Abstract
80% of Australian children do not engage in recommended minima of physical activity levels, contributing to an alarming trend in obesity levels and associated diseases in adult life. We created iEngage, an innovative health education program for 10–12 year old school children that blends a learning app, wearable technology, feedback, goal setting and gamification with practical activities to promote knowledge and behavioural changes with regards to physical activity and to guide children at their own pace towards World Health Organisation’s recommended minima of daily moderate to vigorous physical activity. We present how the activity trackers are used to provide objective feedback and support the learning activities and the individual goal setting. We conducted a controlled pilot study in two Australian schools. Post-tests using research-grade accelerometer devices reveal a significant increase in moderate and vigorous activities in the experimental group, compared to none in the control group.
Access provided by CONRICYT-eBooks. Download conference paper PDF
Similar content being viewed by others
1 Introduction
The sharp increase and affordability of human-centred technologies are extending the range and transforming the way education programs can be delivered. In particular, domains where the learning does not only occur through a computing interface but also through physical activity can now be supported with wearable technology [1] and it becomes feasible to build smart educational systems that also build on wearable sensors collecting physical student data to drive instruction.
One of these areas is children’s education with regards to Physical Activity (PA) and health literacy. Engaging in healthy levels of daily PA, especially at Moderate to Vigorous Physical Activity (MVPA) levels, is an important factor for health and wellbeing [2]. Sadly, studies show that children often do not meet World Health Organisation’s recommended minima of PA levels, especially for MVPA, creating a significant health risk for them later in life [3]. Whilst eHealth intervention programs exist (e.g. [4]), these trends persist. An obvious reason for this is that children may be unable to associate these recommended levels with what they actually do and feel, nor know how to achieve these recommended levels, which are expressed in terms of number of steps (12,000 daily) and intensity (60 min per day of MVPA). It is therefore important to design learning systems that enable children to understand and experience what these minima mean for them.
Our multi-disciplinary team created iEngage, an innovative health education program for 10–12 year old school children that blends a learning app, wearable technology, feedback, goal setting and gamification with practical activities to increase health literacy and encourage behavioural changes with regards to physical activity at an individual pace. In iEngage, the activity trackers are only connected to the learning app and provide objective feedback to support the learning activities and the goal setting. In this paper, we describe how the technology was used and report on our pilot experiment in an Australian urban primary school.
2 Overview of iEngage
iEngage comprises ten learning modules of 45 min, each on a specific topic, delivered at school over several weeks. The learning contents are research-informed, with real physical activities, immediate feedback, goal setting and gamification.
The Technology: Learning App and Activity Trackers.
The iEngage app, built on the BePatient platform [5], is accessible on android tablets. It connects seamlessly to a background app that synchronises (via Bluetooth) the learner’s activity tracker and uploads the data immediately onto the iEngage server (via wifi or 3/4 G), hence allowing real-time data to be used in the learning activities. A commercial wrist-worn activity tracker (Misfit Ray [6]) is worn continuously during the whole duration of the program. It is waterproof, runs on 6 month-life batteries and is fairly secure around the wrist to prevent losses or breakages. The step count is extracted per minute.
A Child-Friendly Learning Interface Design.
Each functional part of the program is guided by the same animal mascot so that children quickly recognise what is expected of them: a giraffe presents learning content, a bison gives quizzes, a kangaroo (Fig. 1(c)) guides the activity tracker synchronisation and data readings, a tiger is in charge of actual physical activity, a penguin supports goal setting and a bird rewards success with cues for a secret message hunt. The visualisation interfaces of PA are designed specifically to support the child’s reflection and self-monitoring during the learning activities. They strictly contain the information that the child needs, without additional information that commercial interfaces of the tracker would typically have.
3 Personalised Learning Supported by Wearable Tracker
Learning about physically-related knowledge and skills cannot remain theoretical and abstract. The role of the activity tracker is to provide objective PA data input into the learner model, and support the child’s personalised learning.
Association Between Perceived Exertion and PA Intensities.
There are different intensity levels of physical activity, broadly called light, moderate and vigorous. Health recommendations are expressed in amount of time (minimum or maximum), that people should spend in each. Figure 1(a) shows a scale of perceived exertion. An important skill that children need to learn is what each intensity level means for them and recognise them as they engage in each.
iEngage offers experiential learning activities where children are instructed to do various physical exercises or games, and then explore their data in a simple, child-accessible way. Figure 1(b) shows one interface, where children can explore their PA in the last 15 min. The intensity for each of these minutes is color-coded: grey for sedentary times, orange for light activities and green for moderate and vigorous.
A typical learning task is to (i) leave the tablet and carry out specific exercises of specific intensity for 10 min, (ii) come back to the app and answer some questions about how they feel and how they perceived their effort (iii) synchronise their tracker and explore their own data in the last 15 min (iv) compare their perceived effort with the objective data.
Association of Perceived Activity with Step Counts.
Similarly, children can learn and associate what step counts mean for them and evaluate for themselves how many they achieve in their regular daily schedule, and what reaching the 12,000 minimum step count mean in their own context. Here, daily summaries are shown per day, week, month. Students reflect on how many steps they accumulate in their typical days, and how many more they could achieve by doing PA learned during the iEngage sessions.
Setting Achievable Goals.
At the end of each module, students set individual goals for their daily physical activity. They can choose from lightly increasing their current PA and intensity level to more challenging goals. The aim is to guide them to reach recommended levels, gradually increase goals throughout the program, and then maintain them. At the beginning of the next module, the system guides students to synchronise their tracker, reflect on their activity and check whether they have achieved their goals. If they have, they receive a reward or just receive encouragements.
4 Pilot Study
Experiment.
We conducted a pilot study in 2 urban primary schools in Sydney with 59 children aged 10–12 (27 girls and 33 boys). In both groups, girls had a distribution of aerobic fitness equivalent to the international normative values, and boys had an over-representation in the ‘poor’ aerobic fitness category. Both schools were similar in terms of socio-economic background, academic achievements and area. We used a pre- and post-test design to evaluate the efficacy of iEngage. In both schools, children’s physical activity was measured on 5 consecutive days before and after the program with a research-grade accelerometer (Geneactiv [7]) measuring activity at 60 Hz unobtrusively, i.e. without showing any feedback to the child. In the experimental school (EXP), children (N = 33) were given access to the iEngage program during school time and followed 10 learning modules over 30 days. In the control school (CTL), children (N = 26) did not follow any particular health education program.
Data.
The four raw Geneactiv datasets contained 60 Hz three-dimensional accelerometer data. The raw data was processed into 1 s epoch SVMg data points, before being categorised into intensity for each second. Specific cutoffs for identifying activity levels in children were used [8]. As we were interested in comparing daily behaviours, we sliced the data into 24 h periods and counted the number of seconds spent in each intensity level for each child, each day. We filtered out any day where the tracker was not worn all day.
Results.
The average daily times spent in vigorous (V), moderate (M), light (L), sedentary (S) and sleep (Z) by each group, before and after the program, are shown in Table 1. Both Pre-test groups had similar distribution. In the EXP group, the percentage of time per day spent in MVPA increased significantly, whereas the CTL group did not (and even had a small decrease in vigorous activities). No significant effect was found in other activities. Furthermore we found that EXP students spent a lot more time (1079 s vs 682 s), in long bouts of continuous MVPA (longer than 30 s) after the program ended, which suggests that they intentionally engaged in more MVPA and that the increase is not solely due to very short bursts of activity.
5 Discussion and Conclusion
We have presented a case of smart educational tool that harnesses wearable devices to provide experiential and personalised learning. Results of our pilot study suggest that iEngage program can create positive behavioural changes in the targeted area of MVPA, and this through educating children rather than prescribing activity. More analysis is needed to better understand how these changes occur, and whether they are sustained in the long-term.
References
Santos, O.C.: Training the body: the potential of AIED to support personalized motor skills learning. Int. J. Artif. Intell. Educ. 26, 730–755 (2016)
Hardy, L.L., Okely, A.D., Dobbins, T.A., Booth, M.L.: Physical activity among adolescents in New South Wales (Australia). Med. Sci. Sports Exerc. 40(5), 835–841 (2008). https://doi.org/10.1249/MSS.0b013e318163f286
Ekelund, U., Luan, J., Sherar, L.B., Esliger, D.W., Griew, P., Cooper, A., Collaborators, I.: Moderate to vigorous physical activity and sedentary time and cardiometabolic risk factors in children and adolescents. J. Am. Med. Assoc. 307(7), 704–712 (2012). https://doi.org/10.1001/jama.2012.156
Kerner, C., Goodyear, V.: The motivational impact of wearable healthy lifestyle technologies: a self-determination perspective on fitbits with adolescents. Am. J. Health Educ. 48(5), 287–297 (2017). https://doi.org/10.1080/19325037.2017.1343161
BePatient. https://www.bepatient.com/. Accessed 14 Apr 2018
Misfit. https://misfit.com/fitness-trackers/misfit-ray. Accessed 14 Apr 2018
Geneactiv. https://www.activinsights.com/. Accessed 14 Apr 2018
Phillips, L., Parfitt, G., Rowlands, A.: Calibration of the GENEA accelerometer for assessment of physical activity intensity in children. J. Sci. Med. Sport 16, 124–128 (2013)
Acknowledgements
This work was funded by Diabetes Australia Research Trust. We thank colleagues and Bepatient for their various contributions in this experiment.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Yacef, K., Caillaud, C., Galy, O. (2018). Supporting Learning Activities with Wearable Devices to Develop Life-Long Skills in a Health Education App. In: Penstein Rosé, C., et al. Artificial Intelligence in Education. AIED 2018. Lecture Notes in Computer Science(), vol 10948. Springer, Cham. https://doi.org/10.1007/978-3-319-93846-2_74
Download citation
DOI: https://doi.org/10.1007/978-3-319-93846-2_74
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-93845-5
Online ISBN: 978-3-319-93846-2
eBook Packages: Computer ScienceComputer Science (R0)