Classroom-Based Physical Activity and On-Task Behavior : Translational Journal of the American College of Sports Medicine

Secondary Logo

Journal Logo

Original Investigation

Classroom-Based Physical Activity and On-Task Behavior

Mahar, Matthew T.

Author Information
Translational Journal of the ACSM: September 1, 2019 - Volume 4 - Issue 17 - p 148-154
doi: 10.1249/TJX.0000000000000099
  • Free



Health benefits of physical activity are widely known (1,2); however, less is known about the additional benefits or potential limitations of physical activity when it is incorporated into the classroom. Research on the effectiveness of classroom-based physical activity began to be published around the mid-2000s. Integrating physical activity into the classroom can help youth meet the U.S. Physical Activity Guidelines (3). Part of the Physical Activity Guidelines recommends that youth participate in at least 60 min of moderate-intensity physical activity per day. At least half of this physical activity should be achieved during the school day (4).

Because approximately 95% of youth spend upward of 6 h·d−1 at school for about half of the days each year for 13 yr, schools represent an ideal location in which to implement policies that can improve student health (5). The main goal of most school districts is to teach academic skills. That, of course, does not mean that school personnel, such as teachers and administrators, are uninterested in children’s physical health, but it does appear that the demands made on teachers are so great that they must focus on the main goal, academic learning, leaving little time to focus on physical and mental health. A wide-angled view of the education of youth, however, will show that multiple variables affect academic learning (6–9). As such, it is logical that consideration of the whole child should involve thoughts about the benefits of physical activity or movement on academic outcomes, behaviors that affect academic outcomes, and health behaviors.

Classroom-based physical activity can conveniently be subdivided into physical activity that is integrated into the academic lesson (sometimes called active lessons, e.g., Energizers) and physical activity breaks (distinct movement episodes interspersed between periods of [typically seated] instruction, e.g., FUNtervals). Potential outcomes related to movement in the classroom can be classified as proximal outcomes or distal outcomes. Proximal outcomes include classroom behavior, cognitive performance, and physical activity. Distal outcomes include measures of academic achievement and standardized test scores. This article will focus on proximal outcomes, particularly academically focused on-task behavior.


When we originally developed activities that integrated academic concepts for classroom teachers in the early 2000s (10), our team worked with Kymm Ballard, who was then with the North Carolina Department of Public Instruction. At that time, few activities were available that integrated physical activity and academic learning, although individual teachers throughout the country were no doubt using movement-based learning in their classrooms. We felt strongly that classroom-based physical activities could become an important component of a comprehensive school physical activity plan. In January of 2003, the North Carolina State Board of Education passed the Healthy Active Children Policy (HSP-S-000), which provided local school districts with resources to promote coordinated school health programs. In April 2005, this policy was revised to mandate that schools provide a minimum of 30 min of physical activity for all K-8 students daily. We used this information to demonstrate to teachers how this mandate could be met with physical education, recess, and classroom-based physical activity. With grant support from the North Carolina Department of Public Instruction, we developed a collaborative team of university-based physical activity experts with teaching experience, coupled with current elementary school teachers to develop classroom-based physical activities that we called “Energizers.”

We then trained classroom teachers to implement Energizers and assessed the effect of these classroom-based physical activities on physical activity and on-task behavior (11). We assessed physical activity with hip-worn pedometers and on-task behavior with direct observation in the classroom. On-task behavior is a time-intensive measure with a particular set of measurement issues, but it represents a measure that can be closely linked with the behavioral intervention (i.e., a proximal outcome) and one that teachers inherently understand and value (12–14).


Relatively few studies have assessed on-task behavior via direct observation, mainly because it is difficult and time consuming. For example, first a system of observation must be developed or adopted. Then observers must be trained to implement the direct observation system. This training starts in the laboratory (e.g., learn the definition of on-task and off-task behavior, learn what specific behaviors represent on-task and off-task behavior, observe training videos, process feedback, and discuss agreements and disagreements) and progresses to training in the school-based environment (practice recording measures of on-task and off-task behavior on real students in an authentic educational setting). Training does not stop there, however. Because humans tend to drift in their observations, periodic retraining of observers is essential to obtaining reliable observations for interventions that use longitudinal designs (15,16).

Any school-based intervention requires buy-in from the school principal, as well as the individual teachers who are expected to implement the intervention. Generally, teachers will need to be convinced that the activities are likely to help them do their job better before they fully subscribe to the proposed program. A high level of fidelity in delivering the classroom-based physical activities as intended will only result from a high level of school and teacher buy-in. We agree with Bartholomew et al. (17) that teachers are more likely to implement activities that demonstrate academic-related outcomes than increased physical activity levels. Thus, the importance of assessing outcomes that speak the language of teachers is of paramount importance to the sustainability of any school-based intervention. The ability to pay attention in class is a key predictor of academic success (18–20). Student attention to appropriate stimuli is often called on-task behavior or time-on-task behavior. The term “on-task behavior” will generally be used throughout this article for clarity and consistency, except when particular studies report only on other variables (e.g., off-task behavior). On-task behavior includes verbal or motor behavior that follows the class rules and is appropriate to the learning situation. As such, on-task behavior is context specific, defined with reference to both the rules of the classroom and the teacher designated academic activity. Examples of on-task behavior might include sitting at one’s desk while working, engaging in group games when appropriate, responding to teacher questions, demonstrating activity to others when expected to do so, or talking during class discussion. Direct observation of live behavior in the classroom may provide observers with the best opportunity to assess on-task behavior because being physically present in the classroom allows observers to understand the context of the situation.

In direct observation studies of on-task behavior, interobserver reliability must be assessed. If credibility of the observation data is uncertain, then interpretations of these data are questionable. Credibility of observation data usually depends on researchers demonstrating adequate levels of interobserver reliability (15). In observation studies, reliability refers to the agreement of two observers who have observed the same student at the same time. The premise is that if two independent observers record the same code (e.g., on-task) when they observe the same student at the same time, then confidence that the coded observation is probably appropriate is enhanced. If levels of agreement between observers are low or not reported, then readers should be skeptical of the correctness of the coded observation. Universal agreement about acceptable values for interobserver reliability does not exist. Interobserver agreement depends on multiple factors, including the variability of the measure being observed, the incidence of the behavior of interest in the observed sample, the number of categories in which the observed behavior can be classified, and the consequences of the evaluation. An average interobserver reliability of 85% is often considered acceptable, with agreement between any two observers less than 70% reason for concern (14). The strongest evidence of interobserver reliability is provided when two independent observers observe and record data on some percentage of the actual observations during the data collection phase of the study. The number of reliability observations required to provide highly credible evidence is typically between 20% and 40% of total observations (11,15). Evidence of interobserver reliability for each observation system provided in a previously published study does not mean a new study using different observers necessarily produces reliable data. It is recommended that evidence of interobserver reliability be provided for each study and that reviewers of observational research studies require that interobserver reliability be reported for each study (16). Evidence of interobserver reliability for the new study based only on training observations is less convincing than evidence provided on data collected during the data collection phase of the study.


In this section, key studies that focused on the effect of classroom-based physical activity on on-task behavior are examined. Although various methods of quantifying on-task behavior are used in these studies (e.g., various durations of observation intervals, various types of group comparisons), findings are quite consistent. No published studies have documented a negative effect of physical activity on directly observed attentional behavior in the classroom. All studies, in one way or another, demonstrate a beneficial effect of classroom-based physical activity compared with traditional sedentary instruction on directly observed attentional classroom behavior. In one of the first studies to examine the effect of classroom-based physical activity on student attentional processes, Mahar et al. (11) conducted a multiple-baseline across-classrooms design to examine the effect of Energizers classroom-based physical activities on physical activity and on-task behavior. Observations were made under baseline conditions for 4 wk for two classes and for 8 wk for two classes. The classes began participation in Energizers activities after the baseline period and continued participation in at least one Energizers activity per day for the remainder of the 12-wk study.

On-task behavior was directly observed during academic instruction for 30 min before and for 30 min after participation in Energizers activities. During the baseline period before Energizers activities were implemented, on-task behavior was directly observed for 30 min before and for 30 min after a break from observation was taken. Observations of individual students were made for 10 s using momentary time sampling. Teachers continued with academic instruction during the break in observation during the baseline period. The break in observation corresponded to the time when Energizers were performed during the intervention period.

The multiple-baseline design is a common design for behavior modification studies (e.g., 21,22), but this approach has not been duplicated in studies designed to examine the effect of classroom-based physical activity on on-task behavior. The multiple-baseline design allows for development of graphs of on-task behavior across time (12 wk in this study), with replication over multiple classrooms. A sample graph of observation data for one class is provided in Figure 1. The baseline and intervention periods are delineated on the graphs to facilitate visual inspection of on-task behavior before and after the classroom-based physical activity intervention is systematically applied to a classroom. During the baseline period when no physical activity was performed, visual inspection of the graphs demonstrated an inconsistent pattern over weeks 1–4. This indicates that no systematic changes in on-task behavior were noted during the baseline period. Conversely, improvements in on-task behavior were seen from observations before (pre-Energizers) and after (post-Energizers) application of classroom-based physical activity across all classes during the intervention period. Assessment during the baseline period allows researchers to demonstrate no timing effect before introducing the intervention. Because the potential effect of timing is likely ruled out, credible evidence that the intervention probably did have an effect is exhibited (15). If evidence of interobserver reliability is provided, this design and visual analysis provides strong and replicated evidence that participation in classroom-based physical activity improves on-task behavior.

Figure 1:
Mean percentage of intervals of on-task behavior during baseline and intervention conditions for one fourth-grade class. Reprinted with permission from Mahar et al. Effects of a classroom-based program on physical activity and on-task behavior. Med Sci Sports Exerc. 2006;38:2086–2094.

In the original Energizers study (11), Mahar et al. (11) quantitatively assessed on-task behavior by calculating mean on-task behavior for four observation periods: before and after the break in observation during the baseline interval and before and after Energizers activities during the intervention period. A significant time×observation period interaction was observed (see Fig. 2). The 8.3% increase in on-task behavior from pre-Energizers activity to post-Energizers activity was significant with a moderate effect size (ES = 0.60). The 3.1% decrease in on-task behavior from preobservation no classroom-based physical activity to postobservation no classroom-based physical activity was not significant and the ES was small (ES = −0.20). Further analyses were conducted to determine whether classroom-based physical activity had a greater effect on on-task behavior for the least on-task students. Because no assessment of attention deficit was made in this study, the least on-task students were identified as those who were on-task less than 50% of the time at baseline. Analysis of these students indicated a significant and large improvement (ES = 2.20) in on-task behavior after the implementation of classroom-based physical activity, representing an improvement in on-task behavior of approximately 20%. This provided some of the first preliminary evidence that classroom-based physical activity is effective for the students who need it the most (i.e., the least on-task students), and this finding has been supported in subsequent research (23).

Figure 2:
Mean percentage of on-task behavior during academic instruction time for the four observation periods. (Graph created based on data from Mahar et al. Effects of a classroom-based program on physical activity and on-task behavior. Med Sci Sports Exerc. 2006;38:2086–2094.)

Although teachers are presumably influenced by the evidence related to academically related behavior in schools, it is also important to document the amount of physical activity in which students engage to allow for comparisons with other studies, possible evaluation of dose–response relationships, and policy recommendations. To assess physical activity during Energizers activities, pedometer steps accumulated were recorded before the Energizers activities were performed, after the Energizers activities were completed, and at the end of the school day. The intervention group that participated in Energizers activities took significantly more school day steps than the control group that did not participate in Energizers. The ES of the difference in school day steps between groups was moderate (ES = 0.49). In support of the assertion that all groups of students benefit, Calvert et al. (24) reported no gender disparity in the number of steps taken during classroom-based physical activities (including Energizers) for students in grades 1–5.

More recent studies have examined large samples of classrooms and students with strong research designs. Bartholomew et al. (25) provided robust evidence for the effect of active learning on on-task behavior. They implemented a cluster randomized control trial in which schools were randomly assigned to one of two active learning conditions or a control condition that consisted of traditional seated instruction. Teachers in the intervention schools were trained to lead active lessons from the Texas Initiative for Children’s Activity and Nutrition (I-CAN!) program. I-CAN! activities include the use of physically active games to teach new information or practice factual information in either math or language arts. The activities are designed to take 10–15 min and to engage students in moderate-intensity activity, although it is recognized that the specific activity chosen and the classroom teacher both affect how intensely an activity is performed. Results demonstrated a moderate effect of the I-CAN! intervention on school day steps (ES = 0.44), a result similar to previous findings (11,26,27). Researchers in the I-CAN! intervention also assessed the effect of classroom-based physical activity on on-task behavior. On-task behavior was directly observed for 15 min before and 15 min after the I-CAN! activities or control teaching condition using a momentary time sampling technique. Observations of individual students were made for 5 s. The individual observation was coded as “on task” when the student was deemed on-task for the entire 5 s. The observation was coded as “off task” if the student was deemed off-task for any part of the 5-s interval. The main finding was that students in the intervention condition spent significantly more time on task than students in the traditional control condition. The size of this effect (ES = 0.32) was consistent with the moderate effect (ES = 0.60) reported by Mahar et al. (11).

Bartholemew et al. (25) demonstrated that higher levels of physical activity (quantified by steps during the intervention) were associated with higher levels of on-task behavior in the intervention group. This relationship between physical activity and on-task behavior, however, was not found when physical activity was quantified as minutes of moderate to vigorous physical activity (MVPA) assessed via accelerometer, decreasing confidence in the existence of a dose–response relationship between classroom-based physical activity and on-task behavior.

Another important finding from multiple studies was that classroom-based physical activity had similar positive effects on subgroups of students. For example, Bartholomew et al. (25) reported that the positive relationship between classroom-based physical activity and on-task behavior was similar for sex, race/ethnicity, and socioeconomic status subgroups. Grieco et al. (28) demonstrated that changes in on-task behavior found over conditions of various intensities (sedentary physical activity to MVPA) were not related to aerobic fitness, body mass index (BMI), or sex categories. The overall conclusion from these findings is that classroom-based physical activity can improve on-task behavior for all subgroups of students.

Donnelly and colleagues conducted two notable intervention studies that have made a substantial contribution to the literature on classroom-based physical activity and on-task behavior: Physical Activity Across the Curriculum (PAAC [29,30]) and Academic Achievement + Physical Activity Across the Curriculum (A + PAAC [31]). The A + PAAC intervention was a 3-yr cluster randomized intervention, in which on-task behavior was assessed via direct observation in the classroom using momentary time sampling with 10-s observation intervals. Students were observed for 5 to 20 min before and for 5 to 20 min after an academic lesson (delivered either via A + PAAC activities or the control traditional teaching condition). On the basis of the strong observational design and large sample size (nine intervention schools and eight control schools) followed for 3 yr, with process measures to evaluate the fidelity to the planned intervention, the authors reported convincing results. In general, students in this sample (>500 students in year 1) had high rates of on-task behavior before and after the physical activity intervention. At years 1, 2, and 3, students in the A + PAAC intervention had an increase in on-task behavior of 4.8%, 2.4%, and 2.7%, respectively, after physical activity participation. Figure 3 shows visually that small increases in percent of time spent in on-task behavior were seen for all 3 yr for the A + PAAC intervention group. Conversely, small decreases in percent of time spent in on-task behavior were seen for all 3 yr for the control group. Although changes in on-task behavior appear small, the magnitude of the increases each year for the A + PAAC intervention group is important relative to the comparatively high levels of on-task behavior before A + PAAC activities were introduced. Another finding from this investigation that hints toward a possible dose–response relationship between physical activity and on-task behavior was that the percent of time spent in MVPA during A + PAAC activities was significantly associated with the percent of time spent in on-task behavior, a finding that may have policy implications.

Figure 3:
Changes in time spent in on-task behavior for intervention and control conditions over 3 yr. (Graph created based on data from Szabo-Reed et al. Impact of 3 years of classroom physical activity bouts on time-on-task behavior. Med Sci Sports Exerc. 2017;49:2343–50.)

Most classroom-based physical activity programs were designed to promote MVPA in students. Ma et al. (32,33) developed high-intensity interval physical activity breaks they called “FUNtervals.” These activities are characterized by 20 s of high-intensity activity followed by 10 s of rest. This pattern is repeated eight times; hence, the entire interval activity lasts 4 min. Effects of this activity on off-task behavior were evaluated in one class of second-grade students and one class of fourth-grade students in a single-group, repeated crossover design. For the 2-wk intervention, days on which FUNtervals were implemented and days on which no activity was implemented were alternated. Although the FUNtervals activity itself lasted 4 min, the intervention time lasted approximately 10 min (i.e., setup, FUNtervals, and preparation for return to regular classroom activity). The no activity break consisted of a 10-min lecture by the research team on a topic unrelated to the classroom lesson (e.g., healthy eating, physical activity, and history of sport).

Off-task behavior was defined as “behavior that is disengaged from and unrelated to the learning task at hand” (p. 1333) and was assessed with the Behavioral Observation of Students in Schools instrument (34). This instrument divides academic engagement into (a) active engagement or (b) passive engagement. When the observed student is disengaged from appropriate academic behavior, the observed behavior is categorized as off-task behavior. Off-task behavior can be subdivided into “off-task motor,” “off-task verbal,” and “off-task passive.” Off-task behavior was recorded using partial interval sampling, indicating that off-task behavior was recorded if it occurred at any time during the observation interval. Ma et al. (32) indicated that individual students were observed for 30 s followed by a 15-s break. Observers recorded the occurrence and duration of off-task behavior during the 30-s interval. Their study protocol included a secondary observer to allow estimates of interobserver reliability during the data collection phase of the study, which were high (≥0.90 for both classrooms). This is consistent with others who have reported high interobserver reliability with this instrument (35).

Figure 4 shows significantly lower mean percent off-task behavior (i.e., better classroom behavior) for grade 2 students after participation in FUNtervals compared with the control no activity condition. Numbers above the bars represent the difference in percent off-task behavior for FUNtervals intervention versus control no activity conditions for each grade level for each type of off-task behavior. Ma et al. (32) reported ES values for grade 2 comparisons ranging from moderate (ES = 0.45) to large (ES = 1.08). For grade 4 comparisons, students exhibited significantly lower mean passive off-task behavior and motor off-task behavior after participation in FUNtervals compared with the control no activity condition. ES values for grade 4 comparisons were small to moderate (ES = 0.31 for passive off-task comparison and ES = 0.48 for motor off-task comparison). Verbal off-task behavior was not significantly different between FUNtervals and control no activity conditions and the ES was small (ES = 0.20).

Figure 4:
Changes in time spent in three types of off-task behavior for two grade levels for intervention and control conditions. (Graph created based on data from Ma et al. Classroom-based high-intensity interval activity improves off-task behavior in primary school students. Appl Physiol Nutr Metab. 2014;39:1332–37.)

Similar to the results of Mahar et al. (11), Ma et al. (32) demonstrated that effects of the physical activity condition were stronger for students who were most off-task at baseline. Ma et al. categorized students in the highest quartile of off-task behavior for each grade level as “high off-task.” Differences for high off-task students between FUNtervals and control no activity conditions were typically more than twice as large as differences between conditions for all students. This is welcome news for teachers who wish to incorporate physical activity into the classroom in an effort to engage students in academic-related tasks.

Ma et al. (32) recognized their small sample (two classrooms) and relatively small sample size (24 students in one classroom and 20 students in a second classroom) as limitations, but these findings considered in combination with findings of others (11,31,36) provide convincing evidence that classroom-based physical activity benefits academic task engagement for all students but may work better for students who need it the most (i.e., students who are least on-task or least academically engaged). Because academically off-task students are also the students who are often disruptive in the classroom, teachers who incorporate classroom-based physical activity can expect more academically focused students who are less disruptive.

Several other research teams have provided evidence for the beneficial effects of classroom-based physical activity on on-task behavior (36,37). Mullender-Wijnsma et al. (37) directly observed on-task behavior (5-s observation intervals) in 81 second-grade and third-grade students after regular classroom lessons and after “Fit & Vaardig op school” (F&V; fit and academically proficient at school) lessons. F&V lessons were physically active academic lessons in which students stood near their desks and performed basic and specific exercises for 10–15 min. Basic exercises included marching, jogging, or hopping in place. Specific exercises were conducted in response to a specific academic task (e.g., jumping the number of times that answered a multiplication question). Observations were made at the start, midway, and end of the 22-wk intervention, and F&V lessons were held three times per week. On-task behavior was significantly higher after participation in F&V lessons (postintervention) compared with after regular inactive lessons (postcontrol) midway through the intervention and at the end of the intervention, but not at the start of the intervention. The ES values of the mean differences at the midway point (ES = 0.60) and at the end of the intervention (ES = 0.59) were moderate.

Grieco et al. (36) examined the effects of classroom-based physical activity and BMI on on-task behavior. Students were directly observed for 15 min before and after both physically active classroom lessons and inactive control lessons. On-task behavior was assessed with momentary time sampling using 5-s observation intervals. Interobserver reliability was estimated before the study and at a 3-month follow-up assessment but was not estimated during the data collection phase of the study. Significant decreases in on-task behavior were noted for students in all BMI categories after inactive control lessons. The overweight group had a greater magnitude of effect (i.e., greater decrease in on-task behavior) after inactive control lessons compared with the normal weight group. Small (ES ranged from 0.13 to 0.26), nonsignificant improvements in on-task behavior were recorded after physically active lessons. The authors noted that observations taken for only 15 min after the lessons was a study limitation that could have underestimated the effects of classroom-based physical activity on on-task behavior. Within the limitations of their study, Grieco et al. (36) concluded that physically active academic lessons may prevent the reduction in on-task behavior seen after traditional inactive lessons.

Take 10! is another program that integrates 10-min physical activities into academic lessons. Goh et al. (38) evaluated the effect of Take 10! on on-task behavior in nine classes (third, fourth, and fifth grades). On-task behavior was directly observed before and after participation in Take 10! activities using momentary time sampling with 5-s observation intervals. After 4 wk of baseline where students did not receive the Take 10! intervention, classroom teachers implemented the activities for the following 8 wk. Classes were observed during the last week of baseline and during the last 4 wk of the intervention period for 30 min before and 30 min after activity. The researchers documented high levels of interobserver agreement (i.e., 96%), with reliability observations taken during 50% of observation sessions. During baseline, observations were made for 30 min during classroom instruction. Then a 15-min break in observation was taken, corresponding to the time Take 10! was delivered during the intervention period. This was followed by another 30 min of observation. During baseline, a significant decrease in on-task behavior of 7.7% was reported from preobservation no activity to postobservation no activity. A significant increase in on-task behavior of 7.2% was noted after participation in Take 10! These results are in agreement with those reported by others (11,31,32,36).

The final study presented here is notable because the authors reported the effects of different durations of classroom-based physical activity (i.e., Brain BITES) on on-task behavior in two fourth-grade and two fifth-grade classrooms (39). All students were observed before and after 10 min of seated classroom activity and before and after 5, 10, and 20 min of classroom-based physical activity. Video cameras were placed in front of the classroom to record classroom behavior. On-task behavior was later observed from the video recordings by noting the direction of the students’ gaze using 15-s observation intervals. Interobserver agreement was fair (81%), but the authors reported that disagreements between observers were reevaluated to reach consensus. The researchers reported that on-task behavior was assessed while students performed paper-and-pencil tests, which may differ from regular academic instruction led by a teacher. Activities were led by research staff rather than by classroom teachers. Although this may limit the generalizability of results relative to authentic classroom experiences led by classroom teachers, it did allow the researchers to have greater control over the actual duration of the physical activity, enhancing fidelity to the prescribed dose of the classroom-based physical activities. On-task behavior assessed after the four conditions was compared, after adjusting for on-task behavior assessed before the implementation of the intervention and other covariates. The authors reported a significant improvement in on-task behavior for all four classrooms combined after the 10-min dose of activity, with a moderate ES (0.50) compared with the seated condition, which is similar to results reported by others (11,25,32,37). The nonsignificant improvement in on-task behavior after the 20-min dose of activity had a small to moderate effect (ES = 0.32) compared with the seated condition. On-task behavior did not change after the 5-min dose of activity, relative to the seated condition. When examined separately, average on-task behavior of individual classrooms had variable responses to the intervention conditions, emphasizing the need for further study relative to the optimal dose of physical activity for most students.


The relatively consistent finding that classroom-based physical activity results in better on-task behavior than sedentary seated lessons has important implications for school policy. It is likely that teacher-administered physically active lessons will increase on-task behavior relative to traditional sedentary lessons and that the increased on-task behavior will result in enhanced learning. This is a message worth amplifying and disseminating to educational practitioners from school superintendents to teachers.

To enhance the use of classroom-based physical activity by teachers and further develop the research base in this area, the following recommendations are offered:

  1. The message that incorporation of classroom-based physical activity enhances on-task behavior that will ultimately result in better learning should be delivered clearly and unambiguously to school personnel.
  2. University teacher preparation programs should consider training new teachers to deliver lessons via classroom-based physical activity.
  3. Classroom-based physical activities should be made available for free or at a low cost to promote inclusion of schools with minimal financial resources.
  4. Physical education teachers in schools should be used as a resource to help classroom teachers lead physically active lessons.
  5. Teachers are a key to uptake of classroom-based physical activities. Teacher perceptions of classroom-based physical activity and perceived barriers to the implementation of classroom-based physical activity should be considered and addressed.
  6. Additional research is needed to examine the effect of different intensities of movement on attention (i.e., dosage of physical activity).
  7. Further research should investigate the duration of physical activity needed to result in substantial improvements in on-task behavior.
  8. Research is needed to examine how long the positive effect of classroom-based physical activity on attention lasts after the bout of physical activity is performed.
  9. Research to understand teachers’ perceived barriers to the implementation of classroom-based physical activity is essential.
  10. Examination of the effects of classroom-based physical activity on subgroups of students needs additional work.
  11. The effect of school policies to promote classroom-based physical activity on uptake and improved learning should be examined.

Although further research is necessary to better understand the potential differential effectiveness of various intensities of classroom-based physical activity, the accumulated evidence suggests that on-task behavior of students can be improved with activity from low to moderate intensity (11,25,28,31) and with activity of vigorous intensity (32). At this point, the message to classroom teachers need not be complicated by recommendations for different intensities of physical activity or by whether specific academic content is included in the delivery; current research evidence suggests that the implementation of classroom-based physical activity will improve on-task behavior for all students relative to traditional sedentary classrooms.

The views of this article do not constitute endorsement by the American College of Sports Medicine. The author has no conflicts of interest to disclose.


1. Biddle S, Gorely T, Stensel D. Health-enhancing physical activity and sedentary behaviour in children and adolescents. J Sports Sci. 2004;22(8):679–701.
2. Must A, Tybor DJ. Physical activity and sedentary behavior: a review of longitudinal studies of weight and adiposity in youth. Int J Obes (Lond). 2005;29(Suppl 2):S84−96, S96.
3. U.S. Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. Washington (DC): U.S. Department of Health and Human Services; 2008.
4. Institute of Medicine. Educating the Student Body: Taking Physical Activity and Physical Education to School. Washington (DC): National Academies Press; 2013.
5. U.S. Department of Education, National Center for Education Statistics. Back to School by the Numbers; 2018 [cited 8 Nov 2018]. Available from:
6. Best JR, Miller PH. A developmental perspective on executive function. Child Dev. 2010;81:1641–60.
7. Crosnoe R, Johnson MK, Elder GH. School size and the interpersonal side of education: an examination of race/ethnicity and organizational context. Soc Sci Q. 2004;85:1259–74.
8. Duckworth AL, Quinn PD, Tsukayama E. What No Child Left Behind leaves behind: the roles of IQ and self-control in predicting standardized achievement test scores and report card grades. J Educ Psych. 2012;104(2):439–51.
9. Godwin KE, Almeda MV, Seltman H, et al. Off-task behavior in elementary school children. Learning and Instruction. 2016;44:128–43.
10. Mahar MT, Kenny RK, Shields AT, Scales DP, Collins G. Energizers: Classroom-Based Physical Activities. Raleigh (NC): North Carolina Department of Public Instruction; 2004. Available from:
11. Mahar MT, Murphy SK, Rowe DA, Golden J, Shields AT, Raedeke TD. Effects of a classroom-based program on physical activity and on-task behavior. Med Sci Sports Exerc. 2006;38:2086–94.
12. Helme S, Clarke D. Identifying cognitive engagement in the mathematics classroom. Math Educ Res J. 2001;13:133–53.
13. Peterson PL, Swing SR. Beyond time on task: students’ reports of their thought processes during classroom instruction. Elem School J. 1982;82:481–91.
14. Van Houten R, Hall RV. The Measurement of Behavior: Behavior Modification. 3rd ed. Pro-ed, Inc: Austin (TX); 2001.
15. Mahar MT. Impact of short bouts of physical activity on attention-to-task in elementary school children. Prev Med. 2011;52(Suppl 1):S60–4.
16. McKenzie TL. Seeing is believing: observing physical activity and its contexts. Res Q Exerc Sport. 2010;81:113–22.
17. Bartholomew JB, Jowers EM, Roberts G, Fall A-M, Errisuriz VL, Vaughn S. Active learning increases children's physical activity across demographic subgroups. Trans J Am Col Sports Med. 2018;3(1):1.
18. Frederick WC, Wahlberg HJ. Learning as a function of time. J Educ Res. 1980;73:183–94.
19. Greenwood CR, Horton BT, Utley CA. Academic engagement: current perspectives in research and practice. School Psych Rev. 2002;31:328–49.
20. Karweit N, Slavin RE. Measurement and modeling choices in studies of time and learning. Am Educ Res J. 1981;18:157–71.
21. Katz RC, Singh NN. Increasing recreational behavior in mentally retarded children. Behav Modif. 1986;10(4):508–19.
22. Shimabukuro SM, Prater MA, Jenkins A, Edelensmith P. The effects of self-monitoring of academic performance on students with learning disabilities and ADD/ADHD. Educ Treatment Child. 1999;22:397–414.
23. Ma JK, Le Mare L, Gurd BJ. Classroom-based high-intensity interval activity improves off-task behavior in primary school students. Appl Physiol Nutr Metab. 2014;39:1332–7.
24. Calvert HG, Mahar MT, Flay B, Turner L. Classroom-based physical activity: minimizing disparities in school-day physical activity among elementary school students. J Phys Act and Health. 2018;15:161–8.
25. Bartholomew JB, Golaszewski NM, Jowers E, et al. Active learning improves on-task behaviors in 4th grade children. Prev Med. 2018;111:49–54.
26. Erwin HE, Abel MG, Beighle A, Beets MW. Promoting children's health through physically active math classes: a pilot study. Health Promot Pract. 2011;12(2):244–51.
27. Oliver M, Schofield G, McEvoy E. An integrated curriculum approach to increasing habitual physical activity in children: a feasibility study. J Sch Health. 2006;76(2):74–9.
28. Grieco LA, Jowers EM, Errisuriz VL, Bartholomew JB. Physically active vs. sedentary academic lessons: a dose response study for elementary student time on task. Prev Med. 2016;89:98–103.
29. Donnelly JE, Greene JL, Gibson CA, et al. Physical activity and academic achievement across the curriculum (A + PAAC): rationale and design of a 3-year, cluster-randomized trial. BMC Public Health. 2013;13:307.
30. Donnelly JE, Hillman CH, Greene JL, et al. Physical activity and academic achievement across the curriculum: results from a 3-year cluster-randomized trial. Prev Med. 2017;99:140–5.
31. Szabo-Reed AN, Willis EA, Lee J, Hillman CH, Washburn RA, Donnelly JE. Impact of 3 years of classroom physical activity bouts on time-on-task behavior. Med Sci Sports Exerc. 2017;49:2343–50.
32. Ma JK, Le Mare L, Gurd BJ. Classroom-based high-intensity interval activity improves off-task behavior in primary school students. Appl Physiol Nutr Metab. 2014;39:1332–7.
33. Ma JK, Le Mare L, Gurd BJ. Four minutes of in-class high-intensity interval activity improves selective attention in 9- to 11-year olds. Appl Physiol Nutr Metab. 2015;40(3):238–44.
34. Shapiro ES. Academic Skills Problems Workbook. New York: Guildford; 1996.
35. Ota KR, DuPaul GJ. Task engagement and mathematics performance in children with attention-deficit hyperactivity disorder: effects of supplemental computer instruction. Sch Psych Q. 2002;17:242–57.
36. Grieco LA, Jowers EM, Bartholomew JB. Physically active academic lessons and time on task: the moderating effect of body mass index. Med Sci Sports Exerc. 2009;41(10):1921–6.
37. Mullender-Wijnsma MJ, Hartman E, de Greeff JW, Bosker RJ, Doolaard S, Visscher C. Moderate-to-vigorous physically active academic lessons and academic engagement in children with and without a social disadvantage: a within subject experimental design. BMC Public Health. 2015;15(1):404.
38. Goh TL, Hannon J, Webster C, Podlog L, Newton M. Effects of a TAKE 10! classroom-based physical activity intervention on third- to fifth-grade children’s on-task behavior. J Phys Act and Health. 2016;13:712–8.
39. Howie EK, Beets MW, Pate RR. Acute classroom exercise breaks improve on-task behavior in 4th and 5th grade students: a dose–response. Ment Health Phys Act. 2014;7:65–71.
Copyright © 2019 by the American College of Sports Medicine