Youth attending urban public schools in low-income areas are at high risk of both inadequate physical activity (PA) and poor of academic achievement (1). This is especially concerning given the bidirectional nature of the relationship between PA and academic success; adequate PA affects cognitive and behavioral development (2), whereas academic success is linked to improved health behaviors in late adolescence and adulthood, including PA behaviors and resultant fitness (1). Thus, this dual risk in childhood helps to amplify lifetime health disparities experienced by urban youth (3).
Physical education (PE) programming is crucial to helping youth living in underresourced, urban neighborhoods achieve PA guidelines. Although interventions to expand opportunities for integrated PA during academic classroom time, such as Take 10! (4), continue to be developed and tested, less attention has been paid to approaches to optimize time available for PA, particularly moderate to vigorous PA (MVPA), in PE class (5). This is a critical gap because studies have shown that school policies and facilities, including duration, frequency, and equipment available for PE, have a highly significant effect on student PA behaviors, particularly in economically disadvantaged areas where a focus on PE policies can nearly double the number of days that youth achieve 30 min or more of PA (6). Specifically, one study of urban schools in St. Louis, Missouri, found that students got approximately 15 min more of MVPA per day on days they had PE class; this is an impactful amount, given that so few urban youth meet national PA guidelines (7). At the same time, urban public schools often have large class sizes, inadequate PE infrastructure, and personnel; a systematic review in the Journal of School Health found that the ability to affect PA via PE was a critical missing link in academic achievement in urban and low-income public schools (8).
Meanwhile, underfunded urban public schools continue to struggle to close the gap in science learning. The National Assessment of Educational Progress and the Trends in International Mathematics and Science Study document racial achievement gaps linked to socioeconomic status, school resources, and infrastructure deficits by fourth grade (9). Other studies using the Early Childhood Longitudinal Survey have found that among third graders, African American students score approximately 1 SD lower than their white peers (10), whereas the science gap between Latin/Hispanic and White third graders is about 0.8 SD (11).
Active Science (AS) is a novel program that attempts to address this dual burden. The overarching goal of AS is to economically deliver a developmentally appropriate and engaging intervention to increase PA intensity and duration among high-risk youth, while simultaneously priming them to engage the science curriculum content delivered each session, thereby decreasing gaps in academic achievement (12). Using a low-cost, tablet-based application, AS combines child-friendly PA self-measurement with science, technology, engineering, and math (STEM) learning modules (Fig. 1 and Supplemental Materials, http://links.lww.com/TJACSM/A99 and http://links.lww.com/TJACSM/A100) mapped to Massachusetts science standards (see Supplemental Materials for the AS curriculum map, http://links.lww.com/TJACSM/A99). This integrated exerlearning technology, where exerlearning is defined as programming specifically designed to increase both exercise and student academic content knowledge, is currently being used in YMCA after-school, camp, and childcare settings across the country (12).
AS is not a PA curriculum. Instead, it gamifies existing PA or PE activities, with the goal of increasing children’s MVPA through the incentives of PA tracking, games, and virtual rewards. This approach allows a cost-effective and setting-flexible intervention that can be adapted to site-specific staffing, space, equipment, available time, and population needs. The tablet-based application uses accelerometer data collected during site-specific PA programming to guide children through interactive STEM content lessons, and it includes integrated pre- and postassessments of science knowledge (Integrated Active Science Knowledge Assessment [IASKA]). Inexpensive, commercially available accelerometers are used, which clip onto students clothing. The application stores students’ PA data as they progress through the curriculum, using it to demonstrate STEM concepts such as hypothesis formation and testing, and creation of tables and graphs. Children are awarded virtual trophies, and short games are unlocked for achieving a variety of progressive PA benchmarks as they advance through the STEM curriculum (see screen shots in Supplemental Materials, http://links.lww.com/TJACSM/A99 and http://links.lww.com/TJACSM/A100). The use of accelerometers for self-tracking of PA intentionally builds on several theoretical self-efficacy constructs and is backed by empirical evidence supporting effects on increased PA quantity and intensity among youth (12). Aligned with the framework for comprehensive school physical activity programming (13), AS has potential for adaptation to PE curricula to increase student engagement in MVPA and science learning.
AS has recently been introduced into urban public school systems, where it has been implemented during PE classes in which student PA engagement in terms of both quantity and intensity has been viewed as challenging by teachers and administration. It does not require the purchase of new equipment such as smart boards; the application can be downloaded onto tablets already in use in most schools. The program was also seen as an opportunity to reinforce classroom science learning with engaging technology in a fun setting (PE class vs academic homeroom). However, this is an untested approach and warrants evaluation. Specifically, although AS may improve the intensity of PA in PE class because of the self-measurement of PA data and virtual rewards, the use of the tablet-based application during PE class may decrease the time available for PA. Also, the use of the science learning application in a PE setting may decrease the science content knowledge gains made by students. Therefore, the specific aims of this study were as follows:
- 1) to explore how the use of AS affects the duration and intensity of PA in PE classes; and
- 2) to determine whether the use of AS in PE classes improves student attitudes toward science and Massachusetts third-grade science curriculum content knowledge.
MATERIALS AND METHODS
Design and Setting
This study was designed as a two-arm pragmatic evaluation of new and ongoing PE programming within the Lawrence Public Schools. Lawrence, Massachusetts, has a population of just over 80,000, 77.1% of whom are Hispanic. The city has a mean per capita income of $17,059 per year; approximately 26.4% of its citizens live in poverty (U.S. Census Bureau, 2017). Through an existing partnership with the Merrimack Valley YMCA, five Lawrence elementary schools were piloting the AS program in third-grade PE classes during the 2017–2018 school year. One school was randomly selected to compare MVPA and STEM learning outcomes in PE classes using AS to standard PE classes. Within that school, six third-grade classrooms were randomly assigned to either the intervention condition (the use of AS in PE class once every six school days for 8 wk) or the waitlist control (standard PE for 8 wk and then AS in PE for 8 wk). Written parental consent and study assent were obtained for all participants. Because only previously planned school programming was being evaluated, this study is not considered a clinical trial. All study protocols and materials were approved by both the Merrimack College Institutional Review Board as well as the City of Lawrence Public Schools.
PE PA outcomes assessed included total steps, steps per minute, minutes of PA, minutes of PA as a percentage of available class time, minutes of MVPA, and minutes of MVPA as a percentage of available class time during PE classes. Attitude toward science was measured via the Modified Attitudes toward Science Inventory (MATSI), a 25-item inventory using a 5-point Likert scale, which has been tested and found valid and reliable (α = 0.7) among in urban, elementary school students (14). Science content knowledge was measured using the Integrated Active Science Knowledge Assessment (IASKA), composed of five questions assessing content contained in the first curriculum module of the AS program, which is mapped to Massachusetts science curriculum standards for third grade (Supplemental Materials, http://links.lww.com/TJACSM/A99 and http://links.lww.com/TJACSM/A100). The IASKA is not a validated instrument; it represents selected questions from Massachusetts curriculum assessments by grade level, which are in turn mapped to each level of STEM lessons presented by the AS curriculum.
There were 142 eligible third-grade students in six classrooms; of those students, N = 135 were consented to participate in the study and contributed PA data during PE classes. The average age was 8.9 yr (SD = 0.4), and participants were 54.1% female. Of the consented students, n = 66 completed both pre- and post-MATSI and IASKA testing. There were very limited class periods available for student data collection to minimize effects on academic and PE time, and student absenteeism rates are very high in this very low-income school district. If a student was absent for either baseline or follow-up data collection on the day their class was scheduled for testing, they were dropped from the study.
Baseline fitness was measured using the PACER test during PE class before collecting baseline PA data. Baseline PA was collected via ActiGraph wGT3X accelerometers during PE classes in the 2 wk before program start. Research assistants fitted the ActiGraphs to students upon arrival to school and removed them before student departure. To collect data on PE PA and MVPA in both the intervention and the control conditions, research assistants fitted children with the EKHO digital accelerometers used by AS as they entered the gymnasium for their 48-min PE classes. Over the 8-wk study period, students then engaged in established PE curriculum, which changed by week and focused on standard PA competencies. Because of the 6-d rotation schedule, school holidays, and staff development days, students could attend a maximum of six PE classes over the 8-wk period. The PE curriculum was taught by the same teacher to all classes, and both control and intervention groups undertook the same activities in their respective classes. During the 8-wk period, the PE curriculum included 4 wk of sport-specific skill building (basketball and soccer) and 4 wk of station rotation. Accelerometers were collected at the end of PE class, and digital readouts were recorded with pen and paper by research assistants. Surveys were administered pre- and postintervention using a one-on-one format by trained research assistants in a quiet testing space. Students were first administered the MATSI, followed by the IASKA. A simple process survey evaluating student perceptions of AS was administered during follow-up data collection.
PE programming was the same during PE classes for both intervention and control groups. However, the intervention group had 15 min reserved at the end of class to use the tablet-based AS programming. At the beginning of that 15-min period, students in the intervention group entered their EKHO digital accelerometer data into the AS application to initiate a virtual AS lesson, which used their PA data to illustrate scientific concepts; screenshots of example lessons are contained in Supplemental Materials, http://links.lww.com/TJACSM/A99 and http://links.lww.com/TJACSM/A100. After recording their data and completing their AS lesson, students in the intervention group received access to a short (<3-min) tablet game as an incentive for completing their lesson.
A priori power analyses were based on science learning outcomes (α = 0.05, 1 − β = 0.8) and indicated a minimum sample size of n = 34. However, because this was a pragmatic evaluation of ongoing public school PE programming where student absenteeism is high and PA outcomes might include clustering effects, all third-grade classes (n = 142, k = 6) were included. All statistical analyses were completed in STATA 13.
Descriptive statistics were used to describe sample demographics (age and gender). Two-sample t-tests were used to assess potential baseline group differences in age, preintervention MATSI scores, preintervention fitness, and preintervention PE PA. A two-proportion z-test was used to assess potential baseline group differences in gender. Because it is a discrete, nonnormally distributed variable in which the median significantly differs from the mean, a nonparametric equality of medians test was applied to assess potential baseline group differences in preintervention IASKA scores.
For specific aim 1, effects of intervention participation on PA outcomes were assessed using multilevel mixed effects regression for panel data collected each day of PE (average of four PE classes per participant). Intervention condition was included as a fixed effect, whereas individual variability and classroom-level variability were included as random effects, the latter due to potential group effects on individual PA engagement during group-based PE activities. For specific aim 2, changes in pre- and postintervention attitudes toward science learning (MATSI scores) were assessed using a paired t-test, whereas changes in science content knowledge (IASKA score) were assessed using the more conservative nonparametric Wilcoxon signed rank test.
Table 1 shows participant characteristics by intervention and control group. There were no significant differences in age, gender, baseline fitness, attitudes toward science, science content knowledge, or baseline PE PA between groups. Of the n = 66 students who completed posttesting, 84% reported that they liked doing AS in PE class, whereas 68% reported that they would additionally like to use AS in science class (23% reported that they were unsure). No student participation refusals were reported during the intervention period.
PE PA Outcomes
Over the course of the 8-wk measurement period, students participated in an average of four out of a possible six PE classes. There were six snow days over the intervention period, which also lowered the possible number of PE classes for some classrooms. Overall mean steps, minutes of PA, and minutes of MVPA per class and by intervention condition are shown in Table 2. Results of PA models are shown in Table 3. Because of the 15-min reduction in time available for PA due to tablet use in the intervention classrooms, participants in intervention classes had significantly fewer steps, minutes of PA, and minutes of MVPA as compared with students in control classes. However, students in the intervention classes had significantly higher steps per minute of available PA time, as well as PA minutes as a percentage of available PA time. Intervention status had no effect on MVPA as a percentage of available PA time.
Science Learning Outcomes
Results of paired t-tests indicated no significant difference in pre- and post-MATSI scores in either control (P = 0.47) or intervention group (P = 0.84) (Fig. 2). However, Wilcoxon signed rank testing showed that although there was no significant change in pre- and post-IASKA scores in the control group (P = 0.64), intervention participation was associated with a significant improvement (P = 0.005), indicating increased third-grade science content knowledge (Fig. 3). Sensitivity analyses showed no difference in average scores by group for either baseline or follow-up MATSI or IASKA if nonpaired student scores were included.
This study examined the effects using the AS program in PE classes on student PA levels, attitudes toward science learning, and science content knowledge. AS uses student-worn digital accelerometers to track PA metrics such as steps and minutes of MVPA, which students then input into a tablet-based application to complete science lessons mapped to Massachusetts science curriculum standards. The application also tracks student PA and awards virtual trophies and game incentives for achievement of specific PA benchmarks. The pragmatic evaluation took place in third-grade classrooms within an urban public school system initiative to improve PA levels and close academic achievement gaps.
Average minutes of PE PA and PE MVPA were low in both the intervention and the control classrooms. Students averaged only 11.7 min of PA per 48-min PE class (24% of available time), and only 6.3 min of MVPA (13% of available time). Although consistent with other studies of PE PA in urban schools serving low-income neighborhoods, increasing PE PA levels should be a priority.
We found that the time necessary to enable students to complete lessons on the tablet-based application significantly decreased the overall steps, minutes of PA, and minutes of MVPA that students averaged during the 48-min PE classes. Although over time student proficiency would likely decrease the amount of time necessary for tablet use, this finding is of concern. However, the intensity of PA appeared to increase in AS classrooms, perhaps because of student self-monitoring and tracking of PA metrics, as well as virtual awards and game incentives within the application. Additional research should be conducted to better understand the relative role of self-monitoring and tracking versus incentives, as well as whether these effects are sustained over longer time periods.
Although AS participation did not appear to improve attitudes toward science learning, participating students did significantly improve Massachusetts third-grade science curriculum content knowledge relative to students in control classrooms. Although the IASKA used within the AS program is not a validated measure, this is an encouraging finding that should be confirmed in future studies of AS use in public schools. However, given that students only took part in an average of four AS lessons, future research should take place over longer periods of time to allow better evaluation of effects of AS participation on student science knowledge within the public school setting.
Given the already limited time available for PE in most public schools and the importance of PE for achieving PA guidelines among underserved students, our findings indicate that optimal effects of AS may be obtained by using accelerometers during class, but moving their data input and tablet-based science learning application outside of PE class. Because lessons are self-guided, AS tablet use could easily occur during classroom transition or individual work times. This would allow students to garner the intensity benefits of the AS program in PE class, while maximizing minutes of PE class available for PA. This approach may have another benefit, namely, that students may better associate the use of the AS tablet application—which has been shown to be popular with students—with science learning within the classroom. This may help to elicit the improvements in attitudes toward science learning that were absent from this study.
Although an important step in evaluating this innovative program, this study had important limitations. The 8-wk study period took place during winter months, and so school cancellations and student absences for illness limited the average number of exposures per student from the planned six exposures to only four exposures. Because of the importance of not disrupting PE programming, PA metrics were collected using the EKHO digital accelerometers used by the AS program. However, the EKHO accelerometers were validated against ActiGraph w-GT3x accelerometers during baseline >data collection. Lastly, data on student language were not collected; because AS is delivered in English only, student engagement is likely modified by English proficiency level. Future research should evaluate differences in programmatic effectiveness for children for whom English is not their first language. If AS use continues to be advanced in urban, low-income school districts, creating multilingual programming versions should be a priority.
AS is a promising programming approach for use in public schools aiming to economically improve student PA and science achievement. Unlike interventions to integrate additional PA into classroom time, AS gamifies existing PE curriculums to increase the intensity of PE PA. This approach is both low burden and leverages existing school resources. This study showed significant short-term gains in third graders’ science content knowledge with AS use. However, targeted changes to programming sequencing—specifically moving tablet use out of PE class and into academic transitional or science instruction periods—would help better achieve the goal of improving PA quality and duration.
This study was made possible through generous funding by the New Balance Foundation. The authors thank the dedicated teachers and administration in the Lawrence Public Schools, as well as the AS staff at the Merrimack Valley YMCA, who made this study possible.
The authors declare that they have no conflict of interest. Presentation of the results of the present study in this manuscript does not constitute endorsement by the American College of Sports Medicine.
1. Fiscella K, Kitzman H. Disparities in academic achievement and health: the intersection of child education and health policy. Pediatrics
2. Lubans D, Richards J, Hillman C, et al. Physical activity for cognitive and mental health in youth: a systematic review of mechanisms. Pediatrics
3. Ickovics JR, Carroll-Scott A, Peters SM, Schwartz M, Gilstad-Hayden K, McCaslin C. Health and academic achievement: cumulative effects of health assets on standardized test scores among urban youth in the United States. J Sch Health
4. Mavilidi M, Okely A, Chandler P, Paas F. Infusing task-relevant physical activities into the classroom: effects on preschool children's geography learning. J Sci Med Sport
5. Sutherland R, Campbell E, Lubans DR, et al. Physical education in secondary schools located in low-income communities: physical activity levels, lesson context and teacher interaction. J Sci Med Sport
6. Ganzar LA, Ranjit N, Saxton D, Hoelscher DM. Association of School Physical Activity Policies with Student Physical Activity Behavior. J Phys Act Health
7. Racette SB, Dill TC, White ML, et al. Influence of physical education on moderate-to-vigorous physical activity of urban public school children in St. Louis, Missouri, 2011–2014. Prev Chronic Dis
8. Basch CE. Healthier students are better learners: a missing link in school reforms to close the achievement gap. J Sch Health
9. Curran FC, Kellogg AT. Understanding science achievement gaps by race/ethnicity and gender in kindergarten and first grade. Educ Res
10. Kohlhaas K, Lin HH, Chu KL. Science equity in third grade. Elem Sch J
11. Morgan PL, Farkas G, Hillemeier MM, Maczuga S. Science achievement gaps begin very early, persist, and are largely explained by modifiable factors. Educ Res
12. Finn K, Yan Z, Martin E, McInnis K. Active Science pilot study: promoting physical activity and science learning among children. J Behav Health
13. Carson RL, Castelli DM, Kulinna PH. CSPAP professional preparation: takeaways from pioneering physical education teacher education programs. J Phys Educ Recreat Dance
14. Weinburgh MH, Steele D. The modified attitudes toward science inventory: developing an instrument to be used with fifth grade urban students. J Women Minor Sci Eng