See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/288836831

The impact of 10-min activity breaks outside the classroom on male students'
on-task behaviour and sustained attention: a randomised crossover design

Article  in  Acta Paediatrica · December 2015

DOI: 10.1111/apa.13323

CITATIONS READS

45 1,009

5 authors, including:

Andrew Wilson Tim Olds

University of South Australia University of South Australia
10 PUBLICATIONS   159 CITATIONS    432 PUBLICATIONS   21,268 CITATIONS   

SEE PROFILE SEE PROFILE

Kurt Lushington Jim Dollman

University of South Australia University of South Australia
160 PUBLICATIONS   5,043 CITATIONS    169 PUBLICATIONS   4,205 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Rural Environments and Community Health View project

Health of Young Victorians Study View project

All content following this page was uploaded by Andrew Wilson on 27 October 2016.

The user has requested enhancement of the downloaded file.

 Acta Pædiatrica ISSN 0803-5253

REGULAR ARTICLE

The impact of 10-minute activity breaks outside the classroom on male

students’ on-task behaviour and sustained attention: a randomised
crossover design
Andrew N. Wilson ([email protected])1, Timothy Olds1, Kurt Lushington2, John Petkov3, James Dollman1
1.Samson Institute for Health Research – Alliance for Research in Exercise, Nutrition and Activity (ARENA), Adelaide, SA, Australia
2.School of Psychology, Social Work and Social Policy, University of South Australia, Adelaide, SA, Australia
3.Centre for Regional Engagement, University of South Australia, Adelaide, SA, Australia

Keywords ABSTRACT
Elementary students, On-task behaviour, Physical Aim: The aim was to evaluate the impact of a brief activity bout outside the classroom on
activity, School based, Sustained attention
boys’ attention and on-task behaviour in the classroom setting.
Correspondence
Methods: Fifty-eight boys (mean age 11.2  0.6 years) were recruited from a boys’
Dr Andrew Wilson, Division of Health Science,
University of South Australia, City East Campus, elementary school in Adelaide, South Australia. Two year 5 and, similarly, two year 6 classes
P6-15. Adelaide, SA 5001, Australia. were assigned using a crossover design to either four weeks of a 10 minute Active Lesson
Tel: +61881693900 |
Break followed by four weeks of a 10 minute Passive Lesson Break (reading) or visa versa.
Email: [email protected]
Attention was quantified using a computerised psychomotor vigilance task, and on-task
Received
behaviour by direct observation.
10 July 2015; revised 2 November 2015;
accepted 28 December 2015. Results: Neither the Active Lesson nor the Passive Lesson condition significantly affected
DOI:10.1111/apa.13323
sustained attention or on-task behaviour, and there were no significant differences
between conditions.
Conclusion: There was no impact on participants’ sustained attention or on-task behaviour
after a short activity break between lessons. Brief activity breaks outside the classroom do
not compromise participants’ on-task behaviour or attention levels upon returning to the
classroom, although improvement in these variables is not seen either. However, the
results suggest that active breaks are effective for accruing moderate-to-vigorous physical
activity without compromising classroom behaviours.

INTRODUCTION increased physical activity (improved musculoskeletal

Increasing children’s daily physical activity to recom- health or reduced hypertension, etc.) to school administra-
mended levels is seen as important in children’s overall tors have had little success in the past (3).
health and well-being (1,2). Compulsory schooling in many Along with the potential health benefits of increasing
Western countries means that students spend a large opportunities for more moderate-to-vigorous physical activ-
proportion of their childhood and adolescence in the ity (MVPA), there is growing interest among researchers
school system. Students from all cultural and economic and educational professionals regarding the impact of
backgrounds are therefore in a position to be exposed to the physical activity on cognitive development in young people
same health messages and intervention programmes (1), (4,5). However, current literature in this field provides little
making schools ideal locations in which to promote compelling evidence. A review of reviews by Biddle and
physical activity (2). Improving students’ daily physical Asare (4) suggested that physical activity may be associated
activity has been achieved in previous school-based inter-
ventions (2), so the aim of increasing physical activity levels
is feasible. However, efforts to ‘sell’ the health benefits of
Key notes
 Brief active breaks during class time may have both
Abbreviations educational and health benefits.
ALB, Active lesson break; AM7164, Model of Actigraph  This study found that brief active classroom breaks had
accelerometer; CV, Coefficient of variation; I-CAN!, Initiatives no impact on cognitive attention and classroom
for children’s activity and nutrition; MET, Metabolic equivalent behaviours among elementary school boys.
of task; MP3, Moving picture experts group layer-3 audio; MVPA,  Future research should evaluate the effectiveness of
Moderate-to-vigorous physical activity; PDA, Personal digital brief active bouts on classroom behaviours among
assistant; PLB, Passive lesson break; PVT, Psychomotor vigilance
children who are prone to inattention.
task; RT, Response time; VPA, Vigorous physical activity.

©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 1
Short activity breaks and boys attention Wilson et al.

with improved cognitive performance, but effect sizes academic outcomes such as grades and test results (15) or
were generally small and inconsistent. Reviews by other proximal behaviours such as on-task behaviour (11,13). As
researchers like Erwin et al. (2) suggested that inconsisten- grades and test results can vary according to measurement
cies in cognitive outcomes or lack of true experimental protocols, the teacher and/or the content taught, this study
evidence makes it difficult to draw any conclusions from focused on objective behaviour and cognitive measures
the literature. Other reviews cautiously suggested positive associated with academic learning – on-task (or off-task)
findings for cognitive performance (3,6), but expressed behaviour and sustained attention.
concerns about inconsistencies in study design and out- Off-task behaviour is perceived to be problematic in a
come measures. school setting, due to the potential to limit learning
Equivocal findings also exist for academic performance opportunities (16). Further to this, Salmeron and Garcia
outcomes, particularly in relation to effect sizes. The review (17) reported a link between sustained attention and
by Singh et al. (7) reported strong evidence of positive reading, suggesting it was essential for reading printed text
relationships. Erwin et al. (2) reported that in classroom- due to the incremental nature of comprehension. The
based studies, although limited in number, the impact of implication was that failure to attend to or identify impor-
physical activity on learning outcomes was positive, signif- tant ideas at one point in the text may limit comprehension
icant and moderate in size. A review by Biddle and Asare of forthcoming text. Previous studies (11,13) have shown
(4) reported weaker associations. As with the cognitive that classroom-based physical activity has improved on-
outcomes, caution is warranted when interpreting these task behaviour in elementary students, while literature
findings due to inconsistencies in methodological design reviews suggest that physical activity not only improves
among the studies. Putting methodological issues aside, all classroom behaviour but also the ability of the student to
reviews reported little or no evidence that allocating better attend to and concentrate on the task at hand (4,18).
additional time to physical activity during the school day Moreover, as active curriculum strategies require exten-
detracted from academic performance. The systematic sive planning and therefore add substantially to teachers’
review by Rasberry et al. (5) into school-based physical workloads, the current study focused on the impact of short,
activity and academic or cognitive performance reported discrete bouts of physical activity that were not integrated
negative associations for only 1.5% of the 251 associations into learning activities. In summary, the aim of this study
reviewed. Overall, 50.5% of all associations examined in was to evaluate the impact of a brief physical activity bout
this review were positive. (10-minute ALB) outside the classroom on boys’ attention
While recent reviews (7) cautiously support a positive and on-task behaviour in an Australian elementary school.
relationship between physical activity and classroom Based on the extant literature, it is hypothesised that the
behaviour, cognitive performance and academic achieve- ALB will improve these measures relative to a passive
ment, obstacles remain for long-term implementation of control condition.
interventions in schools, such as budgetary implications (6);
perceived skill level of classroom teachers to administer
physical activity interventions (1); and competition for time PARTICIPANTS AND METHODS
in the curriculum. In many countries, physical education Sample
(PE) is diminishing in the curriculum (8), with this trend Research suggests that boys are more prone than girls to
often attributed to increasing pressures from vocation- inattention the longer the duration of academic work
centred learning areas and the associated standardised without a break (19). A convenience sample of 58 boys
testing (8). This is likely to be due to the perception of some (mean age 11.2  0.6 years) was therefore recruited from
administrators and teachers that time devoted to an two year 5 (mean age 10.9  0.7 years) and two year 6
academic task is directly related to success in that learning classes (mean age 11.5  0.3 years), in a boys’ elementary
area (9). Finally, much of the formal physical activity in school in Adelaide, South Australia. Effect sizes from
elementary schools is supervised by classroom teachers, previous studies of physical activity bouts and on-task
many of whom do not have specialist training in PE (1), behaviour are d = 0.6 (11) and 0.13 (13). Based on these
potentially leading to lower confidence or willingness to effect sizes, an effect size of f = 0.2 (d = 0.4) was proposed
teach this component of the curriculum (1). Funding for this study. With a power of 80% and an alpha of 0.05, a
constraints compound the issue by limiting access to sample size of 52 was required. Approval for this study was
specialist PE teachers and equipment. granted by the University of South Australia’s Human
There is growing interest in the feasibility and impact of Research Ethics Committee and the governing body’s
small bouts of supervised activity during class time (9–14) to educational office. Parents and participants gave informed
increase students’ physical activity and academic outcomes. consent to participate in the study.
These brief (generally 10–15 minutes) episodes of physical
activity begin to address the limitations of time, cost and The intervention
teacher expertise that plague PE or sport-mediated inter- Active lesson break
ventions. Much of the current research in this field has Active Lesson Breaks in the context of this study were 10-
focused on the impact of integrating physical activity into minute bouts of physical activity administered by the class-
the curriculum (‘active curriculum’) on distal student room teacher outside the regular classroom environment.

2 ©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
Wilson et al. Short activity breaks and boys attention

They were designed to minimise the extra work for the Passive lesson break
teacher and the time impact on the curriculum. The 10- A Passive Lesson Break (PLB) was included as the control
minute duration was based on previous strategies such as condition during which participants sat outside their class-
‘Take 10!â’ (10) and ‘Energizers’ (11), which were devel- room for 10 minutes and read a self-chosen book for
oped to improve opportunities for participants to increase 10 minutes. The distance the participants walked was no
their daily physical activity. Teachers were provided with a more than 20 m from their classroom.
menu of activities from which they could develop each
ALB. The compendium of activities was designed by a Testing protocol
registered physical educator (first author) to limit equip- The study utilised a crossover design, with all four class
ment requirements to four to six cones (markers) and to groups participating in the Active and Passive Lesson
maximise individual participation. Activities included tag/ Breaks. One class from each year level was randomly
chasing games or invasion-type games (a summary of the assigned, using a web-based randomising program
activities is available from the corresponding author). As (www.randomizer.org), to complete four weeks of the
the emphasis was on active involvement rather than skill ALBs, followed by four weeks of the PLBs (Table 1). The
development, specialist PE training for teachers was not remaining two classes completed the study in the reverse
assumed. Prior to the intervention, teachers were trained to order. All four classes had a two-week “washout” period
administer the ALBs by an experienced physical educator between conditions (coinciding with the school’s two-week
(first author). The ALBs were administered once per day for term holidays). To minimise the influence of temperature
three out of the five school days (Monday, Tuesday and and ventilation on academic performance, work output and
Thursday) over a period of four weeks. reaction time measures, the ambient temperatures and
The intensity of physical activity during the ALBs was airflow in the classrooms were controlled during the eight-
assessed using the Actigraph AM7164 (Pensacola, FL, USA). week study using the room’s climate control units (22°C).
The validity of the AM7164 has previously been reported as The four classrooms are located next to each other in a row
acceptable in children (20). The Actigraph is a uniaxial and have the same architectural design and lighting. The
accelerometer designed to detect normal human motion curriculum units taught during this testing period were the
while excluding high-frequency stimuli from other sources. same within each year level.
The filtered signal is digitised, with the magnitude summed
over a user-specified time. The cycle (epoch) in this study Outcome measures
was set at 15 seconds. An epoch of 15 seconds has the Sustained attention
capacity to capture short intermittent bursts of activity Sustained attention was recorded before and after Active
characteristic of young children (20). All participants were and Passive Lesson Breaks using the five-minute psychomo-
given instructions on the correct wearing of the device tor vigilance task (PVT). The PVT measures sustained
(right hip, just above iliac crest) by the researcher. The attention based on repeated reaction time trials over a
participants then wore the devices on an elastic belt for a period of five minutes (21), with the change in mean
trial period of one day before the intervention to ensure that reaction time (mean RT) representing mental fatigue (22).
they were familiar with the correct procedures. Delayed responses (lapses) to these reaction time trials are
All class groups wore the Actigraph AM7164 accelerom- used to measure changes in behavioural alertness (23). The
eters during the third week (five school days) of the ALB five-minute PVT software (24) developed by the Walter
intervention to monitor whether the three ALBs during that Reed Army Institute of Research (USA) was ported onto a
week were being conducted at a suitable intensity (MVPA). personal digital assistant (PDA) using a Palm (Palm Inc,
This enabled the measurement of 3 9 10 minutes of each Sunnyvale, CA, USA) operating system. The mean response
participant’s physical activity during the ALBs of that week time (MRT) (ms) to a repeating visual stimulus (black and
(Table 1). white bull’s eye target) presented randomly at one- to

Table 1 Timeline of outcome measurement

Week 1 (7) Week 2 (8) Week 3 (9) Week 4 (10) Week 5 + 6

Test day M T Thu M T Thu M* T* Thu* M T Thu Washout

Measurement tool Ob 5-min Ob 5-min Ob 5-min Ob 5-min Ob 5-min Ob 5-min Period
Intervention† ALB (PLB) ALB (PLB) ALB (PLB) ALB (PLB)

*Intensity of ALBs measured using accelerometry.

†
Grade 5a + 6a (n = 30) started with the ALB intervention, Grade 5b + 6b (n = 28) started with PLB.
Parenthesis represents second phase of the crossover design (i.e. 7 = week 7) for a particular class.
ALB = 10-min Active Lesson Break; PLB = 10-min Passive Lesson Break (reading).
Ob = % Off-task behaviour (in-class observation); five minutes = Sustained attention measures (5-min PDA-PVT).

©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 3
Short activity breaks and boys attention Wilson et al.

five-second intervals over a five-minute period was used classroom. When the observers had completed the 30 min-
to measure participants’ sustained attention. Responses utes of post-break observation, they removed themselves
≥500 ms were recorded as lapses in attention, and later from the classroom at the nearest break in teaching to
converted to a percentage of total responses. Instructions to minimise disruption. It was not possible to blind observers
participants and three practice trials were delivered in the to the intervention.
week prior to data collection. Three practice sessions were
included on the basis of previous research findings (25) to Inter-observer reliability
familiarise participants with the device and to minimise Prior to the study, all observers underwent rigorous train-
practice effects in the data. This is likely to be a conservative ing. The observers were given definitions of the four
approach, as recent research (25) reported no statistically possible behaviours (one on-task and three off-task) that
significant difference between the second and third prac- they were to record, and shown examples of these four
tices when using the five-minute PDA-PVT. behaviours in a video taken of a younger grade 4 class at the
During the test days (Table 1) when the five-minute PVT same school. Once each observer could correctly identify
was used, the timing of the intervention breaks remained the four behaviours [on-task, motor off-task (out of their
the same. However, five minutes before exiting the class- seat), talking off-task and off-task (other)], they watched a
room, the participants completed the first of two-five- one-hour (2 9 30 minutes) videotape of the grade 4 class
minute PVTs. The second five-minute PVT was started and recorded the behaviour of five nominated participants.
immediately after all participants had returned from the The use of training videos continued until an inter-observer
10-minute Active or Passive Lesson Break. Each participant reliability of ≥ 80% was reached. On two separate days, one
was allocated the same PDA throughout the study. week prior to the beginning of the study, the observers sat in
on the classes to be used in the study and observed the
On-task behaviour on-task behaviour of the participants. The observers sat in
An in-class observer used a 10-second time sampling pairs, with one designated as the primary observer, the
procedure to record the percentage of time participants other the secondary observer. This habituation period
were on-task or off-task for a 30-minute period, both before enabled the participants to become familiar with the
and after an Active or Passive Lesson Break. The observa- observers being in their classroom, as well as providing
tion format and criteria adopted were similar to those used additional in-class inter-observer reliability testing. In total,
in a recent study of participants’ on-task behaviour (11). 25% of the observation periods involved both a primary and
Using a pre-recorded digital MP3 audio file, which directed secondary observer. A two-way audio jack allowed two sets
observers when to observe and when to record, observers of headphones (with long leads to ensure that observers
systematically recorded a nominated participant’s beha- could not see the other observer’s recordings) to operate
viour in four 15-second intervals (inclusive of a five-second from the same MP3 player, ensuring synchronised record-
period to physically record the observation). At the com- ing. The mean agreement between observers during the
pletion of the one-minute observation period, the obser- study was 94 (83–100)%. This figure is compatible with
vation was rotated to the second of five nominated agreements found in similar studies (11, 13).
participants. In the 30-minute observation period, the five
nominated participants were observed 24 times for a total of Statistical analysis
six minutes each. The sustained attention data (mean RT and % of lapses)
During testing days, observers entered an empty class- were positively skewed and were normalised by a logarith-
room 10–15 minutes before the participants to minimise mic transformation. In total, 5.3% of data were missing,
disruption. They sat and waited until five minutes after either through a malfunction of the PDA device or through
teachers started the lesson before beginning observations. participants missing a testing period. The distribution of the
The observations alternated between a morning session missing data was assessed prior to analysis and it did not
(9.00 am) and an afternoon session (after lunch at indicate that there was an increase or decrease as days
1.55 pm). A web-based program (www.randomizer.org) progressed, so was deemed to be a random occurrence, an
was used to randomly select participants who had con- important criterion for mixed-model analysis.
sented to participate in the study for observation from each The data for the percentage of time on-task were
class at the start of the day. Neither the participants nor negatively skewed; reversing the data and adding a value
the teacher in charge of the class was aware of the five of one to each score prior to a logarithmic transformation
participants chosen. Participants’ names were not returned created a more normalised data set. Reversing the data
to the random draw until all consenting participants had resulted in an ‘off-task’ behaviour score rather than the
been observed in the current round, ensuring that everyone usually published on-task score. For example, being on-task
was observed twice on average per condition (active or 80% of the recorded time would mean that the participant
passive). Each participant received at least 24 minutes of was off-task 20% of the time. Off-task rather than on-task
observation per condition (12 minutes pre- and 12 minutes behaviour was used in all subsequent analyses. For the
post-break during the four-week period). The observers observational measure (off-task behaviour), a mean value
remained in the classroom at the completion of the 30- was calculated for each participant for each condition
minute pre-break period when the participants left the both for the pre-break and post-break periods that is, the

4 ©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
Wilson et al. Short activity breaks and boys attention

pre- and post-observation values recorded during the ALB

Table 2 Test of fixed effects from mixed-model analysis (mean reaction time)
period were averaged to give a mean pre-break score and a
Source df F Sig.
mean post-break score for the four weeks of the ALB
intervention. The process was repeated with the passive Intercept 69.80 30708.44 <0.0001
intervention weeks. The design of the study did not allow Condition 1064.01 0.26 0.610
for individual scores to be used to track changes over the Time 61.03 2.19 0.257
four-week period, as complete observation of the subjects Condition9Time 1057.25 0.003 0.955
each week was not possible.
Condition (active vs. passive), Time (pre vs. post).
A mixed-model design was used to analyse the repeated
measures for the five-minute PVT measures (mean RT and
percentage of lapses) and the observational data.
Accelerometer data were assessed using a custom- Table 3 Test of fixed effects from mixed-model analysis (percentage of lapses)
designed program and syntax from Deakin University Source df F Sig.
(Melbourne, Victoria). The intensity of the ALBs was
Intercept 65.59 109.97 <0.0001
expressed as MVPA minutes. The cut-off point for MVPA
Condition 1061.50 0.22 0.642
was set at ≥3 METs (unit of energy expenditure) using an Time 59.28 4.10 0.249
age-specific equation (26). The coefficient of variation (CV) Condition9Time 1059.77 0.22 0.641
was also reported for the ALBs along with other periods
during those days in which participants had opportunities Condition (active vs. passive), Time (pre vs. post).
to be physically active: recess, lunch and after school (until
6.00 pm) to create a brief activity profile. The 10-minute
segment of the data corresponding to the time of the ALB respectively), meaning that no statistically significant dif-
(Mon, Tues and Thurs) was extracted by programming the ference between the two conditions (ALB and PLB) was
download software to specifically target this time period. evident. Accordingly, the study hypothesis is rejected.
Specific periods of time during the day (Recess, Lunch, after
school, ALB) could be targeted because each recorded Off-task behaviour
epoch is time stamped. A valid day constituted ≥10 hours of The pre- and post-break mean values for the observational
wear. Five consecutive minutes (non-zero) indicated the (off-task behaviour) measures are presented in Table 4. The
start of each day, while 20 consecutive minutes indicated mean difference (Table 5) for off-task behaviour for the
non-wear. ALB condition was negligible (d = 0.11); however, for the
PLB condition, the size of the mean difference was small
(d = 0.20). The condition was not statistically significant
RESULTS (p = 0.36). Observations between the PLB and ALB condi-
The response rate was 71% of the total sample invited to tion revealed no significant differences between the two
participate, with Caucasian (67%) and Asian (26%) the despite participants in the ALB condition being observed to
major cultural backgrounds of participants. The intensity of have less off-task behaviour. The overall coefficient for time
accelerometry-measured physical activity sampled during was negative, meaning that scores were lower in the post-
ALBs (three episodes during week 3 – Table 1) revealed test environment, yet not at a statistically significant level
that participants were engaged in MVPA for approximately (p = 0.23). The interaction effect (condition x time) also
60% of the time (mean 6.02  1.53 minutes). The inter- revealed no statistically significant difference between pre-
individual coefficient of variation for the ALBs was lower and post-break values (p = 0.42). Once again, the study
(25.4%) than other potentially active periods during the hypothesis is rejected.
same day: recess (37.9%), lunch (49.4%) and after school
(54.3%), indicating that most children were equally active
during the ALBs. DISCUSSION
Prolonging lesson time without breaks reduces attention in
Sustained attention students, and inattention in the classroom is a serious
The results of the mixed-model analysis are presented in hindrance to academic learning (27). Objective measures of
Tables 2 and 3 (mean RT) (% lapses). Pre- and post-break the effect of physical activity on students’ sustained atten-
values for the sustained attention measures are presented in tion are sparse in the literature, perhaps due to a perception
Table 4. The difference for mean RT (Table 5) was negligi- that these measures cannot be collected from groups of
ble for both the Active (Cohen’s d = 0.01) and Passive students at the same time. However, a psychometric test
(d = 0.06) Lesson Break conditions. The size of the mean was validated for use in a classroom setting (11), making it
difference for % lapses was also negligible for both the possible to address this lack of classroom-based physical
Active (d = 0.09) and Passive (d = 0.06) Lesson Break activity research (2). Using this recently validated measure-
conditions. ment protocol, the current study found that a brief episode
Mixed modelling for the mean RT and % lapses revealed of physical activity outside the classroom had no impact on
that the condition was not significant (p = 0.61 and 0.64 sustained attention upon returning to the classroom among

©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 5
Short activity breaks and boys attention Wilson et al.

Table 4 Mean (SD) values of classroom behaviour outcomes before and after a 10-minute break* from classroom lessons
Sustained attention (n = 58)† Off-task behaviour (n = 58)††

Mean reaction time (ms) Lapses (%) % of time

Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post
ALB ALB PLB PLB ALB ALB PLB PLB ALB ALB PLB PLB
477 479 511 526 23.0 24.7 27.5 28.7 13.6 14.8 13.4 15.6
(285) (200) (258) (279) (19.4) (20.4) (20.7) (20.3) (10.0) (11.6) (11.6) (11.1)

*The breaks are 10 minutes outside the classroom – Active (ALB) or Passive (PLB).
†
measured using the five-minute PVT before and after the 10-minute break.
††
Measured using in-class observation, 30 minutes pre and post a 10-minute break outside the classroom.

Table 5 Mean differences in classroom behaviour outcomes (post-break minus pre- in the active condition. In these two previous studies, the
break) “Energizers” (11) and “Texas I-CAN” (13) programmes
Off-Task
were linked to the curriculum and conducted inside. The
behaviour ALBs in the current study were conducted outside the
Sustained attention (n = 58)* (n = 58)† classroom. The intensity of activities outside the classroom
in this study may have been different, given that activities
Mean reaction were not restricted by space and students were more freely
time (ms) Lapses (%) % of time able to run and sprint. The intensity of the physical activity
ALB PLB ALB PLB ALB PLB was not measured by either Mahar et al. (11) or Grieco
et al. (13) and there is currently no consensus on the
Mean difference 1.63 15.09 1.78 1.23 1.14 2.17 optimal intensity of activity for cognitive benefits in
Effect size (d) 0.01 0.06 0.09 0.06 0.11 0.20 children. It has been suggested that the intensity ‘threshold’
*Measured using the five-minute PVT before and after the 10-minute break. for cognitive or academic gains is in the vigorous range by
†
Measured using in-class observation, 30 minutes pre and post a 10-minute some researchers (18). Others propose a much wider
break outside the classroom. intensity range that included moderate activity (28). Fur-
ALB = Active lesson break (10 minutes of physical activity outside thermore, Grieco and colleagues (13) identified an interac-
classroom); PLB = Passive lesson break (10 minutes reading a book outside tion of response with weight status, such that active breaks
classroom). resulted in negligible changes in on-task behaviour among
normal-weight children (d = 0.13) and a larger effect
among overweight children (d = 0.26), who perceive the
elementary school-age boys. This observation may encour- intensity of an absolute exercise challenge to be higher
age teachers interested in improving their students’ daily compared with normal-weight children. The lack of con-
physical activity levels to consider implementation of short sensus on this issue highlights the need for more research
physical activity breaks by alleviating concerns about on the intensity of the physical activity that strikes a balance
reduced attention due to the disruption to classroom between cognitive outcomes and wide acceptability among
routine. To appeal to education policymakers, the current children.
research needs to be extended to a wider range of school The duration of an ALB-focused intervention for benefits
settings and to include longer term outcomes such as to be observed is yet to be firmly established. A four-week
academic performance. period was chosen for the current study after considering the
The current study found no evidence that short breaks protocols adopted in previous studies. Mahar et al. (11)
from academic lessons for either physical activity or reading utilised both eight- and four-week active intervention blocks.
quietly outside altered off-task behaviour immediately after The study by Hill et al. (14) of participant attention reported
participants returned to the classroom. There was a non- the changes during one week of activity, while the study by
significant tendency for participants to spend more time Grieco et al. (13) of on-task behaviour utilised only one
‘off-task’ after a passive compared with an active break active day from each participant in their analysis. Both of
(Cohen’s d = 0.20 vs. 0.11). Mahar and colleagues (11) these studies with brief exposures reported improvements in
reported a more substantial increase (d = 0.60) in on-task attention (13) and on-task behaviour (14) with only one to
behaviour (or a decrease in ‘off-task’ behaviour) during the five days of recording. More research is needed to determine
post-active break period among younger grades 3 and 4 the time course of changes over the longer term, and the
children. Grieco et al. (13) also reported statistically signif- extent to which ongoing time commitments and potential
icant differences in time on task between a treatment and ‘fatigue’ among students and teachers impact long-term
control group, although this was attributable to a decrease implementation of the ALB strategy.
in the on-task behaviour of participants in the control group The sample in this study (mean age 11.5 years) was older
(normal classroom routine) rather than an increase of those than in previous studies (9.1 years, [11]; 8.7 years [13]). A

6 ©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
Wilson et al. Short activity breaks and boys attention

brief physical activity break may have a larger effect on of physical activity is optimised through shorter epoch
younger participants due to developmental factors. Work- lengths (30). The limited storage data capacity of the
ing memory, inhibition control (29) and sustained attention AM7164 model accelerometer meant 15 seconds was the
(27) are all factors that develop further during childhood. shortest epoch available, and so is a possible limitation of
Sustained attention, for example, develops rapidly through- this study.
out childhood, before beginning to plateau after the age of
10 and into adolescence (27). It could therefore be possible
that the younger participants in the previous studies (11,13) CONCLUSION
were less able to sustain attention in class for extended The findings of this study have added to the limited
periods of time, so more benefit (better on-task behaviour) literature by introducing an objective measure of sustained
was gained from a break in routine. The findings of a recent attention, and through investigating the feasibility of short
meta-analysis by Erwin et al. (2) suggested that enhanced physical activity bouts outside the classroom, instead of
learning outcomes in physical activity interventions were integrated into the curriculum. This study found no
more often observed among elementary students than evidence that either off-task behaviour or sustained atten-
secondary students, adding support to this hypothesis. tion was changed in this sample of Australian boys as a
The ALB contributed approximately six minutes to daily result of them spending 10 minutes away from their desk to
MVPA in this sample, which is 10% of the recommended undertake physical activity. The results suggest that active
daily minimum MVPA levels for children and adolescents classroom breaks make a substantial contribution to daily
(2). Furthermore, the coefficient of variation for MVPA in physical activity for a high proportion of participants,
the ALB was narrower than observed for free play periods without the need for extensive teacher training and exper-
(recess and lunch breaks) during the same day, suggesting tise, and without disruption to classroom behaviour in the
that the ALB engages a higher proportion of participants in immediate aftermath.
activity than occurs when young people can freely choose
how they use their time. In turn, this suggests that regularly
scheduled active breaks in class time are potentially CONFLICTS OF INTEREST
important sources of MVPA for otherwise sedentary The authors had no financial help in conducting this study.
children. In addition, the authors have no financial or other interest
There were several limitations of the study that should be in Actigraph (formerly Computer Science Application),
acknowledged. Firstly, this study needs to be replicated in a maker of the accelerometer (model 7164) used in this study.
wider range of schools and age ranges. As boys are more
prone to inattention during extended academic study References
periods, the current findings may not generalise to girls.
Direct observation of students in classrooms may result 1. Booth M, Okely A. Promoting physical activity among children
and adolescents: the strength and limitations of school based
in some reactivity, where participants may modify their
approaches. Health Promot J Austr 2005; 16: 52–4.
behaviour as a consequence of being observed. The obser- 2. Erwin H, Fedewa A, Beighle A, Ahn S. A quantitative review of
vation protocol was designed to minimise reactivity by physical activity, health, and learning outcomes associated with
preventing participants and teachers from knowing who classroom-based physical activity interventions. J Appl Sch
was being observed at any given time. This was achieved by Psychol 2012; 28: 14–36.
positioning observers out of the direct line of sight of 3. Howie EK, Pate RR. Physical activity and academic
participants, using observers who were independent of the achievement in children: a historical perspective. J Sport Health
Sci 2012; 1: 160–9.
school with no previous history of interaction with the par-
4. Biddle SJH, Asare M. Physical activity and mental health in
ticipants and placing observers in classrooms in the week children and adolescents: a review of reviews. Br J Sports Med
prior to the study as a familiarisation period. For logistical 2011; 45: 886–95.
reasons, observers were not blinded to the intervention 5. Rasberry CN, Lee SM, Robin L, Laris BA, Russell LA, Coyle
condition. To blind the observers would have required them KK, et al. The association between school-based physical
exiting and re-entering the class during teaching periods to activity, including physical education, and academic
eliminate visible cues such as elevated breathing rate and performance: a systematic review of the literature. Prev Med
2011; 52: s10–20.
sweating associated with physical activity. This would have
6. Sibley BA, Etnier JL. The relationship between physical activity
led to unacceptable disruption to classroom functioning. and cognition in children: a meta-analysis. Pediatr Exerc Sci
However, the non-blinding of observers raised the possi- 2003; 15: 243–56.
bility of observer bias. To address this concern, secondary 7. Singh A, Uijtdewilligen L, Twisk JWR, van Mechelen W,
observers were used for 25% all observation periods and Chipinaw MJM. Physical activity and performance at school. A
inter-observer reliability was high (94%) across the entire systematic review of the literature including a methodological
quality assessment. Arch Pediatr Adolesc Med 2012; 166: 49–
study, similar to the level of inter-observer reliability
55.
reported in previous similar studies (11,13). 8. Dollman J, Boshoff K, Dodd G. The relationship between
Boys’ physical activity bouts are often brief (30), and curriculum time for physical education and literacy and
consensus (especially in younger children) appears to numeracy standards in South Australian primary schools. Eur
support the idea that capturing these frequent, brief bouts Phy Educ Rev 2006; 12: 151–63.

©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 7
Short activity breaks and boys attention Wilson et al.

9. Ahmed Y, MacDonald H, Reed K, Naylor P-J, Liu-Ambrose T, achievement in children. Med Sci Sports Exerc 2006; 38:
McKay H. School based physical activity does not compromise 1515–9.
children’s academic performance. Med Sci Sports Exerc 2007; 19. Pellegrini AD, Huberty PD, Jones I. The effects of recess timing
39: 371–6. on children’s playground and classroom behaviours. Am Educ
10. Stewart JA, Dennison DA, Kohl HW, Doyle A. Exercise level Res J 1995; 32: 845–64.
and energy expenditure in the TAKE 10!â in-class physical 20. Pate RR, Almeida MT, McIver KL, Pfeiffer KA, Dowda M.
activity program. J Sch Health 2004; 74: 397–400. Validation and calibration of an accelerometer in preschool
11. Mahar MT, Murphy SK, Rowe DA, Golden J, Shields AT, children. Obesity (Silver Spring) 2006; 14: 2000–6.
Raedeke TD. Effects of a classroom-based program on physical 21. Roach GD, Dawson D, Lamond N. Can a shorter psychomotor
activity and on-task behaviour. Med Sci Sports Exerc 2006; 38: vigilance task be used as a reasonable substitute for the ten
2086–94. minute psychomotor vigilance task? Chronobiol Int 2006; 23:
12. Donnelly JE, Greene JL, Gibson CA, Smith BK, Washburn RA, 1379–87.
Sullivan DK, et al. Physical Activity Across the Curriculum 22. Tyagi R, Shen K, Shao S, Li X. A novel auditory working
(PAAC): a randomized controlled trial to promote physical memory vigilance task for mental fatigue assessment. Saf Sci
activity and diminish overweight and obesity in elementary 2009; 47: 967–72.
school children. Prev Med 2009; 49: 336–41. 23. Sforza E, Haba-Rubio J, De Bilbao F, Rochat T, Ibanez V.
13. Grieco LA, Jowers EM, Bartholomew JB. Physically active Performance vigilance task and sleepiness in patients
academic lessons and time on task: the moderating effect of with sleep-disordered breathing. Eur Respir J 2004; 24:
body mass index. Med Sci Sports Exerc 2009; 41: 1921–6. 279–85.
14. Hill L, Williams JHG, Aucott L, Milne J, Thomson J, Greig J, 24. Thorne DR, Johnson DE, Redmond DP, Sing HC, Belenky G,
et al. Exercising attention within the classroom. Dev Med Child Shapiro JM. The Walter Reed palm-held psychomotor vigilance
Neurol 2010; 52: 929–34. test. Behav Res Methods 2005; 37: 111–8.
15. Riley N, Lubans DR, Holmes K, Morgan PJ. Findings from the 25. Wilson A, Dollman J, Lushington K, Olds T. Reliability of the 5-
EASY minds cluster randomized controlled trial: evaluation of min psychomotor vigilance task in a primary school classroom
a physical activity Integration program for mathematics in setting. Behav Res Methods 2010; 42: 754–8.
primary schools. J Phys Act Health 2015. doi:10.1123/ 26. Freedson P, Pober D, Janz KF. Calibration of accelerometer
jpah.2015-0046 output for children. Med Sci Sports Exerc 2005; 37: s523–30.
16. Godwin KE, Fisher AV. Allocation of attention in classroom 27. Betts J, McKay J, Maruff P, Anderson V. The development of
environments: consequences for learning. In: Carlson L, sustained attention in children: the effect of age and task load.
Holscher C, & Shipley TF (Eds.), Proceedings of the 33rd Child Neuropsychol 2006; 12: 205–21.
Annual Conference of the Cognitive Science Society (2806-12). 28. Brisswalter J, Collardeau M, Rene A. Effects of acute physical
Austin, TX: Cognitive Science Society, 2011. exercise characteristics on cognitive performance. Sports Med
17. Salmeron L, Garcia V. Children’s Reading of printed text and 2002; 32: 555–66.
hypertext with navigation overviews: the role of 29. Diamond A. Executive Functions. Annu Rev Psychol 2013; 64:
comprehension, sustained attention, and visuo-spatial abilities. 135–68.
J Educ Comput Res 2012; 47: 33–50. 30. Sanders T, Cliff D, Lonsdale C. Measuring adolescent boys’
18. Coe DP, Pivarnik JM, Womack CJ, Reeves MJ, Malina RM. physical activity: bout length and the influence of
Effects of physical education and activity levels on academic accelerometer epoch length. PLoS One 2014; 9: 1–8.

8 ©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd

View publication stats