Students work with big data, survey teen sleep

Project Leader: Dr Linda McIver, Australian Data Science Education Institute

How many hours of sleep do you lose due to your mobile phone each night? The Australian Science Data Education Institute (ADSEI) might have an answer.

ADSEI created a sleep survey on teenage mobile phone use before bed. It asked respondents around the world to rate the hours of sleep they get and how long before bed they put away their phones.

The sleep survey is also upskilling high school students in data literacy and data visualisation. It’s overcoming the fear of big numbers and getting students to understand how data can help themselves and their communities.

This survey is still running, complete it here: https://learn.pawsey.org.au/data/sleep/survey 

 
Countries surveyed
 
The oldest person to complete the survey
 
People completed the survey
Partner Institution: Australian Data Science Education Institute System: Visualisation Areas of science: Data Science Applications used: Python

The Challenge

The 2018 Longitudinal Study of Australian Children Annual Statistical Report found nearly a third of Australian teenagers with internet access in their bedrooms were losing sleep.   The blue light from portable devices, like phones and tablets, suppresses melatonin release. Melatonin is a vital hormone that gets your body ready for bed. Sleep deprivation lowers mood, increases stress, impacts health and decreases academic performance.

Linda McIver is the Executive Director of the ADSEI, a charity that trains kids and teachers in data science.

“I ran a workshop for teachers at Pawsey which Marianne Beattie from All Saints’ College attended. She wanted a project for her Year 9s and 10s to explore the impact of mobile phones on sleep. She wanted to combine that with the opportunity to work with big data,” Linda says.

Data literacy is the ability to understand and communicate facts to solve a problem. Around 20% of Australians are data literate. We fair better than Europe and Japan, but worse than the USA and India. Poor data literacy costs Australian industries roughly $13.9 billion per year.

“We’re working with real data that kids can use to explore data science and build their data literacy skills.”

But it’s not just students who lack data literacy skills. Linda says few teachers have the knowledge to manipulate big data packages or understand high-performance computing (HPC).

“The fear factor is the biggest obstacle. When people hear terms like data science and HPC, they think they’re way outside anything they can cope with. The fact is, it’s actually really accessible if it’s presented in the right way.”

The Solution

Ann Backhaus, Education and Training Manager at Pawsey, had a vision for STEM projects. She wanted to combine real world data with Pawsey’s HPC and visualisation expertise to make STEM subjects more engaging to students. She approached Linda to use the sleep project as the focus for a summer internship to build a platform that would support Data Science projects. Ann and Linda worked with Pawsey Senior Visualisation Specialist, Yathu Sivarajah and visualisation specialists Ali Zamani and Jesse Helliwell to supervise a Pawsey Summer Intern, Thai Nguyen. Thai developed the full stack web site to collect data into a database and create visualisations on the fly to allow students to explore the results.

“From a technology perspective, Pawsey is coordinating all the technical expertise and hosting the platform, so we’re heavily involved. Our plan is to facilitate ADSEI throughout this process, not just by hosting the platform but by providing guidance,” Yathu says.

“Pawsey are providing the funding and support to train teachers to use this platform as well,” Linda says.

The survey is being used by students to practice visualising big data. Students usually deal with easy-to-visualise data sets. A few dozen points can be easily placed on a graph and outliers are obvious. When hundreds or thousands of points become involved, it’s more difficult to read what the data are saying.

The team had Thai build a visualisation platform in React. The platform helps students compare different questions in the sleep study, such as participant country and age. This way, students get familiar with using big data without needing an in-depth understanding of coding.

“Python graphing libraries are really difficult to use for kids. We wanted a way for kids to play with the data and go, ‘What if I compare this field against that field?’ We want them to be able to download and play with the data, to look for relationships and build their programming skills,” Linda says.

The sleep survey isn’t just for data training though. Yathu says schools can use it to explore different variables, such as the impact of exercise or stressors, on students’ sleep patterns.

“You could have students contribute at different times. For example, you could get them to answer questions about how they’re sleeping before and after exams, then maybe trial ways to prepare students for test taking by directly involving them in the discussion.”

Outcome

So far, over 500 participants, mostly Australian teens, have responded to the survey. It will remain open indefinitely at this point. Linda hopes to get thousands more students contributing to the dataset, so it’s a true representation of working with big data.

Linda is also hosting a series of workshops for teachers to train them in using ADSEI data sets in schools. These workshops feature Pawsey staff introducing teachers to the concept of HPC.

“We’re planning to do these workshops online to make them accessible right across Australia. And the goal is to run four of them this year.”

Pawsey Head of Scientific Platforms, Mark Gray was featured in one of the recent workshops.

“He talked about some of the applications of big data and the teachers were just blown away by it. Just hearing the different ways supercomputing impacts the world is so powerful. A lot of teachers said that was the best part of the workshop.”

The data manipulation tools will be used for a variety of future studies for students. They’re designed to need little reworking to adapt to new big data sets, so national or international education programs can use them.

“The same workflow can be repeated across studies. We want this platform to be widely accessible, so it can be used across Australia, potentially globally, for school curriculums,” Yathu says.

Pawsey is continuing to host the ADSEI sleep survey and is working with Linda to empower students with data literacy. By knowing how to look at data, students are preparing themselves for their future careers and giving themselves the tools they need to contribute to their communities.

We’re working with real data that kids can use to explore data science and build their data literacy skills
Dr Linda McIver, Australian Data Science Education Institute,
Project Leader.