An Experimental Test of the Theory of Planned Behavior

Page created by Kristen Sandoval
 
CONTINUE READING
APPLIED PSYCHOLOGY: HEALTH AND WELL-BEING, 2009, 1 (2), 257–270
                      doi:10.1111/j.1758-0854.2009.01013.x

           An Experimental Test of the Theory of
                    Planned Behavior
                                 Falko Sniehotta*
                             University of Aberdeen, UK

   The Theory of Planned Behavior (TPB) is one of the leading theories of health
   behavior, yet supporting evidence is exclusively based on correlational research.
   This study aims to test the TPB experimentally. N = 579 participants were
   randomised to receive persuasive messages addressing salient beliefs elicited
   in a pilot study, following a 2*2*2 factorial design. Participants were
   randomised to a behavioral-belief-intervention (BBI) or not, a normative-
   belief-intervention (NBI) or not, and a control-belief-intervention (CBI) or not.
   The primary outcome was objectively recorded attendance at university sports
   facilities over 2 months; and the secondary outcomes were post-intervention
   TPB measures. Main effects of the BBI on attitudes and of the NBI on subjec-
   tive norm, PBC, attitudes, and intentions were found. The CBI did not alter
   post-intervention cognitions, but was the only intervention to change behavior
   not mediated by cognitions. While the findings support the TPB’s assumptions
   on intention formation, behavior change results are not in line with the theory
   and therefore further question the TPB’s leading role in behavioral science.

   Keywords: physical activity participation,           randomised      trial,   theory
   development, Theory of Planned Behavior

                                 INTRODUCTION
Developing the science of behavior change is paramount for health psychol-
ogy. Behavior plays an increasingly important role in the prevention and
management of chronic disease. To date, the main causes of mortality and
morbidity in the US and the developed world are behaviors such as poor diet,
physical inactivity, smoking, and alcohol abuse (Mokdad, Marks, Stroup, &
Gerberding, 2004). Effective behavioral interventions addressing today’s key
health challenges should be based on tested scientific theory, rather than on
researchers’ and practitioners’ common sense and intuition (Campbell et al.,
2000; Rothman, 2004; Michie & Abraham, 2004).

  * Address for correspondence: Falko Sniehotta, University of Aberdeen, School of Psychol-
ogy, William Guild Building, Aberdeen AB24 2UB, UK. Email: f.sniehotta@abdn.ac.uk

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology. Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ,
UK and 350 Main Street, Malden, MA 02148, USA.
258     SNIEHOTTA

   The Theory of Planned Behavior (TPB; Ajzen, 1991) is among the leading
theories used to predict a range of health behaviors, such as physical activity
(Armitage, 2005; Eng & Martin Ginis, 2007; Johnston et al., 2007), eating
(Conner, Norman, & Bell, 2002), and other health behaviors (Kiene, Tennen,
& Armeli, 2008; van den Berg et al., 2008). The TPB is a social cognition
model that proposes that behavior is a linear function of behavioral inten-
tions and perceived behavioral control (PBC), the perception of individual
control over performing the behavior. Intentions, in turn, are assumed to be
a linear function of three types of cognitions: Attitude (positive or negative
evaluation of the behavior), subjective norm (perceived approval of perform-
ing the behavior), and PBC. Attitudes, subjective norm, and PBC are based
on a set of more specific salient behavioral, normative, and control beliefs
that reflect perceived outcomes associated with the target behavior (behav-
ioral beliefs), approval of important others (normative beliefs), and barriers
and facilitators (control beliefs) (Ajzen, 1991; Sutton, 2002).
   Over the last three decades, researchers have conducted hundreds of
studies applying and testing the TPB with regard to an array of social and
health behaviors (Armitage & Conner, 2001; Noar & Zimmerman, 2005).
This research is still dominated by “shortitudinal” research designs with very
short follow-up periods and by self-reported measures of behavior. This body
of research has not stimulated changes or development of the theory, i.e. the
evidence found did not convince the scientific community to modify, extend,
or abandon the TPB. This may be interpreted as an indication that the theory
is either very successful, or that the research conducted testing this theory has
been lacking rigor.
   The current evidence base for the TPB is dominated by between-
participant correlational studies (Noar & Zimmerman, 2005; Weinstein,
2007). These studies show that behavioral intention and PBC are indeed
predictive of behavior, accounting for about 28 per cent of its variance
(Armitage & Conner, 2001; Hagger, Chatzisarantis & Biddle, 2002; Sheeran,
2002). Surprisingly, rigorous tests of the theory from randomised experimen-
tal studies have been remarkably rare. A systematic review of studies apply-
ing the TPB to develop and test behavior change interventions found a lack
of studies systematically using the TPB to develop interventions and very
little evidence for their effectiveness (Hardeman et al., 2002). Since Hard-
eman’s review, some additional studies have been published testing the effects
of interventions developed based on the TPB; for example, Stead and colle-
ages found in a non-randomised longitudinal cohort study in Scotland that
the introduction of their intervention to reduce speeding was associated with
improved attitudes and affective beliefs (Stead, Tagg, MacKintosh, &
Douglas, 2005). More recently, the ProActive UK study showed that seden-
tary older adults with a parental history of type 2 diabetes randomly assigned
to an intensive 1-year behavior change intervention program based on an

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                259
extended TPB, delivered either face to face or via the telephone, did not show
higher levels of objectively measured physical activity than participants in a
measurement only control group (Kinmonth et al., 2008). Chatzisarantis and
Hagger (2005) tested two persuasive communications targeting salient vs.
non-salient behavioral beliefs about physical activity among young people.
While the intervention targeting salient behavioral beliefs improved attitudes
as hypothesised, this effect did not translate into significant differences in
intention and behavior. This is in line with Webb and Sheeran’s (2006)
findings that interventions that are successful in changing behavioral inten-
tions result in effects on behavior which are much smaller in effect size than
those on intentions.
   To provide a more stringent test of the TPB, full-factorial randomised
studies are needed that test independent effects of manipulating behavioral,
normative, and control beliefs, and use appropriate randomisation, sufficient
sample sizes, and thoroughly developed interventions (Sutton, 2002; Wein-
stein, 2007). The rationale for behavioral interventions within the TPB is to
address and modify the salient beliefs underlying attitudes, subjective norms,
and PBC, i.e. the most prevalent behavioral, normative, and control beliefs,
hypothesised to affect attitudes, subjective norms, and PBC, and, in turn,
intentions and, eventually, behavior (Ajzen, 2006).
   To our knowledge, no study has tested the TPB using a full-factorial
experimental design. McCarty (1981) tested the precursor of the TPB, Fish-
bein’s (1967) Theory of Reasoned Action, using an orthogonal experimental
design aiming to independently manipulate attitudes and subjective norms
using persuasive messages about contraceptive use. These interventions,
however, failed to affect the immediate targets, attitudes and subjective
norms. Thus the study does not provide evidence supporting the theory.

Aims and Hypotheses
This study provides the first full-factorial experimental test of the TPB using
a 2 (behavioral belief (BB) intervention vs. control) * 2 (normative belief (NB)
intervention vs. control) * (2 control belief (CB) intervention vs. control)
factorial design with post-intervention measure of TPB cognitions, an addi-
tional no-measurement control group to control for possible effects of this
measurement, and an objective record of the outcome behavior over 2
months following the intervention. A physical activity behavior was chosen
as the outcome to reflect the vast correlational evidence for the TPB in
predicting physical activity behaviors (Hagger et al., 2002). It was hypoth-
esised that (1a) the BB intervention will change attitudes, (1b) the NB inter-
vention will change subjective norms, and (1c) the CB intervention PBC,
(2) the interventions that successfully changed cognitions will also modify

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
260      SNIEHOTTA

intentions, and (3) the BB, NB, and CB interventions will increase behavior
and that this effect will be mediated by effects on intentions.

                                    METHODS

Design
Participants were randomly assigned following a 2 (behavioral belief inter-
vention vs. no intervention) * 2 (normative belief intervention vs. no inter-
vention) * (2 control belief intervention vs. no CB intervention) factorial
design with immediate post-intervention measurement of TPB cognitions and
subsequent objective recording of behavior.

Participants
Participants (N = 579; mean age 23.2 years, SD = 7.6; n = 382, 66% women)
were undergraduate students of a Scottish campus university participating in
a university-based online survey about physical activity and lifestyle behav-
iors at the beginning of the academic year (September/October 2007) prior to
randomisation.

Interventions
Interventions consisted of brief, online-delivered persuasive communications
providing information and addressing salient behavioral, normative, and
control beliefs identified in a previous belief elicitation study with 52 under-
graduate students of the same university with varying levels of physical
activity (diversity sampling), using a standard interview protocol (Ajzen,
2006; Francis et al., 2004).
  All intervention components were introduced as follows:

   We are interested in learning what factors are relevant for your decision to use
   the university’s sport and recreation services or not. Based on feedback from
   previous studies we have compiled a couple of statements summarising infor-
   mation about the services and we would ask you to tell us to what degree you
   were already aware of this information and how important this information is
   for your decision to use the sport and recreation services or not (link to webpage
   of S&R services).

For each type of belief, four brief paragraphs were presented providing
persuasive information addressing salient behavioral, normative, or control
beliefs elicited in the pilot study, respectively. Where possible, reference was
made to previous findings from the ongoing survey about students’ physical
activity and lifestyle behaviors.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                261
   The behavioral belief intervention emphasised: (a) the positive effects of
regular physical activity on health, fitness, mood, stress, and ability, (b) that
students attending the S&R (Sports and Recreation) services are more likely
to be physically active and to maintain physical activity over the term time
than students who do not attend the program (based on previous evidence
from a larger ongoing study), (c) the safety and low injury risk of physical
activity in the S&R facilities due to induction, monitoring, and availability of
trained support, and (d) the wide range of exercise classes and individual
activities (i.e. swimming, fitness suite, squash, etc.) and the flexibility of the
timetable with a website that allows one to tailor the program to one’s fitness
level, interests, and availability.
   The normative belief intervention emphasised: (a) that most students report
that not having somebody to actively participate in physical activity with them
is a major barrier to becoming more physically active. Students who attend the
S&R services find that they meet people who do approve, support, and
participate with them in physical activity which creates the necessary level of
social approval and support, (b) that most friends and family actually approve
of involvement in physical activity which is safe, secure, and healthy, (c) that
misperceptions about others’ disapproval for involvement in regular physical
activity often comes down to competing expectations about the time spent for
oneself rather than disapproval for the activity, and that participants in the
pilot study felt that this could often be turned into approval by explicitly
talking to important others and addressing potential time conflicts together,
and (d) that perceptions of regular S&R users as fit, well-trained super-athletes
who look down on regular people trying to get fitter are far from reality and
that the users of the services are a cross-section of the university community.
   The control belief intervention addressed the four key barriers to participa-
tion found in the initial pilot study: (a) costs: our pilot studies have shown
that students generally overestimate the costs for admittance. We therefore
provided information about the actual costs of GBP£1.60 (approx.
USD$2.65) for a single visit and GBP£82 (approx USD$135) for an annual
membership for students to address this barrier, (b) time: students’ main
concern of not having enough time was addressed by emphasising the long
and flexible opening hours that allow students to schedule sessions before,
between, and/or after their lectures, or over lunchtime, or choose from over
35 classes each week throughout the day, every day, using the services website
as well as the possibility of combining socialising with physical activities, (c)
access: ease of access was emphasised and information about locations,
parking, and public transport from the university’s remote hospital campus
were provided, and (d) feelings of discomfort and embarrassment about
exercising in public were addressed. This issue came up frequently in the pilot
study and showed that the idea about the typical user of the facilities is often
quite biased.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
262      SNIEHOTTA

                                   TABLE 1
        Bivariate Correlations between Post-Intervention TPB Measures
       and Attendance Records, Means, Standard Deviations, Range and
                               Cronbach’s Alphas

                         1             2             3             4             5

1. Attitude                          .56**         .63**         .75**         .22**
2. Subjective norm                                 .61**         .67**         .19**
3. PBC                                                           .79**         .31**
4. Intention                                                                   .34**
5. Attendance
Cronbach’s alpha         .95           .78           .90           .95            –
Mean (SD)            5.46 (1.44)   5.10 (1.52)   4.75 (1.64)   4.97 (1.87)   1.05 (2.15)
Range                   1–7           1–7           1–7           1–7           0–9

  In order to facilitate elaboration of the persuasive messages, two questions
were asked following each paragraph: (1) Were you aware of this information?
(answer scale: yes/no) and (2) How important is this information for your
decision to personally use the S&R facilities? (Not important at all (1)–very
important (7)).

Control Measures
Attitudes, subjective norms, PBC, and behavioral intention were measured at
baseline to perform manipulation checks and mediation analyses using a
validated physical activity-based TPB questionnaire by Armitage (2005). The
behavior reference in each item was modified to “participate in the univer-
sity’s sports and recreation program throughout the next two months” to
optimise correspondence between predictors and behavioral measure (Ajzen,
2006). Means, standard deviations, ranges, and Cronbach’s alphas are
displayed in Table 1.

Outcomes
Primary outcome was the number of weeks that participants attended the
university’s sport and recreation facilities over 2 months between baseline
measurement and the university’s 2007 Christmas break, based on objective
attendance records. As the primary outcome measure is based on objective
records, no attrition was encountered. Students are required to swipe their
student ID cards to get access to S&R facilities.

Randomisation
Participants were randomly assigned to receive the BB intervention or not, to
receive the NB intervention or not, and to receive the CB intervention or not

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                263
following a 2*2*2 factorial design. Randomisation occurred after completing
a general health-related online questionnaire using a pearl script integrated
into the online questionnaire.

Blinding
Double-blinded randomised trial: Participants were unaware that the study
would include assignment to groups. Assessment of baseline measures was
blinded by online measurement and outcomes were measured objectively.

Ethical Approval
The study received ethical approval from the School of Psychology Ethical
Review Board at the University of Aberdeen.

Statistical Methods
Data were analysed using SPSS 16. Possible intervention effects on control
measures and outcomes were tested using analyses of variance (ANOVAs)
based on intention-to-treat analyses. Relationships between post-
intervention TPB measures and subsequent records of S&R attendance were
tested using linear regression analyses.

                                  RESULTS
Table 1 shows bivariate correlations between post-intervention TPB mea-
sures and attendance records as well as means, standard deviations, and
Cronbach’s alphas. Generally, social cognitions towards using the sports and
recreation facilities were positive (above the mean of the 1–7 scales). Atten-
dance measures showed a mean of 1.05 weeks attended throughout the study
period. All post-intervention TPB variables showed significant intercorrela-
tions with large effects and moderate to large correlations with attendance,
with the highest correlations of r = .34, p < .001 for intentions and r = .31,
p < .001 for PBC. Using linear regression analyses, behavioral intentions were
accurately predicted by attitude (b = .37; p < .001), subjective norm (b = .20;
p < .001), and PBC (b = .43; p < .001) accounting for 75.5 per cent of variance
in intentions. Post-intervention TPB measures were predictive of attendance
over the following 2 months, with intention (b = .27; p < .001), but not PBC
(b = .10; p = .14), significantly contributing to the prediction (R2 = .124;
R2adjusted = .121; p < .001). This is in line with meta-analytical reviews of TPB–
behavior relationships in which the predictive power is lower for objective
outcome measures and longer follow-up periods (Hagger et al., 2002).

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
264     SNIEHOTTA

Adding attitude and subjective norm to the equation did not improve the
prediction.

Randomisation Checks (Effects on TPB Measures)
Participants allocated to the behavioral belief intervention showed more
favorable attitudes towards using the sport and recreation facilities
(F(1, 530) = 6.05; p = .014; h2 = .011). Estimated marginal mean attitude for
the behavioral belief condition was 5.59 (SE = .09) vs. 5.28 (SE = .09) for
controls on a scale ranging from 1 to 7. No main effect of the control belief
intervention was found (F(1, 530) = .295; ns) and the normative belief inter-
vention approached significance (F(1, 530) = 3.69; p = .055) (Figure 1a).
Neither of the two-way interactions nor the three-way interaction was
significant.
   The normative belief intervention successfully manipulated subjective
norms (F(1, 531) = 4.43; p = .036; h2 = .008), estimated marginal mean
subjective norms was 5.22 (SE = .10) for the normative belief intervention
and 4.94 (SE = .09) for controls (Figure 1b). While neither the control belief
intervention (F(1, 531) = .08; ns) nor the behavioral belief intervention
affected subjective norms (F(1, 531) = 1.53; ns), an unexpected interaction
between the behavioral belief and the control belief interventions reached
significance (F(1, 531) = 4.97; p = .026; h2 = .009), showing that the attitude
intervention increased subjective norm in the absence of the PBC interven-
tion. No other interaction was significant.
   PBC was not affected by the intervention addressing control beliefs (F(1,
533) = .06; ns). However, the normative belief intervention had a positive
main effect on PBC (F(1, 533) = 5.08; p = .025; h2 = .010); estimated marginal
mean PBC was 4.89 (SE = .10) for the normative belief condition and 4.58
(SE = .98) for controls. No other main or interaction effects were found.
   Behavioral intentions were significantly improved by the intervention
addressing normative beliefs (F(1, 531) = 10.67; p < .001; h2 = .020; estimated
marginal mean NB = 5.23 (SE = .12), estimated marginal mean con-
trols = 4.71 (SE = .11)). No other effect reached significance. In conclusion,
manipulation checks show that the interventions successfully manipulated all
TPB measures including intentions.

Effects on Primary Outcome
The full-factorial ANOVA for the primary outcome, attendance to the sport
and recreation services, showed a small main effect of the control belief
intervention on attendance (F(1, 576) = 4.03; p = .045; h2 = .007). Participants
attended the services an average of 1.22 (SE = .13) weeks in the past 2
months, whereas controls attended 0.86 (SE = .13) weeks (Figure 2). The

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                265

FIGURE 1 a. Post-intervention Theory of Planned Behavior measures for
participants assigned to the behavioral belief intervention vs. controls;
b. Post-intervention Theory of Planned Behavior measures for participants
assigned to the normative belief intervention vs. controls.

interventions focusing on normative beliefs (F(1, 576) = 2.00; p = .16) and
behavioral beliefs (F(1, 576) = .15; p = .70) as well as the two-way and three-
way interactions were not significant.

Mediation Analyses
The effect of the normative belief intervention on intentions was partially
mediated by subjective norms and PBC. An ANCOVA of the intervention
effects on behavioral intentions shows that the main effect on intention
remains significant (F(1, 529) = 3.87; p = .049) even after subjective norms
(F(1, 529) = 95.63; p < .001) and PBC (F(1, 529) = 272.51; p < .001), both
measures modified by the NB intervention, are controlled for as covariates.
All other effects were not significant.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
266     SNIEHOTTA

FIGURE 2. Mean weeks attended the sport and recreation facilities in the 2
months following intervention for participants assigned to the control belief
intervention vs. controls.

   The control belief intervention was the only intervention to affect the
behavioral outcome. As the control belief intervention did not affect cogni-
tive TPB measures and, in turn, the normative belief and behavioral belief
interventions’ success in affecting TPB cognitions were unsuccessful in chang-
ing behavior, the conditions for mediation analyses were not fulfilled (Baron
& Kenny, 1986). The TPB measures did not mediate the intervention effect
on behavior.

                                DISCUSSION
This study conducted the first full-factorial experimental test of the TPB
using post-intervention assessment of TPB cognitions and objectively
recorded behavioral outcome measures over a meaningful time period
following the interventions. The three brief persuasive communications to
address modal salient behavioral, normative, and control beliefs were devel-
oped based on standard procedures within the TPB, including a prior belief
elicitation study using a standard protocol (Ajzen, 2006). These brief inter-
ventions, delivered online as part of a survey, resulted in small changes in
cognitions related to the use of the university’s sport and recreation facilities.
Both attitudes and subjective norms were successfully increased by the inter-
ventions addressing the respective behavioral and normative beliefs. PBC was
also manipulated, but not as hypothesised by the control belief intervention
but rather by the subjective norm intervention. This may reflect the proposed
overlapping nature of TPB cognitions, supported by an abundance of studies
showing that they are highly correlated (Armitage & Conner, 2001). Beliefs

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                267
elicited by questions about social approval might also bring up issues that
have implications for the controllability and ease of performing a behavior.
For instance, in our pilot study, questions eliciting both normative and
control beliefs brought up beliefs about social disapproval from stereotypical
users of sports services (e.g. lean, muscular, athletic) serving as a social
barrier to participation, causing social embarrassment of new users perceiv-
ing themselves as less in shape. This might explain the unexpected side effect
of the subjective norm intervention. The question of the discriminant con-
struct validity (Pollard, Johnston, & Dieppe, 2006) of interventions is a
challenge to social cognition models that will need addressing in the future.
   All three TPB predictors of behavioral intentions were manipulated by the
interventions. The normative belief intervention increased intentions, which
was partly mediated by changes in subjective norm and PBC. These effects,
therefore, confirm the TPB’s assumptions regarding intention formation.
However, the effect of the normative belief intervention on intention did not
translate into an effect on behavior. Instead, the control belief intervention
had a small significant effect, increasing attendance behavior. This is in line
with evidence that interventions targeting control or self-efficacy beliefs are
effective in changing behavior directly (Bandura, 1997). It has also previously
been shown that interventions targeting behavioral control beliefs may influ-
ence behavior without affecting self-report measures of control beliefs
(Johnston et al., 2007). It is also possible that the effect might have been
mediated by more specific control beliefs than those measured in the present
study (Rodgers, Murray, Courneya, Bell, & Harber, 2009).
   All effect sizes were very small. This partly reflects the challenge of changing
cognitions about physical activity behaviors and service use among well-
informed students who will likely have formed stable beliefs prior to entering
the study. However, it also indicates that the proposed procedures for inter-
vention development within the TPB (Ajzen, 2006) and the prevalent research
traditions associated with the theory do not provide effective tools and tech-
niques to change cognitions (Sniehotta, 2009). For example, the effectiveness
of persuasive messages on behavior change is very limited (Hillsdon, Foster,
Cavill, Crombie, & Naidoo, 2005). Thus, theories explicitly incorporating
behavior change techniques such as Social Learning Theory (Bandura, 1986)
are more useful for theory testing and intervention design. This study shows
that small changes in cognitions do not lead to behavior change; it is possible
that large changes in cognitions do (cf. Webb & Sheeran, 2006). However, the
TPB provides no guidance on how to achieve such large changes in cognitions.
   In contrast to the experimental findings, regression of intention and behav-
ior on post-intervention TPB measures confirms TPB assumptions that atti-
tudes, subjective norms, and PBC are highly (cross-sectionally) predictive of
intentions, and PBC and intention (in our study only intentions) are predictive
of behavior. While these predictive findings are well in line with existing

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
268     SNIEHOTTA

evidence (Armitage & Conner, 2001; Sheeran, 2002), they are misleading
because the experimental findings in the same study suggest that the TPB does
not explain behavior (see Weinstein, 2007). This study suggests that TPB
assumptions do not stand up to experimental tests, despite strong predictive
effects using correlational approaches. While the changes in intention found
in this study are in line with TPB assumptions, the theory does not account for
the findings on behavior change found in this experimental study. This limits
the TPB’s potential to contribute to the science of behavior change.
   So what is wrong with the TPB? The theory’s various conceptual problems
have been discussed (Ogden, 2003; Sniehotta, 2009). The key shortcomings
from an experimental and behavior change point of view are that (a) the
theory does not specify techniques to modify hypothesised cognitive deter-
minants of intention and behavior, (b) possible changes in beliefs will be
attenuated through the hypothesised causal chain of events from beliefs, to
intention, to behavior caused by the imperfect empirical relationships
between these variables, and (c) the TPB does not account for intention–
behavior discrepancies (Sniehotta, 2009). Thus, the Theory of Planned
Behavior is neither about planning, as it does not address how people trans-
late their intention into behavior, nor an accurate theory of behavior
(Schwarzer, 2008; Sniehotta, Scholz, & Schwarzer, 2005). While future
research is needed to confirm these findings, it calls into serious question the
leading role of the TPB in the health psychology literature.
   It is time for a new research agenda to be set, aimed at testing and devel-
oping theories of health behavior by using more rigorous tests and designs
(Weinstein, 2007) and setting criteria for abandoning theories which fail these
tests (West, 2005).

                         ACKNOWLEDGEMENTS
The author would like to thank Silje Skår for her support in conducting this
study, Justin Presseau for proofreading the manuscript, and Vera Araújo-
Soares and Justin Presseau for helpful comments on a previous version of this
manuscript.

                                REFERENCES
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
   Decision Processes, 50, 179–211.
Ajzen, I. (2006). Designing a TPB questionnaire. Retrieved 10 October 2008 from
   http://people.umass.edu/aizen/pdf/tpb.measurement.pdf.
Armitage, C.J. (2005). Can the theory of planned behaviour predict the maintenance
   of physical activity? Health Psychology, 24, 235–245.
Armitage, C.J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A
   meta-analytic review. British Journal of Social Psychology, 40, 471–499.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
TEST OF THE THEORY OF PLANNED BEHAVIOR                    269
Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ:
   Prentice Hall.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.
Baron, R.M., & Kenny, D.A. (1986). The moderator–mediator variable distinction in
   social psychological research: Conceptual, strategic, and statistical considerations.
   Journal of Personality and Social Psychology, 51, 1173–1182.
Campbell, M., Fitzpatrick, R., Haines, A., Kinmonth, A.L., Sandercock, P.,
   Spiegelhalter, D. et al. (2000). Framework for design and evaluation of complex
   interventions to improve health. British Medical Journal, 321, 694–696.
Chatzisarantis, N., & Hagger, M. (2005). Effects of a brief intervention based on the
   theory of planned behaviour on leisure time physical activity participation. Journal
   of Sport and Exercise Psychology, 27, 470–487.
Conner, M., Norman, P., & Bell, R. (2002). The theory of planned behavior and
   healthy eating. Health Psychology, 21, 194–201.
Eng, J.J., & Martin Ginis, K.A. (2007). Using the theory of planned behavior to
   predict leisure time physical activity among people with chronic kidney disease.
   Rehabilitation Psychology, 52, 435–442.
Fishbein, M. (1967). Readings in attitude theory and measurement. New York: Wiley.
Francis, J.J., Eccles, M.P., Johnston, M., Walker, A.E., Grimshaw, J.M., Foy, R.
   et al. (2004). Constructing questionnaires based on the theory of planned behav-
   iour. In A manual for health services researchers. Centre for Health Services
   Research, University of Newcastle upon Tyne.
Hagger, M.S., Chatzisarantis, N.L.D., & Biddle, S.J.H. (2002). A meta-analytic
   review of the theories of reasoned action and planned behavior in physical activity:
   Predictive validity and the contribution of additional variables. Journal of Sport
   and Exercise Psychology, 24, 3–32.
Hardeman, W., Johnston, M., Johnston, D.W., Bonetti, D., Wareham, N.J., & Kin-
   month, A.L. (2002). Application of the Theory of Planned Behaviour in behaviour
   change interventions: A systematic review. Psychology and Health, 17, 123–158.
Hillsdon, M., Foster, C., Cavill, N., Crombie, H., & Naidoo, B. (2005). The
   effectiveness of public health interventions for increasing physical activity among
   adults: A review of reviews. London: Health Development Agency (http://
   www.publichealth.nice.org.uk/page.aspx?o=505281).
Johnston, M., Bonetti, D., Joice, S., Pollard, B., Morrison, V., Francis, J.J. et al.
   (2007). Recovery from disability after stroke as a target for a behavioural inter-
   vention: Results of a randomised controlled trial. Disability and Rehabilitation, 29,
   1117–1127.
Kiene, S.M., Tennen, H., & Armeli, S. (2008). Today I’ll use a condom, but who
   knows about tomorrow: A daily process study of variability in predictors of
   condom use. Health Psychology, 27, 463–472.
Kinmonth, A.-L., Wareham, N.J., Hardeman, W., Sutton, S., Prevost, A.T., Fan-
   shawe, T. et al. (2008). Efficacy of a theory-based behavioural intervention to
   increase physical activity in an at-risk group in primary care (ProActive UK): A
   randomised trial. The Lancet, 371, 5–7.
McCarty, D. (1981). Changing contraceptive usage intentions: A test of the Fishbein
   model of intention. Journal of Applied Social Psychology, 11, 192–211.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
270      SNIEHOTTA

Michie, S., & Abraham, C. (2004). Interventions to change health behaviours:
   Evidence-based or evidence inspired? Psychology and Health, 19, 29–49.
Mokdad, A.H., Marks, J.S., Stroup, D.F., & Gerberding, J.L. (2004). Actual causes
   of death in the United States, 2000. Journal of the American Medical Association,
   291, 1238–1245.
Noar, S.M., & Zimmerman, R.S. (2005). Health behavior theory and cumulative
   knowledge regarding health behaviors: Are we moving in the right direction?
   Health Education Research, 20, 275–290.
Ogden, J. (2003). Some problems with social cognition models: A pragmatic and
   conceptual analysis. Health Psychology, 22, 424–428.
Pollard, B., Johnston, M., & Dieppe, P. (2006). What do osteoarthritis health
   outcome instruments measure? Impairment, activity limitation, or participation
   restriction? Journal of Rheumatology, 33, 757–763.
Rodgers, W., Murray, T., Courneya, K.S., Bell, G., & Harber, V.J. (2009). The
   specificity of self-efficacy over the course of a progressive exercise programme.
   Applied Psychology: Health and Well-Being, 1.
Rothman, A.J. (2004). Is there nothing more practical than a good theory? Why
   innovations and advances in health behavior change will arise if interventions are
   more theory-friendly. International Journal of Behavioral Nutrition and Physical
   Activity, 1, 11.
Schwarzer, R. (2008). Modeling health behavior change: How to predict and modify
   the adoption and maintenance of health behaviors. Applied Psychology: An
   International Review, 57, 1–29.
Sheeran, P. (2002). Intention–behaviour relations: A conceptual and empirical review.
   European Review of Social Psychology, 13, 1–36.
Sniehotta, F.F. (2009). Towards a theory of intentional behavior change: Plans,
   planning and self-regulation. British Journal of Health Psychology, 14, 261–273.
Sniehotta, F.F., Scholz, U., & Schwarzer, R. (2005). Bridging the intention–behaviour
   gap: Planning, self-efficacy, and action control in the adoption and maintenance of
   physical exercise. Psychology and Health, 20, 143–160.
Stead, M., Tagg, S., MacKintosh, A.M., & Douglas, E. (2005). Development and
   evaluation of a mass media Theory of Planned Behaviour intervention to reduce
   speeding. Health Education Research, 20, 36–50.
Sutton, S. (2002). Testing attitude–behaviour theories using non-experimental data:
   An examination of some hidden assumptions. European Review of Social
   Psychology, 13, 293–323.
van den Berg, M., Timmermans, D.R.M., Knol, D.L., van Eijk, J.T.M., de Smit, D.J.,
   van Vugt, J.M.G. et al. (2008). Understanding pregnant women’s decision making
   concerning prenatal screening. Health Psychology, 27, 430–437.
Webb, T.L., & Sheeran, P. (2006). Does changing intentions engender behavior
   change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132,
   249–268.
Weinstein, N.D. (2007). Misleading tests of health behavior theories. Annals of
   Behavioral Medicine, 33, 1–10.
West, R. (2005). What does it take for a theory to be abandoned? The transtheoretical
   model of behaviour change as a test case. Addiction, 100, 1048–1050.

© 2009 The Author. Journal compilation © 2009 International Association of Applied
Psychology.
You can also read