypical. Actual savings, results and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software.