Journal of Adolescent Health 49 (2011) 547–549
www.jahonline.org
Adolescent health brief
A Test of the Theory of Planned Behavior to Explain Physical Activity in a Large
Population Sample of Adolescents From Alberta, Canada
Ronald C. Plotnikoff, Ph.D.a,b,*, David R. Lubans, Ph.D.a, Sarah A. Costigan a, Linda Trinh, M.A.c,
John C. Spence, Ph.D.d, Shauna Downs, M.Sce, and Linda McCargar, Ph.D.e
a
School of Education, Faculty of Education and Arts, University of Newcastle, Callaghan, New South Wales, Australia
School of Public Health, Faculty of Physical Education, University of Alberta, Edmonton, Alberta, Canada
c
Behavioral Medicine Laboratory, University of Alberta, Edmonton, Alberta, Canada
d
Faculty of Physical Education and Recreation, University of Alberta, Edmonton, Alberta, Canada
e
Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Alberta, Canada
b
Article history: Received September 10, 2010; Accepted March 9, 2011
Keywords: Adolescents; Physical activity; Theory of planned behavior; Gender differences; Web-based survey
A B S T R A C T
Purpose: To test the Theory of Planned Behavior (TPB) in a large population sample of adolescents from
Alberta, Canada.
Methods: 4,073 adolescents completed a self-administered web-based survey related to physical activity
(PA).
Results: TPB explained 59% and 43% of the variance for intention and behavior, respectively. Moderating (by
gender) and mediating tests were supported.
Conclusions: TPB is useful for understanding PA in this population.
䉷 2011 Society for Adolescent Health and Medicine. All rights reserved.
The Theory of Planned Behavior (TPB) [1] is a major socialcognitive theory that has been applied to explain physical activity
(PA) behavior in numerous populations. Briefly, the TPB proposes
that a person’s intention to perform a behavior is the central determinant of that behavior because it reflects the level of motivation a
person is willing to exert to perform the behavior [1]. Intention is
hypothesized to be determined by attitude, subjective norm, and
perceived behavioral control. Attitude is reflected in a positive or
negative evaluation of performing the behavior. Subjective norm is
defined as the perceived social pressure to perform the behavior,
whereas perceived behavioral control is defined as the perceived
ease or difficulty of performing the behavior. Perceived behavioral
control is also hypothesized to directly predict behavior. However,
the core tenets of the TPB have not been thoroughly tested for PA in
adolescent populations. The primary objective of this study was to
examine the explanatory power of the TPB to explain PA behavior in
a large population sample of adolescents. Secondary objectives
were (i) to examine the moderating effects of gender on the TPB,
and (ii) to test the mediating effects of intention on the relationship
between attitude, subjective norm, and perceived behavioral control with behavior.
Methods
The Web-Survey of Physical Activity and Nutrition was selfadministered to students in grades 7–10, across the province of
Alberta, Canada. Of the 59 school boards in the province, 85% agreed
to have schools participate. Of 271 schools contacted, 109 schools
across 37 school boards participated, reflecting a school board and
school response rates of 64% and 40%, respectively. Overall, 4,073 of
a potential 9,071 adolescents consented and completed the survey,
yielding a student response rate of 44.9%. Ethics approval was obtained from the University of Alberta.
Measures
* Address correspondence to: Ronald C. Plotnikoff, Ph.D., School of Education,
Faculty of Education and Arts, University of Newcastle, University Drive, Callaghan, NSW 2308, Australia.
E-mail address: [email protected] (R.C. Plotnikoff).
Validated, short measures (5-point response options) of core
TPB constructs (the measures of attitude, subjective norm, and
intention were modified from Hagger et al [2]) were used to
1054-139X/$ - see front matter 䉷 2011 Society for Adolescent Health and Medicine. All rights reserved.
doi:10.1016/j.jadohealth.2011.03.006
548
R.C. Plotnikoff et al. / Journal of Adolescent Health 49 (2011) 547–549
Table 1
Path coefficients, significance levels, and significance of the indirect effects of attitude, subjective norm, and perceived behavioral control
Variables
Single mediator modelsc
Attitude
Subjective norm
Perceived behavioral control
Multiple mediator modelsd
Attitude
Subjective norm
Perceived behavioral control
Effect sizea
d
Unstandardized regression coefficients and confidence intervals
Significance of indirect
effect
␣ (SE)
95% CI
␤ (SE)
95% CI
␣␤
.73 (.01)*
.34 (.01)*
.56 (.01)*
.70–.75
.32–.37
.54–.59
.23 (.01)*
.33 (.01)*
.16 (.01)*
.21–.26
.31–35
.14–19
.17
.11
.09
.15–.18
.10–.12
.08–.10
.20
.18
.12
.78 (.05)**
.05 (.02)**
.18 (.03)**
.70–.88
.02–.08
.12–.24
.08 (.01)**
.08 (.01)**
.08 (.01)**
.06–.11
.06–.11
.06–.11
.06
.004
.01
.05–.08
.001–.007
.01–.02
.07
.01
.02
95% CIb
␣ ⫽ estimate of unstandardized regression coefficient of attitude, subjective norm and perceived behavioral control predicting intention; ␤ ⫽ estimate of the
unstandardized regression coefficient of intention predicting physical activity with attitude, subjective norm or perceived behavioral control in the model; SE ⫽
standard error; 95% CI ⫽ 95% confidence interval; ␣␤ ⫽ product of coefficients estimate.
* p ⬍ .001, **p ⬍ .01.
a
Effect sizes calculated using the product of the standardized coefficients.
b
95% asymmetric confidence intervals of the mediated effect calculated using the Product Confidence Limits for the Indirect Effect program.
c
Single mediator models tested using Statistical Package for the Social Sciences and ordinary least squares regression.
d
Multiple mediator models tested in the full model using Analysis of Moment Structures output.
minimize response burden. A two-item construct was used to
assess attitude on the basis of enjoyment and PA importance [2].
A single-item assessed subjective norm [2] which asked: “Most
people important to me think I should take part in regular PA.” As
a proxy to perceived behavioral control, a 4-item self-efficacy
measure (␣ ⫽ .83) [3] assessed confidence in participating in PA
under various circumstances (e.g., when tired, having homework). A single-item construct was used to assess intention, for
instance, “I plan to be physically active on a regular basis over the
next month.” The validated Physical Activity Questionnaire for
Older Children [4] was used to assess PA levels of the participants
during the previous 7-day period.
Results
Overview
The sample reflects the age/sex distribution for Alberta youth.
The mean age of our sample was 13.6 (⫾1.4) versus 13.3 (⫾1.3)
years in the total population of Alberta of students in grades 7–10
(N ⬃ 130,000). The proportion of boys in our sample was 46%
versus 51% (grades 7–10) for the province. The sample (N ⫽
4,073) included 1,785 boys and 2,270 girls (18 students did not
provide gender).
Model testing
Model testing (primary objective)
Structural equation models were examined using analysis
of moment structures (AMOS) 17.0. To correct for the clustering of effects at the school level, all variables were adjusted for
school, using multiple linear regression, and the unstandardized residuals were used in the analyses. The proposed model
was tested using maximum likelihood analysis in AMOS and
the Physical Activity Questionnaire for Older Children score
was used as the dependent variable.
Moderation and mediation analyses (secondary objective)
Gender was identified as a potential moderator of the TPB
model. Multigroup moderation analyses were conducted using a
series of models, starting from unrestricted to fully constrained
[5]. A ⌬ comparative fit index (CFI) ⱕ ⫺.01 indicates that the null
hypothesis of invariance should not be rejected [6] for testing
multigroup invariance.
Intention is thought to mediate the relationship between
attitude, subjective norm, perceived behavioral control, and the
behavior itself. The indirect effects of attitude, subjective norm,
and perceived behavioral control were examined using single
and multiple mediator models and asymmetric confidence intervals tested the significance of the indirect effects [7]. The product
of the standardized coefficients was calculated to provide an
estimate of effect size and interpreted as small (d ⫽ .2), medium
(d ⫽ .5), and large (d ⫽ .8) [8].
Although the ␹2 statistic was significant (␹2 ⫽ 364.57, df ⫽ 22,
p ⬍ .001), the TPB model represented an excellent fit to the data
based on varying indices (goodness of fit [GFI] ⫽ .98, adjusted
goodness of fit [AGFI] ⫽ .96, CFI ⫽ .98, root mean square error of
approximation [RMSEA] ⫽ .06). All the pathways were significant (p ⬍ .001) and the model explained 59% and 43%, respectively, of the variance for intention and behavior (Figure 1).
Moderation analyses
The model was tested separately for boys and girls. In both
groups, the model was found to be an excellent fit to the data
(boys: ␹2 ⫽ 121.12, df ⫽ 22, p ⬍ .001, GFI ⫽ .99, AGFI ⫽ .97,
CFI ⫽ .99, RMSEA ⫽ .05; girls: ␹2 ⫽ 214.88, df ⫽ 22, p ⬍ .001,
GFI ⫽ .98, AGFI ⫽ .96, CFI ⫽ .98, RMSEA ⫽ .06). Model 1
(unrestricted model) was not significantly different from
model 2 (measurement of equivalent model which includes
equal factor loading across sub-samples). Model 3 (model
constraints plus equal factor variance and covariances) was
significantly different (⌬CFI ⫽ ⫺.08) from model 2, suggesting
that the relationship between constructs was stronger among
boys. Similarly, model 4 (included model 3 constraints plus
equal paths) was significantly different (⌬CFI ⫽ ⫺.03) from
model 3, indicating stronger path coefficients for boys. Model
5 (fully constrained model with the model 4 constraints plus
equal factor residuals) was also significantly different from
model 4 (⌬CFI ⫽ ⫺.10).
R.C. Plotnikoff et al. / Journal of Adolescent Health 49 (2011) 547–549
549
Figure 1. Standardized parameter estimates for pathways among constructs from the Theory of Planned Behavior in adolescents. Note. All pathways statistically
significant p ⬍ .001. Results for full sample in bold, boys/girls in parentheses.
Mediation analyses
The indirect effects of attitude, subjective norm, and perceived behavioral control are reported in Table 1. Intention to
be active was found to mediate the relationship between these
variables and PA in both single and multiple mediator models.
All the hypothesized mediation pathways were statistically
significant, supporting the role of intention to be physically
active as a mediator of the relationship between attitude,
subjective norm, perceived behavioral control, and PA behavior. The largest effects were found when intention was tested
as a mediator of the relationship between attitude and behavior (␣␤ ⫽ .20 in the single mediator model and ␣␤ ⫽ .07 in the
multiple mediator model).
Discussion
This research appears to be the largest TPB study on youth
conducted with a representative sample. Overall, our results
support the TPB with statistically significant values for all construct pathways. Perceived behavioral control and intention accounted for 43% of the variance in behavior for the overall sample. Attitude, subjective norm, and perceived behavioral control
explained 59% of the variance for intention. These results are
generally consistent with PA-TPB literature in both adolescent
and adult populations [9,10].
Differences existed in the strength of associations between
TPB constructs when comparing the gender-specific models.
However, for both genders, perceived behavioral control was
the strongest correlate of behavior, and attitude was the strongest correlate of intention. This highlights the importance of
practitioners/teachers to provide enjoyable activities and
strategies to enhance adolescents’ confidence to adopt and
sustain regular PA.
Finally, the assumptions for mediation were satisfied for all
three TPB constructs (i.e., attitude, subjective norm, perceived
behavioral control). Although the sizes of the mediated effects
were small, they did have important implications at the population level.
Acknowledgments
R.C.P. and L.T. conceived the study. R.C.P. was responsible for
producing the first draft of the entire manuscript. D.R.L. analyzed
the data and wrote the results section. S.A.C. and L.T. assisted in
writing the manuscript. S.D. was responsible for conducting the
study. J.C.S., L.M., and S.D. provided comments on manuscript
drafts. L.M., R.C.P., and J.C.S. were the co-investigators on the
related grant, funded by the Alberta Heritage Foundation for
Medical Research. All authors reviewed the final version of the
manuscript.
References
[1] Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process
1991;50:179 –211.
[2] Haggar MS, Chatzisarantis N, Biddle SJH, et al. Antecedents of children’s
physical activity intention and behaviour: Predictive validity and longitudinal effects. Psychol Health 2001;16:391– 407.
[3] Spence JC, Blanchard CM, Clark M, et al. The role of self-efficacy in explaining gender differences in physical activity among adolescents: A multilevel
analysis. J Phys Act Health 2010;7:176 – 83.
[4] Kowalski KC, Crocker PRE, Faulkner RA. Validation of the physical activity
questionnaire for older children. Pediatr Exerc Sci 1997;9:174 – 86.
[5] Nigg CR, Lippke S, Maddock JE. Factorial invariance of the theory of planned
behavior applied to physical activity across gender, age and ethnic groups.
Psychol Sport Exerc 2009;10:219 –25.
[6] Cheung GW, Rensvold RB. Evaluating goodness-of-fit indexes for testing
measurement invariance. Struct Equ Modeling 2002;9:233–55.
[7] MacKinnon DP, Fritz MS, Williams J, Lockwood CM. Distribution of the
product confidence limits for the indirect effect: Program PRODCLIN. Behav
Res Methods 2007;39:384 –9.
[8] Cohen J. Statistical power analysis for the behavioral sciences, 2nd edition.
Hillsdale, NJ: Lawrence Earlbaum Associates, 1988.
[9] Hagger MS, Chatzisarantis NLD, Biddle SJH. A meta-analytic review of the
theories of reasoned action and planned behavior in physical activity:
Predictive validity and the contribution of additional variables. J Sport Exerc
Psychol 2002;24:3–32.
[10] Symonds-Downs D, Hausenblas H. The theories of reasoned action and
planned behavior applied to exercise: A meta-analytic update. J Phys Act
Health 2005;2:76 –97.