Vol.5, No.2, 245-252 (2013) Health
http://dx.doi.org/10.4236/health.2013.52033
Direct and indirect effects of multilevel factors on
school-based physical activity among Japanese
adolescent boys
Li He1*, Kaori Ishii2, Ai Shibata2, Minoru Adachi3, Keiko Nonoue4,5, Koichiro Oka2
1Graduate School of Sport Sciences, Waseda University, Tokorozawa, Japan; *Corresponding Author: meethl@moegi.waseda.jp
2Faculty of Sport Sciences, Waseda University, Tokorozawa, Japan
3Graduate School of Education, Okayama University, Okayama, Japan
4Sonan Junior High School, Okayama, Japan
5Waseda Institute for Sport Sciences, Tokorozawa, Japan
Received 26 October 2012; revised 25 November 2012; accepted 4 December 2012
ABSTRACT
Purpose: Background: Physical activity is a
complex behavior which involves the intera ctio n
of multilevel factors at the individual, social and
environmental level. However, previous studies
have largely focused on psychological and/or
social environmental factors and the direct im-
pact of such factors on physical activity. There
are few studies having examined how multilevel
factors may interact to influence activity level.
Therefore, the purpose of the present study was
to examine both direct and indirect effects of
multilevel factors on school-based physical ac-
tivity in Japanese adolescent boys. Methods: In
this cross-sectional surv ey of the Japanese ado-
lescent lifestyles, 379 junior high school boys
were invited to complete self-r epor t measur es of
age, grade, weight, height, self-efficacy, social
support (family, friends and teachers), school
physical environment (equipment, facilities and
safety) and average minutes per week of phy-
sical activity during lunch time and after-school
hours occurring at school. Structural equation
modeling analyses controlling for age were uti-
lized to examine the effects of body mass index
(BMI), self-efficacy, social support and school
physical environmental variables on lunchtime
and after-school physical activity. Results: Dur-
ing lunch time, self-efficacy exhibited direct
positive effects on physical activity. BMI, facili-
ties, and safety were indirectly associated with
lunchtime physi cal a ctivity through s elf -efficacy.
However, there were no significant relationships
of equipment and social support with lunchtime
physical activity. During after-school hours, fa-
mily support and facilities directly affected phy-
sical activity. Self-efficacy was indirectly related
with physical activity through family support.
BMI, equipment, and safety indirectly affected
physical activity through self-efficacy and/or
family support. Conclusion: Effects of multilevel
factor on physical activity among adolescent
boys differed according to context, which im-
plies that interventions to promote physical ac-
tivity should be context-specific. Findings en-
courage the development of future effective in-
terventions to promote physical activity through
self-efficacy during lunch time as well as family
support during after-school hours.
Keywords: Public Health; Physical Activity;
Adolescents; Ecological Model; Structur al Equation
Modeling
1. INTRODUCTION
Regular physical activity during adolescence is known
to have both short- and long-term health benefits [1,2].
However, participation in such activity declines dramati-
cally during this period of development [3]. The majority
of adolescents do not meet the recommended 60 min/day
of physical activity [3,4]. Therefore, promoting their
physical activity through successful interventions has
been a high public health priority.
Identifying modifiable correlates of physical activity
behavior that could act as mediators is an important step
in developing successful interventions. Mediators are the
third variables in a causal path that explain how or why
effects of initial variables on outcome variables occur [5].
However, the mediated (i.e., indirect) effects of corre-
lates on physical activity remain poorly understood in
regard to adolescents. Although the social ecological
model suggests that physical activity behavior is ex-
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252
246
plained by the interaction of personal, social, and envi-
ronmental factors [6], much research is limited to indi-
viduals’ cognitions and perceptions of social environ-
mental influences [7-12]. Few researchers have exam-
ined both direct and indirect effects of multilevel factors
on physical activity across personal, social, and envi-
ronmental levels.
In this respect, Motl et al. reported that the relations
between perceived equipment accessibility at home and
in the neighborhood, social support, and self-reported
physical activity were accounted for by self-efficacy in
adolescent girls [13,14]. In a study by Lubans et al. [15],
while social support was not significantly related to
physical activity in adolescent girls, school physical en-
vironment was found to partially mediate the relation
between self-efficacy and objective physical activity.
However, those studies focused on overall physical ac-
tivity level without distinguishing physical activity in
different contexts, for example at school or out of school
during after-school hours. As the ecological approach
notes, behavior occurs in various “settings or contexts”
(i.e., in certain places and at certain times). Thus, exam-
ined behaviors should be specific to the relevant envi-
ronment (i.e., neighborhood physical environment rele-
vant to out-of-school activity; school physical environ-
ment relevant to school-based physical activity) [16].
Moreover, in those studies, models were tailored for
adolescent girls and did not reflect boys’ specific charac-
teristics. Physical activity declines in boys as well as
girls during adolescence, although boys are typically
more active than girls [17]. Considering that boys and
girls are no longer physically well-matched from early
adolescence [18], personal, environmental, and beha-
vioral characteristics might be unique to different gen-
ders. Therefore, interventions targeting adolescent boys
as well as girls should be guided by relevant theoretical
models. Furthermore, such models provide evidence for
targeting self-efficacy or physical environment as possi-
ble mediators [13-15]. However, whether social support
can account for the associations of self-efficacy and
physical environment with physical activity is still un-
known.
School is an important setting for promoting adoles-
cents’ physical activity at the population level. Students
spend much of the day during the week at school and
have many potential opportunities for daily physical ac-
tivity there in addition to physical education class (e.g.,
break times, including lunch recess and after-school
hours). A previous review reported that involvement in
physical activity during such non-curricular school time
contributes 5% - 40% of the recommended 60 minutes of
daily physical activity [19]. However, an understanding
of the effects of school environment on such physical
activity is limited. Moreover, there are no studies that
have comprehensively examined the effects of personal,
social, and school physical environmental factors that
impact such physical activity among adolescent boys.
Consequently, the present study aimed to explore the
direct and indirect effects of personal, social, and school
physical environmental factors on non-curricular school-
based physical activity among adolescent boys. The pri-
mary hypothesis was that perceived social support would
be a possible mediator of physical activity. The present
model that tested effects of multilevel correlates on con-
text-specific physical activity among adolescent boys can
contribute to the systematic progression of physical ac-
tivity research as well as the development of effective
school-based interventions for boys.
2. METHODS
2.1. Participants and Data Collection
The present study included data on 12 - 15-year-old
adolescent boys who participated in a cross-sectional
survey of Japanese junior high school students’ lifestyle.
The survey aimed to assess the interactions between
personal and environmental factors and adolescents’ life-
style. A total of 761 students participated, including 379
boys. All participants were asked to complete a question-
naire individually during class time. Demographic in-
formation such as age, gender, and grade were collected
with this questionnaire. Informed consent was obtained
from all participants, guardians, and the school. Partici-
pation was voluntary, and confidentiality of the partici-
pants was ensured. The study protocol was approved by
the research ethics committee of Waseda University.
Data collection was conducted between Oct. and Dec.
2010.
2.2. Measures
Measures of weight (kg) and height (cm) were col-
lected through a height and weight measuring scale.
Body mass index (BMI) was calculated. Participants
were grouped into underweight, normal weight, over-
weight, and obese by BMI ranges specific for age and
gender using Centers for Disease Control and Prevention
criteria [20].
The frequency (days per week) and duration (minutes
per day) of physical activity at school during lunch re-
cess and after-school hours in a usual week were re-
ported by adolescents. For analysis, minutes per week of
physical activity in two settings were calculated [21].
School environmental characteristics were assessed
subjectively, using 10 items. These items comprised 3
factors: 1) “Equipment” (3 items), examining the accessi-
bility or usability of physical equipment (e.g., There is
enough equipment for physical activity at school); 2)
“Facility” (4 items), measuring the accessibility or usabil-
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252 247
ity of physical activity facilities (e.g., The school ground
is wide enough for physical activity); and 3) “Safety” (3
items) investigating perceived safety of physical activity
equipment and facilities (e.g., It is safe to engage in phy-
sical activity on the grounds and in the gym at school).
The factorial reliability (Equipment: Cronbach α = 0.72;
Facility: Cronbach α = 0.77; and Safety: Cronbach α =
0.78) of this scale was confirmed by respondents. All
items were measured using a four-point scale from 1)
strongly disagree to 4) strongly agree.
To measure social support for physical activity, stu-
dents were asked to rate support from three sources on a
four-point scale from 1) no support at all to 4) strongly
supported, using the following question: “How do you
rate support for engaging in physical activity from 1)
family, 2) teachers, or 3) friends?”
The measure of self-efficacy (i.e., belief in one’s abi-
lity to be active relative to peers) in the present study
contained 1 item with response choices ranging from 1)
strongly disagree to 4) strongly agree [22]. The statement
was “I am able to do physical activities/exercises/sports
better than my friends”.
2.3. Data Anal ysis
A list-wise deletion procedure was adopted. Means
and standard deviations (S.D.) of variables were then
calculated through descriptive statistics using SPSS 18.0
software. The size of the final sample was adequate to
estimate the models [23]. Finally, structural equation
modeling (S.E.M.) analysis with maximum likelihood
estimation using Amos 17.0 was performed to test the fit
of proposed models.
The original model leading to a good fit of the final
model is described below. The measurement model in-
cluded 1) three latent variables of physical environment:
equipment (3 indicators), facilities (4 indicators), and
safety (3 indicators); 2) relations between latent vari-
ables and their indicators; and 3) correlations between
three latent environmental factors. On the basis of the
hypothesis, the structural model included 1) paths from
perceived physical equipment, facilities, and safety, and
BMI to perceived self-efficacy and self-reported physical
activity, 2) path from self-efficacy to each source of so-
cial support; and 3) paths from self-efficacy and three
sources of social support to physical activity.
Model fit was assessed using the goodness-of-fit index
(GFI), adjusted goodness-of-fit index (AGFI), root mean
square error of approximation (RMSEA), and Akaike in-
formation criterion (AIC). GFI and AGFI are used to
measure how well the model fits the data and vary from
0 to 1, with 0.90 indicating an acceptable model fit and
0.95 indicating a good model fit [24,25]. RMSEA is a
measure of the discrepancy between a population-based
model and a hypothesized model assessed per degree of
freedom. There is good model fit if the RMSEA is less
than or equal to 0.05 with the upper limit of confidence
interval less than 0.08 and the lower 90% confidence
limit including or close to 0 [25]. A lower AIC value re-
flects a better-fitting model compared with competing
models [26]. A model was considered to fit the data when
the following criteria were met: GFI > 0.90, AGFI > 0.90
(AGFI < GFI), RMSEA < 0.05, and lower AIC value
compared with competing models. A p value less than
0.05 was considered statistically significant.
To adjust the original specified model, new free pa-
rameters were added based on the modified indices be-
fore the Wald test to delete all non-significant free pa-
rameters thereby increasing model fitness. Then, only
significant causal paths with corresponding standardized
regression coefficients (β) were shown in the figures of
final structural models that demonstrated a good model
fit. With the standardized regression coefficients, the
magnitude of each factor could be directly compared
with other factors in the model.
3. RESULTS
3.1. Participant Characteristics
Of the 379 boys who returned the questionnaire, 300
with complete data (mean age = 13.5, S.D. = 0.96) com-
prised the final sample. Mean height and weight were
161.85 cm (S.D. = 8.06) and 50.30 kg (S.D. = 11.77),
respectively. The majority of adolescents had normal
weight (5 BMI < 85 percentile, n = 236, 78.7%). More
information about the characteristics of studied variables
is provided in Table 1.
Table 1. Characteristics of participants and physical activity
outcome variable.
Participants (Na = 300)
Mean S.D.a
Age (year) 13.49 0.99
Height (cm) 161.85 8.06
Weight (kg) 50.30 11.77
BMI 19.09 3.83
Lunch-time physical activity 16.80 35.88
After-school physical activity282.98 319.03
Grade (N, %)
Grade 1 98 32.7%
Grade 2 98 32.7%
Grade 3 104 34.7%
Weight status (N, %)
Underweight 27 9.0%
Normal weight 236 78.7%
Overweight 21 7.0%
Obesity 16 5.3%
aN: Number; S.D.: Standard deviation.
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252
Copyright © 2013 SciRes.
248
3.2. Structural Equation Model
The final structural model for lunchtime physical ac-
tivity in Figure 1 demonstrated a good model fit (GFI =
0.96, AGFI = 0.93, RMSEA = 0.03 [90% confidence
interval = 0.004 - 0.045]). The value of AIC was reduced
from 815.83 to 204.70 after the model modifications.
During lunch recess, self-efficacy (β = 0.13) directly and
positively affected physical activity. The standardized
coefficients for the indirect effect of perceived facilities,
safety, and BMI through self-efficacy was –0.03, 0.03
and –0.02, respectively. Their effect sizes on physical
activity were generally low. Perceived equipment and
social support had neither direct nor indirect effects on
lunchtime physical activity. Self-efficacy was the most
important factor and mediator affecting lunchtime phy-
sical activity.
The final structural model for after-school physical ac-
tivity presented in Figure 2 demonstrated a good model
fit (GFI = 0.95, AGFI = 0.93, RMSEA = 0.03 [90% con-
fidence interval = 0.017 - 0.047]). The recalculation of
the model after addition and deletion of free parameters
reduced the AIC value from 859.15 to 237.16. Family
support (β = 0.28) was identified as the most influential
factor directly affecting physical activity during after-
school hours. Self-efficacy (β = 0.06) and perceived
equipment (β = 0.04) indirectly affected physical activity
through family support. The path coefficient for the in-
direct positive effects of perceived safety on physical
activity through self-efficacy and family support was
0.02. The total effects of facilities (β = –0.14) on physi-
cal activity were partially mediated by self-efficacy and
family support. The path coefficient for the indirect
negative effects of facilities through self-efficacy and
family support on physical activity was –0.02. BMI (β =
–0.01) indirectly affected physical activity through self-
efficacy and family support.
4. DISCUSSION
The present study examined cross-sectional effects of
perceived school physical environment, social support,
self-efficacy, and BMI on school-based physical activity
in Japanese adolescent boys. The findings can support
the development of school-based physical activity inter-
vention programs meeting specific needs of adolescent
boys regardless of age. The primary finding of this study
was that the effects of variables on physical activity de-
pended on the context, which implies that the develop-
ment of effective interventions for promoting physical
activity should be tailored for specific contexts.
The present study also identified positive direct and
indirect associations between self-efficacy and physical
activity, although the domain of physical activity exam-
ined was different from previous studies [8,10,13-15].
During lunch recess, self-efficacy was identified fully
mediating the effects of facility, safety, and BMI on phy-
sical activity. This finding supports previous studies
showing that environmental factors affect physical activ-
ity through psychological factors [13,14]. Both current
and previous findings indicate that increasing self-effi-
cacy might be a means of directly increasing physical
activity during lunch recess. However, there were no sig-
nificant associations between social support and physical
activity, indicating that social support was not an effec-
tive mediator. Therefore, exploring other third variables
(e.g., perceived barriers, enjoyment, or friends’ physical
activity behavior [9,27,28]) that act as influential factors
Figure 1. Effects of personal, social, and physical environ-mental factors on lunch-time physi-
cal activity among boys. Only statistically significant paths are indicated in the figure. The sig-
nificance level was set at p < 0.05. Digitals in each path represent standardized path coeffi-
cients. BMI: body mass index; Family: family support; Teacher: teacher support; Friend: friend
support.
OPEN ACCESS
L. He et al. / Health 5 (2013) 245-252 249
Figure 2. Effects of personal, social, and physical environmental factors on after-school physi-
cal activity among boys. Only statistically significant paths are indicated in the figure. The sig-
nificance level was set at p < 0.05. Digitals in each path represent standardized path coeffi-
cients. BMI: body mass index; Family: family support; Teacher: teacher support; Friend: friend
support.
mediating the associations between self-efficacy, physi-
cal environment, and lunchtime physical activity are
warranted in the future.
During after-school hours, in line with the hypothesis,
family support was found to fully mediate effects of
equipment, safety, self-efficacy, and BMI on physical
activity. In addition, perceived accessible facilities di-
rectly or indirectly affected physical activity through
self-efficacy and family support. This finding indicates
that increasing perceived family support might serve as a
beneficial strategy for increasing physical activity during
after-school hours. However, evidence against the hy-
pothesis was that support from friends and teachers did
not significantly affect such physical activity. Much at-
tention has been paid to comparing the influences of
family and friends as sources of social support for ado-
lescent physical activity [11,29-33]. However, conflict-
ing results were found in previous studies [11,29,30,
32,33]. There is a lack of studies taking into account
support from teachers when investigating the relative
importance of family, friends, and teachers on adolescent
boys’ school-based physical activity. Consistent with the
study by Hsu et al. [33], the present study identified fam-
ily support as significantly influencing boys’ physical
activity, while the effect of friend support was not sig-
nificant. This finding was explicable based on the sub-
stantial reliance of adolescent boys on instrumental or
informational and emotional support from family. Boys
may follow their friends in joining activities, but family
support (e.g., cost of equipment, approval, encourage-
ment or praise for behaviors, or talking about activities
frequently) might be more important in removing barri-
ers to being active, especially in early adolescence [10,
29,33,34]. The latest national report regarding family
effects on junior high students’ exercise habits showed
that across Japan, 40.8% of junior high boys talked about
physical activities with their families at least once weekly
[35]. It is understandable that teacher support was not
significantly important because after-school hours repre-
sent free time in which physical activity becomes a lei-
sure choice. The influence of teachers on adolescents
might be more significant in physical education courses
than during free-time choices. Based on the present
findings, designing more effective interventions through
family support should involve more in-depth research
examining types of family support (e.g., encouragement
or tangible assistance) in association with context-spe-
cific physical activity. Moreover, it is necessary to un-
derstand the interaction effects of family, friends, and
teachers (e.g., the combined effect of any two sources
and the full interaction of all three sources) on school-
based physical activity in the future.
Previous research has revealed significant direct and
positive influences of some school environmental char-
acteristics (e.g., availability of play equipment and facili-
ties like playing fields) on overall physical activity or
recess physical activity at school [36,37]. However, un-
derstanding regarding how various components of the
school environment impact school-based physical acti-
vity is limited. In this respect, the present study revealed
that perceived equipment had neither direct nor indirect
effects on lunchtime physical activity, but had indirect
effects on after-school physical activity through family
support. This difference might be explained by the fact
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252
250
that boys are often involved in different types of physical
activity in different time periods during weekdays [38].
In Japan, recess after lunch for students is only about 20
minutes, whereas time at school at the end of day is
about 3 hours. Boys might tend to engage in physical
activities that do not require equipment (e.g., playing or
walking between school buildings/classrooms) during the
short lunch recess. Accordingly, further studies should
take types of activities into account to better understand
how the physical environment affects preferred types of
activities during different time periods.
Against expectation and previous studies [15,39], per-
ceived facilities were found to be inversely associated
with boys’ physical activity in the present study. It is
possible that boys who actually used facilities compared
to boys who did not reported lower scores for available
facilities because the latter were less aware of the real
situation regarding the facilities. Further studies are
needed to test whether perceptions of facilities match ob-
jective measurements to clarify the influence of facilities
on physical activity. In addition, the present study indi-
cates that boys who perceived high safety were more
active than those who reported low perceptions of safety,
regardless of setting. This finding suggests that increas-
ing perceptions of safety as well as equipment is a possi-
ble way to promote self-efficacy and social support and
perhaps ultimately increase physical activity among boys.
This highlights the necessity of improving both the ob-
jective environment to be more activity-friendly at school
and perceptions among students of school resources for
physical activity. There are various strategies that might
be used to increase the awareness of equipment at school
and its safety. For example, school officials might update
news of regular maintenance of facilities and equipment
in a monthly report distributed to students and parents, or
put a sign on school grounds informing students of the
available equipment or facilities. Regardless, more in-
depth research is necessary in regard to manipulating
perceptions of the physical environment to observe
changes in self-efficacy and perceptions of social support
and physical activity.
Finally, the present study indicates that boys with
higher BMI were less active than those with lower BMI
because they perceived less self-efficacy and family
support. Thus, it is important in the future to examine the
interventional effects of self-efficacy and family support
on physical activity in overweight or obese boys. Con-
sidering that factors influencing physical activity might
be different in those overweight, obese, or of normal
weight, more research is needed to explore correlates/
determinants of physical activity in overweight and
obese adolescents. This could facilitate the development
of effective strategies for promoting physical activity
among this specific at-risk group.
4.1. Strengths
The first strength of this study was extending previous
research by simultaneously measuring direct and indirect
effects of multilevel contributing factors on context-spe-
cific physical activity rather than overall physical activity
level. Several sources of social support and various
school physical environmental attributes were examined
concurrently. Second, this study contributed to studies
about adolescents by exploring a behavioral model tai-
lored for adolescent boys’ physical activity behavior.
Finally, this study used SEM, which was helpful in ex-
ploring potential mediators that can be intervened upon,
and allowed the examination of relative contributions of
factors that explain physical activity behavior.
4.2. Limitations
One limitation of this study is the use of a self-report
measure of physical activity, which is subject to error
and bias. Further studies should attempt to combine ex-
isting objective and subjective measures to investigate
context-specific physical activity more accurately. An-
other limitation is the list-wise deletion adopted in the
present study that may have biased the data findings. A
further limitation is that generalizability of findings be-
yond the study location may be limited because data
were collected from a single school. However, the pre-
valence of overweight and obesity among Japanese jun-
ior high students in a national survey (8.4% [40]) was
only slightly lower than in the present study (12.3%).
Therefore, it is likely that the structural models in this
study would fit counterparts across the country. Still an-
other limitation was the cross-sectional data that permit-
ted only estimates of between-person relations among
variables. Therefore, longitudinal or interventional de-
signs are warranted in the future.
4.3. Conclusion
The present study implies that improvement of self-
efficacy and family support can be effective in directly
increasing school-based physical activity in adolescent
boys. Furthermore, these findings can encourage re-
searchers as well as policy makers, both at school and the
national level, to consider resources and improvements
in school physical environments as a means of increasing
perceptions of social support and self-efficacy when de-
veloping programs and strategies that are important to
increasing physical activity indirectly.
5. ACKNOWLEDGEMENTS
This investigation was supported by the Grants-in-Aid for Scientific
Research (No. 22700680) from the Japan Society for the Promotion of
Science, Waseda University Grant for Special Research Projects
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252 251
(2010A-095, 2011A-092), and the Global COE Program “Sport Sci-
ences for the Promotion of Active Life” from the Japan Ministry of
Education, Culture, Sports, Science, and Technology.
REFERENCES
[1] Hallal, P.C., Victora, C.G., Azevedo, M.R. and Wells, J.C.
(2006) Adolescent physical activity and health: A system-
atic review. Sports Medicine, 36, 1019-1030.
doi:10.2165/00007256-200636120-00003
[2] Janssen, I. and LeBlanc, A.G. (2010) Systematic review
of the health benefits of physical activity and fitness in
school-aged children and youth. International Journal of
Behavioral Nutrition and Physical Activity, 7, 40.
doi:10.1186/1479-5868-7-40
[3] Biddle, S.J.H. and Mutrie, N. (2008) Psychology of phy-
sical activity: Determinants, well-being and interventions.
2nd Edition, Routledge, London and New York.
[4] World Health Organization (2012) Global recommenda-
tion on physical activity for health.
http://whqlibdoc.who.int/publications/2010/97892415999
79_eng.pdf
[5] Baron, R. and Kenny, D. (1986) The moderator-mediator
variable distinction in social psychological research: Con-
ceptual, strategic, and statistical considerations. Journal
of Personality and Social Psychology, 51, 1173-1182.
doi:10.1037/0022-3514.51.6.1173
[6] Sallis, J.F. and Owen, N. (2002) Ecological model of
health behavior. In: Glanz, K., Rimer, B.K. and Lewis,
F.M., Eds, Health Behavior and Health Education, 3rd
Edition, Jossey-Bass, San Francisco, 462-464.
[7] Dishman, R.K., Motl, R.W., Saunders, R., Felton, G.,
Ward, D.S., Dowda, M., et al. (2004) Self-efficacy par-
tially mediates the effect of a school-based physical ac-
tivity intervention among adolescent girls. Preventive Me-
dicine, 38, 628-636. doi:10.1016/j.ypmed.2003.12.007
[8] Dishman, R.K., Saunders, R.P., Motl, R.W., Dowda, M.
and Pate, R.R. (2009) Self-efficacy moderates the relation
between declines in physical activity and perceived social
support in high school girls. Journal of Pediatric Psy-
chology, 34, 441-451. doi:10.1093/jpepsy/jsn100
[9] Dishman, R.K., Dunn, A.L., Sallis, J.F., Vandenberg, R.J.
and Pratt, C.A. (2010) Social-cognitive correlates of phy-
sical activity in a multi-ethnic cohort of middle-school girls:
Two-year prospective study. Journal of Pediatric Psy-
chology, 35, 188-198. doi:10.1093/jpepsy/jsp042
[10] Trost, S.G., Sallis, J.F., Pate, R.R., Freedson, P.S., Taylor,
W.C. and Dowda, M. (2003) Evaluating a model of pa-
rental influence on youth physical activity. American
Journal of Preventive Medicine, 25, 277-282.
doi:10.1016/S0749-3797(03)00217-4
[11] Wu, T., Pender, N. and Noureddine, S. (2003) Gender
differences in the psychosocial and cognitive correlates of
physical activity among Taiwanese adolescents: A struc-
tural equation modeling approach. International Journal
of Behavioral Medicine, 10, 93-105.
doi:10.1207/S15327558IJBM1002_01
[12] Motl, R.W., Dishman, R.K., Ward, D.S., Saunders, R.P.,
Dowda, M., Felton, G., et al. (2005) Comparison of bar-
riers self-efficacy and perceived behavioral control for
explaining physical activity across 1 year among adoles-
cent girls. Health Psychology, 24, 106-111.
doi:10.1037/0278-6133.24.1.106
[13] Motl, R.W., Dishman, R.K., Ward, D.S., Saunders, R.P.,
Dowda, M., Felton, G., et al. (2005) Perceived physical
environment and physical activity across one year among
adolescent girls: Self-efficacy as a possible mediator?
Journal of Adolescent Health, 37, 403-408.
doi:10.1016/j.jadohealth.2004.10.004
[14] Motl, R.W., Dishman, R.K., Saunders, R.P., Dowda, M.
and Pate, R.R. (2007) Perceptions of physical and social
environment variables and self-efficacy as correlates of
self-reported physical activity among adolescent girls.
Journal of Pediatric Psychology, 32, 6-12.
doi:10.1093/jpepsy/jsl001
[15] Lubans, D.R., Okely, A.D., Morgan, P.J., Cotton, W., Pug-
lisi, L. and Miller, J. (2012) Description and evaluation of
a social cognitive model of physical activity behaviour
tailored for adolescent girls. Health Education Research,
27, 115-128. doi:10.1093/her/cyr039
[16] Rosenberg, D., Ding, D., Sallis, J.F., Kerr, J., Norman,
G.J., Durant, N., et al. (2009) Neighborhood environment
walkability scale for youth (NEWS-Y): Reliability and
relationship with physical activity. Preventive Medicine,
49, 213-218. doi:10.1016/j.ypmed.2009.07.011
[17] Nader, P.R., Bradley, R.H., Houts, R.M., McRitchie, S.L.
and O’Brien, M. (2008) Moderate-to-vigorous physical
activity from ages 9 to 15 years. JAMA, 300, 295-305.
doi:10.1001/jama.300.3.295
[18] Berk, L.E. (2005) Infants, children and adolescents. 5th
Edition, Allyn & Bacon Inc., Bosten.
[19] Ridgers, N., Stratton, G. and Fairclough, S. (2006) Physi-
cal activity levels of children during school playtime.
Sports Medicine, 36, 359-371.
doi:10.2165/00007256-200636040-00005
[20] Centers for Disease Control and Prevention (2011) About
BMI for children and teens.
http://www.cdc.gov/healthyweight/assessing/bmi/children
s_bmi/about_childrens_bmi.html
[21] He, L., Ishii, K., Shibata, A., Adachi, M., Nonoue, K. and
Oka, K. (2012) Patterns of physical activity outside of
school time among Japanese junior high school students.
Journal of School Health, in Press.
[22] Ryan, G. and Dzewaltowski, D. (2002) Comparing the
relationships between different types of self-efficacy and
physical activity in youth. Health Education & Behavior,
29, 491-504. doi:10.1177/109019810202900408
[23] North Carolina State University (2012) Structural equa-
tion modeling: Statnotes.
http://faculty.chass.ncsu.edu/garson/PA765/structur.htm
[24] Kline, R.B. (1998) Principle and practice of structural
equation modeling. Guilford, New York.
[25] Schumacker, R.E. and Lomax, R.G. (2004) A beginner’s
guide to structural equation modeling. 2nd Edition, Law-
rence Erlbaum Associates, Mahwah.
[26] Hamparsum, B. (1987) Model selection and Akaike’s in-
Copyright © 2013 SciRes. OPEN A CCESS
L. He et al. / Health 5 (2013) 245-252
Copyright © 2013 SciRes. OPEN A CCESS
252
formation criterion (AIC): The general theory and its ana-
lytical extensions. Psychometrika, 52, 345-370.
doi:10.1007/BF02294361
[27] Dishman, R.K., Motl, R.W., Saunders, R, Felton, G.,
Ward, D.S., Dowda, M., et al. (2005) Enjoyment medi-
ates effects of a school-based physical activity interven-
tion. Medicine and Science in Sports and Exercise, 37,
478-487. doi:10.1249/01.MSS.0000155391.62733.A7
[28] Jago, R., Page, A.S. and Cooper, A.R. (2012) Friends and
physical activity during the transition from primary to se-
condary school. Medicine and Science in Sports and Ex-
ercise, 44, 111-117. doi:10.1249/MSS.0b013e318229df6e
[29] Duncan, S., Duncan, T. and Strycker, L. (2005) Sources
and types of social support in youth physical activity.
Health Psychology, 24, 3-10.
doi:10.1037/0278-6133.24.1.3
[30] Hohepa, M., Scragg, R., Schofield, G., Kolt, G.S. and
Schaaf, D. (2007) Social support for youth physical activ-
ity: Importance of siblings, parents, friends and school
support across a segmented school day. International
Journal of Behavioral Nutrition and Physical Activity, 4,
54. doi:10.1186/1479-5868-4-54
[31] Lubans, D.R., Morgan, P.J., Callister, R., Collins, C.E.
and Plotnikoff, R.C. (2010) Exploring the mechanisms of
physical activity and dietary behavior change in the pro-
gram X intervention for adolescents. Journal of Adoles-
cent Health, 47, 83-91.
doi:10.1016/j.jadohealth.2009.12.015
[32] Patnode, C.D., Lytle, L.A., Erickson, D.J., Sirard, J.R.,
Barr-Anderson, D. and Story, M. (2010) The relative in-
fluence of demographic, individual, social, and environ-
mental factors on physical activity among boys and girls.
International Journal of Behavioral Nutrition and Physi-
cal Activity, 7, 79. doi:10.1186/1479-5868-7-79
[33] Hsu, YW, Chou C.P., Nguyen-Rodriguez, S.T., McClain,
A.D., Belcher, B.R. and Spruijt-Metz, D. (2011) Influ-
ences of social support, perceived barriers, and negative
meanings of physical activity on physical activity in mid-
dle school students. Journal of Physical Activity & Health,
8, 210-219.
[34] Ornelas, I.J., Perreira, K.M., and Ayala, G.X. (2007) Pa-
rental influences on adolescent physical activity: A lon-
gitudinal study. International Journal of Behavioral Nu-
trition and Physical Activity, 4, 3.
doi:10.1186/1479-5868-4-3
[35] Ministry of Education, Culture, Sports, Science and Tech-
nology, Japan (2012) 2010’s national survey result of
physical and athletic capacity and exercise habits.
http://www.mext.go.jp/component/a_menu/sports/detail/_
_icsFiles/afieldfile/2010/12/16/1300103_5.pdf
[36] Haug, E., Torsheim, T. and Samdal, O. (2008) Physical
environmental characteristics and individual interests as
correlates of physical activity in Norwegian secondary
schools: The health behaviour in school-aged children
study. International Journal of Behavioral Nutrition and
Physical Activity, 5, 47. doi:10.1186/1479-5868-5-47
[37] Haug, E., Torsheim, T., Sallis, J.F. and Samdal, O. (2010)
The characteristics of the outdoor school environment
associated with physical activity. Health Education Re-
search, 25, 248-256. doi:10.1093/her/cyn050
[38] Stanley, R.M., Ridley, K. and Olds, T.S. (2011) The type
and prevalence of activities performed by Australian
children during the lunchtime and after school periods.
Journal of Science and Medicine in Sport, 14, 227-232.
doi:10.1016/j.jsams.2010.10.461
[39] Durant, N., Harris, S.K., Doyle, S., Person, S., Saelens,
B.E., Kerr, J., et al. (2009) Relation of school environ-
ment and policy to adolescent physical activity. Journal
of School Health, 79, 153-159.
doi:10.1111/j.1746-1561.2008.00384.x
[40] Ministry of Education, Culture, Sports, Science and Tech-
nology, Japan (2012) 2010’s national survey result of
physical and athletic capacity and exercise habits: (4)
physical fitness and BMI.
http://www.mext.go.jp/component/a_menu/sports/detail/_
_icsFiles/afieldfile/2010/12/16/1300270_4_1.pdf