You are on page 1of 7

International Journal of Behavioral

Nutrition and Physical Activity Central


Med
Bio

Debate Open Access


"Is there nothing more practical than a good theory?": Why
innovations and advances in health behavior change will arise
if
interventions are used to test and refine theory
Alexander J Rothman*

Address: Department of Psychology University of Minnesota Minneapolis, MN USA

Email: Alexander J Rothman* - rothm001@umn.edu


* Corresponding author

Published: 27 July 2004 Received: 22 June 2004


Accepted: 27 July 2004
International Journal of Behavioral Nutrition and Physical Activity 2004, 1 :11 doi:10.1186/1479-5868-1-11

This article is available from: http://www.ijbnpa.org/content/1/1/11

2004 Rothman; licensee BioMed Central Ltd.


This is an open-access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Theoretical and practical innovations are needed if we are to advance efforts to persuade and
enable people to make healthy changes in their behavior. In this paper, I propose that progress in
our understanding of and ability to promote health behavior change depends upon greater
interdependence in the research activities undertaken by basic and applied behavioral scientists. In
particular, both theorists and interventionists need to treat a theory as a dynamic entity whose
form and value rests upon it being rigorously applied, tested and refined in both the laboratory and
the field. To this end, greater advantage needs to be taken of the opportunities that interventions
afford for theory-testing and, moreover, the data generated by these activities need to stimulate
and inform efforts to revise, refine, or reject theoretical principles.

Background Investigators continue to advocate for a broad range of


Even with the dramatic advances in our understanding of theories and there has been limited progress in demon-
the biological processes that determine health and illness, strating the unique value of any specific theory. [6-8]
it has never been more clear that rates of disease morbid-
ity and premature mortality reflect people's behavioral Although there may be consensus in the professional
practices. [1] The benefits, both for individuals and the community that there are considerable gaps in our under-
societies in which they live, that would come from system- standing of health behavior change, critiques of the cur-
atic improvements in diet, physical activity, and use of rent state of affairs more often that not reflect the
substances
tantalizing
tiatives
behavior
the
tions
iors
methods
ior
in theories
have
commitment
have
and
tobeen
that
change
advances
been
elicit
of
and
such
produce
health
elusive
modest.
provide
changes
as
of
hastobacco,
in
considerable
behavior
proven
[2-5].
our
sustained
In
ample
inability
particular,
health
At
alcohol,
a have
worthy
motivation
the
improvements
time
tobehavior.
same
also
improve
and
the
and
adversary.
been
time,
specification
illicit
to
effort,
Yet,
develop
health
modest.
innovations
drugs
inhealth
Despite
innova-
behav-
behav-
ini-
are
of behavior
professional
totoregulate
be
interventions
specify
behavior
the
design
that
lament
make
most
people's
specifies
the
interests
and
change.
effective
that
itstructural
are
difficult
implement
the
behavior
not
how
ofAt
(p apreponderance
targets
guided
gthe
ethe
they
and
tonucritic.
discern
lament
same
health
mpsychological
are
bfor
by
e r supposed
nInvestigators
aintervention.
time,
obehavior
what
theoretical
the
t fof
o fact
rinvestigators
theories
factors
ci t ato
processes
that
tinterventions
ioelicit
Page
framework
nwho
Moreover,
too
are
of health
1likely
p u rpmany
strive
owho
of
that
se s7)
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

it is argued that theories are not sufficiently specified to improve the quality of the theoretical models available to
determine when or how to modify factors that are to be guide subsequent intervention efforts.
targeted in an intervention.
Discussion
Of course, concerns regarding the link between theory and When is an intervention effective?
practice are not new and efforts to address this problem Interventions are designed to address important practical
have taken several forms. Considerable effort has been problems (e.g., obesity) and thus their value is inextrica-
given to provide practitioners with a comprehensive and bly linked to their ability to alleviate the targeted problem.
concise understanding of the array of theories that have Interventions need to provide a meaningful return on the
been developed to address health behavior. [9] Moreover, time, money, and effort invested such that the outcomes
conceptual frameworks such as PRECEDE-PROCEED [10] afforded by a intervention strategy are proportional to the
and Intervention Mapping [11] have been developed to resources utilized. Of course, determining what is a suffi-
provide investigators with a structured process to improve cient return on an investment can be a challenge. Small
the accuracy and ease with which theoretical concepts are effects may be impressive if the intervention is directed at
used to address a practical problem. In both cases, these a construct or behavior that is considered difficult to
efforts have targeted improving how theoretical principles move. [12] In addition, interventions can have minimal
are applied and, in doing so, have relied on the assump- impact on an individual's behavior but when dissemi-
tion that current theories of health behavior are useful and nated widely have a dramatic impact at the societal level.
productive. Is this assumption valid? Could the often [13]
repeated plea for investigators to ground their interven-
tion efforts in theory be a sign that there are significant What conditions are likely to facilitate a successful inter-
limitations to the practical principles that can be derived vention? Broadly speaking, an intervention is most likely
from current theories of health behavior? If so, merely to be effective if it is appropriately grounded in the practi-
improving how people use theories will not be sufficient. cal problem targeted. [11] For example, consider an inter-
What is needed is a shift in how we engage the interplay vention to promote healthy food choices. The
between theory and practice, with an emphasis placed on intervention design team must possess a clear understand-
developing initiatives that target opportunities to ing of who is engaging in the targeted behavior (e.g., who
develop, test, refine health behavior theory. is making unhealthy food choices), the underlying nature
of the behavior (e.g., the frequency and function of food
In this paper, I describe and advocate for a model of col- choices), and the context in which the behavior is per-
laboration between basic and applied behavioral scien- formed (e.g., where and with whom do people make
tists. Although I recognize the value of improving the choices about food). In a similar manner, the intervention
manner in which theoretical principles are matched to needs to be appropriately grounded in the biological,
problems and methods, I propose that innovations in our structural and psychological processes that shape and reg-
understanding of and ability to promote health behavior ulate people's behavioral practices. [14-16] For example,
change will not arise if theory is construed as a fixed entity the expected value of altering a feature of the environment
that is delivered to interventionists for implementation. in which people make food choices (e.g., increasing the
To date, although theories may fluctuate in their popular- cost of high-fat foods) is predicated on the assumption
ity, their properties have remained strikingly static over that the intervention will directly, or indirectly through an
time. I believe greater attention must be paid to refining intervening construct, influence people's food choice in
and, when necessary, rejecting theoretical principles. For that setting.
this process to take shape, there needs to be an on-going
series of exchanges between theorists and interventionists Health behavior theories provide an explicit statement of
in which theory is treated as a dynamic entity whose value the structural and psychological processes that are
depends on it being not only applied and tested rigor- hypothesized to regulate behavior (e.g., increasing the
ously,
those
A but
improvements
vention
are
afford
aingiven also
fundamental
data
more
tests.
for
that
theory.
methods refined
receptive
theory
canin
These based
implication
reveal
testing,
depend
both
to the
data
the
health
on on
there the
opportunities
adequacies
will,
of
each
behavior
this
will
in findings
other.
perspective
turn,
be and
atheory afforded
interventions
If
enable
dramatic
investigators
inadequacies
and byof
istheorists
that
increase
inter- to cost
cult
guide
foods
mines aoftheory's
behavior.
the
tician,
designed ahigh-fat
toaccurate
quality
people's
purchase
by making
[17] itfoods
theory's
and
and
account
behavior
Although
them). ewill
implemented,
value?
predictive
more
value
(p aof
If
gand
From curtail
investigators
theories
the
aversive
rests
nuvalue
justify
mfactors
bthe tconsumption
einterventionists
on
r ndescribe
of
perspective
its
or,
ohow
that
aability
fperhaps,
theory.
omay
r an
regulate
ci tthe
intervention
arecognize
to
t ioWhat
of
Page of
factors
more
nprovide
depend these
appeople's
utheore-
2deter-
rp diffi-
oof
that
seon
an
is
s7)
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

behavior is affected by factors at different levels of analysis someone who is satisfied is more likely to sustain a pat-
(i.e., biological, psychological, social, environmental), a tern of behavior does not indicate what causes someone
theory's value is not necessarily predicated on its ability to to be satisfied and, thus, little guidance is given as to what
provide linkages across these levels. Because of this can be done to heighten the satisfaction people derive
emphasis, theory testing tends to occur in controlled con- from changes in their behavior. In the absence of this type
texts, typically a laboratory setting, that afford the social of information, interventionists may find little difference
and behavioral version of a Petrie dish. This approach between developing intervention strategies that are or are
allows investigators to observe the relation between a not grounded in a health behavior theory. In fact, given
given set of constructs with greater precision, but it these practical needs, it is not be surprising that interven-
renders the generalizability and strength of the observed tionists are more likely to rely on health behavior theories
effect difficult to discern. For example, investigators may (e.g., Social Cognitive Theory [22]) that specify the deter-
determine that focusing people's attention on the undesir- minants of its primary constructs and thus provide guid-
able aspects of an object increases their interest in avoid- ance as to how to construct an intervention protocol.
ing it, but be unable to specify the conditions under which
this relation is and is not most likely to obtain. Breakdown in the evolution of health behavior theories
If the design and implementation of intervention strate-
From the perspective of an interventionist, the accuracy of gies rely on assumptions regarding the factors that regu-
the relations specified in a theory is an important but not late people's behavior, why haven't current theories of
sufficient determinant of its value. Interventionists need health behavior evolved in ways that would enable them
theories that are accurate and applicable; that specify not to more effectively guide intervention development? I
only the relation between two constructs, but also believe the critical problem is that there has been a break-
whether that relation does or does not change across con- down in the relation between basic and applied scientists
texts (e.g., does the impact of risk perceptions on behavior who study health behavior. [23] As scholars such as Kurt
differ whether one is examining decisions to test for radon Lewin [24] have asserted, the development and specifica-
or to start smoking?). Given a set of a factors hypothesized tion of theories of human behavior depend upon an iter-
to regulate people's behavior, interventionists need to be ative series of research activities in which theoretical
able to discern which of these factors are the most appro- principles initially formulated by basic behavioral scien-
priate targets for intervention. In fact, a common com- tists are tested and evaluated by applied behavior scien-
plaint regarding theories is that they are not useful (See tists. These tests provide critical information that enables
Jeffery, this issue). A theory may specify a host of factors basic scientists to revise, refine, or reject their initial prin-
that regulate a person's behavior, but in the absence of ciples. Moreover, an applied setting can afford investiga-
information regarding the relative importance of each fac- tors the opportunity to assess the relative impact of
tor leave an interventionist unsure as to where to direct different processes hypothesized to regulate people's
her or his resources. For example, the Theory of Planned behavior. It is through this on-going cycle of specification,
Behavior [18] and Theory of Reasoned Action [19] pro- application, and evaluation that accurate and applicable
pose that people's attitudes toward the behavior and their theoretical models arise.
perceived subjective norm regarding the behavior are crit-
ical determinants of behavior (albeit mediated by behav- To the extent that behavioral theories are not tested in
ioral intention), but the relative contribution of these complex social settings such as those afforded by interven-
constructs is allowed to fluctuate from setting to setting. tions to change health practices, the process by which the-
In any given context, it is unclear how to determine a priori
ories develop is curtailed. Because the manner in which a
which set of constructs should be prioritized as a target for theory is specified reflects, in part, the contexts in which it
intervention. The interest interventionists have shown in has been operationalized and tested, theories that are
stage-based models of health behavior may reflect the fact tested primarily in tightly controlled laboratory settings
that the models attempt to specify the conditions under will likely be characterized by a rich description of the
whichasguidance
Little
critical
outcomesspecific
colleagues
determinant
such
measures
predict constructs
aconstructs
this
behavioral
of
afforded
and
are
of
a construct,
behavioral
Iis
typically
have
also
can
by affect
outcome.
proposed
be
agiven
pattern
in behavioral
predicated
manipulated.
maintenance.
this
asYet,
to
case
of
that
how
behavior
the
on decisions.
satisfaction
satisfaction,
observation
For
or
evidence
[20,21]
even
example,
is awhether
critical
Claims
with
that
that
my[8]
uniquely
the social myriad
ingminimize
affect
choices.
factors of
determining
that factors
environments,
empirically
behavior.
esses
and
The
noise
that
thus that
whether,
laboratory
and
[17] could
optimizes
test
investigators
(p a potential
gThus,
etheoretical
affect
in
can
nusetting
a bin
m affect
more
their
e people's people's
rconfounding
the
nrun
oallows
tprinciples
complex
ability
absence
f the
o r behavior
cirisk nbehavioral
investigators
t to
a t of
setting,
detect
or
ioof
Page
in initiatives
moderat-
develop-
pcomplex
without
3proc-
u rp othey
of
sedo
to
s7)
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

ing a "hot house" theory of health behavior that has lim- behavior is that by providing a context in which some or
ited practical value. all of the facets of a theory can be tested, interventionists
are in a position to generate evidence that will enhance
Interventions afford an invaluable opportunity to discern the accuracy and applicability of theory and thus, over
the context dependence of causal relations that have been time, improve the quality of the theories to which inter-
revealed in the laboratory. Some factors may be shown to ventionists can turn. By systematically testing principles
always be critical, whereas others may be critical only specified in health behavior theories, investigators are
under certain conditions. [25] For example, self-efficacy able to not only verify the accuracy of these predictions,
may be a critical determinant of the decision to initiate a but also develop a better understanding of their practical
new pattern of behavior, but have a limited impact on the value. Across studies, evidence should accumulate that
decision to maintain that behavior over time. [21] It is will allow investigators to differentiate between factors
critical to understand that restricting the conditions under that should and should not be targeted for intervention.
which a construct affects behavior does not mean that a Because current theories of health behavior often provide
given factor is not important. Information that would a list of factors that may affect behavior, the set of poten-
help delimit these conditions would enable theorists to tial mediating variables suggested by a theory may pose a
develop more precise models. daunting if not untenable measurement burden. How-
ever, the implementation of consistent and methodologi-
The case for why interventions should be more receptive to cally sound assessment of these factors should provide the
theory empirical evidence needed to constrain and prioritize the
There are two sets of reasons why we must take better variables on that list.
advantage of the opportunities interventions provide to
implement and test theories of health behavior. One set The characteristics of intervention strategies that prove to
focuses on what theory can do to improve the implemen- be effective should also provide investigators with a better
tation and evaluation of an intervention, whereas the understanding of the determinants of a given construct. As
other set focuses on how interventions can be used to was previously mentioned, theories may propose that a
improve the accuracy and quality of prevailing health construct (e.g., satisfaction) is a critical determinant of
behavior theories. First, by grounding their work on theo- decisions to maintain a new pattern of behavior, but pro-
retical principles regarding processes that regulate peo- vide limited guidance as to how to alter people's standing
ple's behavior, investigators can readily specify the critical on that construct (e.g., how to help feel satisfied with the
assumptions that underlie their intervention protocol. outcomes afforded by their new behavior). [21] An inter-
These formal statements of cause and effect relations not vention protocol that is shown to successfully heighten
only provide a clear justification for the proposed research people's satisfaction with process and outcomes associ-
activities (i.e., why an investigator believes a given inter- ated with weight loss not only has clear practical value,
vention strategy will be effective), but also increase the but also can shed light on the process by which people
likelihood that the proposed methodology will allow the determine whether they are satisfied with their experi-
investigator to detect whether and why the intervention ences. If theorists can develop a more detailed account of
had its intended effect. [10,11] the processes that shape the primary constructs identified
in a health behavior theory, interventionists will find that
When faced with unambiguous evidence of a successful theories can provide a more useful set of guidelines for
intervention effect, investigators might be able to move how to develop strategies to target these constructs.
forward without knowing why the intervention was effec-
tive. However, more often than not, investigators are faced Testing theoretical principles across a diverse array of set-
with the task of determining why an intervention failed to tings and populations will also enable investigators to bet-
produced the desired effect or why it worked under a lim- ter specify the scope of a theory. Although interventions
ited set of conditions. An a priori set of theoretical princi- provide a wonderful opportunity to test theoretical princi-
plesalso
canIn
framework
fective.
tigators
determine
but
decision
The
advantage provide
second
have
assessed
process.
whetherofan
particular,
offor
set important
opportunities
notdetermining
only
them
reasons
an conceptual
itintervention
identified
at
increases
the
why
to
why
appropriate
test
interventions
the
an
the
theories
will and
intervention
constructs
likelihood
prove
pointsanalytic
of should
health
effective,
that
that
in
was
the
may
inves-
take
inef- ples
strain
taken
to
range
atelyin diverse
systematic
these
comparisons.
tors
data
onpowered
together
of
have
types
sets
sampleofsamples
settings
comparisons
should
appropriately
have
tests
size
activities,
With and
orfacilitate
and
people.
of
tested
(pthe asettings,
a gmoderators
eresources.
information
can
it
assessed
nu should
The
mtheoreticalt formal
opportunities
be
b e rdrawn
increase
n oHowever,
can
be
the fthat
oeasier aaand
critical
put
across
rprinciple
ciin
istfor
public
gleanedappropri-
tconsiderable
to
ifstudies
io Page
this
constructs,
ndetermine
investiga-
pacross
utype
access
4from
rp oof
that
seof
s7
a)
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

which moderators are worth testing in a single, appropri- least consider assessing several constructs: the degree to
ately powered study design. which the intervention was implemented (e.g., did the
interventionists consistently provide participants with the
The identification of situational or personal factors that intervention exercises?), the degree to which participants
moderate the impact of a theoretical principal can be correctly identified the emphasis of the intervention (e.g.,
indicative of a number of different scenarios. For example, did participants assigned to the optimistic outcome con-
what might one conclude if an intervention that pro- dition report their was a greater emphasis on favorable
moted the health benefits of eating a balanced diet altered outcomes than did those assigned to the control condi-
the eating habits of college students but not those of high tion?), and finally the degree to which the intervention
school students? It could indicate that health benefits do altered the targeted set of opportunities, thoughts or feel-
not affect what high school students choose to eat. Alter- ings (e.g., did those assigned to the optimistic outcome
natively, it might be that high school students are respon-
condition develop more favorable expectations regarding
sive to perceptions of the health benefits afforded by a the benefits afforded by behavior change than did those
balanced diet, but that other factors (e.g., control over assigned to the control condition?).
access to food) preclude them from acting on those
beliefs. The practical and theoretical conclusions that can Although it is important that interventionists explicitly
be drawn from the identification of moderating factors are specify the constructs that determine the influence of the
dramatically increased if investigators can identify the intervention on participant behavior, the quality of the
causal processes that underlie the observed impact of the evidence that can be gathered depends on the assessment
moderator. In particular, can investigators discern procedures that are utilized. The persuasiveness of any
whether the moderated effect was obtained because the claims regarding the importance (or lack of importance)
moderator altered the ability of the intervention strategy of a particular construct is contingent on the use of meas-
to change the proposed mediating construct (e.g., the ures that have been shown to be reliable and valid. Given
intervention raised perceptions of the health benefits held that many of the constructs specified in theories of health
by college but not high school students) or because it behavior are conceptually similar, it is difficult to draw
altered the effect the mediator has on the primary out- strong conclusions regarding the specific contributions of
come measure (e.g., perceived health benefits predicted different variables in the absence of well-designed meas-
the eating habits of college but not high school students)? ures. [26,27] In addition, the inclusion of a pool of poten-
Greater attention to the causal processes invoked by a tial mediators enables the investigators to make stronger
moderator may also help investigators grapple with the claims as he or she can demonstrate that not only does the
daunting number of potential moderators. It is quite pos- construct specified in the model serve as a mediator but
sible that moderators that differ at the level of description that other factors do not operate as mediators.
(e.g., gender, ethnicity) can be accounted for by the same
underlying process. Adequately testing basic principles also depends on a
well-timed assessment schedule. Assessments are often
Finally, it is important to recognize that progress in theory too infrequent to detect meaningful changes on the con-
development can arise from the failure to obtain evidence struct. This is particularly true if the constructs of interest
in support of a specific prediction. Empirical evidence that are psychological states that both affect and are affected by
provides investigators with a better sense of the potential behavioral practices. However, specifying the optimal
factors that do not affect health practices will allow them time to assess the primary constructs can be difficult. To
to reduce the number of constructs (and, in time, theo- the extent that one wants to determine whether an inter-
ries) invoked to predict and explain health behavior. vention strategy (e.g., a tailored message about dietary
changes) alters the predicted mediating variable (e.g.,
What can be done to make interventions more theory- willingness to modify one's diet), one might consider
friendly? minimizing the length of time between the delivery of the
If one
tions
to
derived
is
tion(s)
that
whether
icated
wasthe assumes
enhance
can
manipulated.
more
appropriate
on
imbedded
from
be
itthe that
reveals
receptive
an
drawn
success
current
in there
intervention's
evaluation
To
from
thethe
to
this is
health
with interest
predicted
theory-testing,
the
intervention.
end,
which inindependent
intervention,
behavior
ability rendering
ofinvestigators
the
pattern
thecritical
toAny
theories?
assess
what interven-
ofregardless
conclusions
results,
manipula-
need
can
principles
One
be
tovariable
isdone
of
at
issue
pred- on
ever,intervention
cases,
detect
an
window
(e.g.,
chance and
investigator's
the
at
related
the
the
the
change
hypothesized
length
other.
of
of
same
to
time the
detecting
each
in
of
Of
time,
between
ability
diet)
these assessment
course,
other
(p amediator
ginterest
one
would
e two
to
and
nuthe
relation
there
madequately
btime
ethus
also
in
two
rand of
nthe
are the
owindows
benefit
efforts mediator.
tassessments.
may
the
association
fpractical
o rassess
outcome
cihinder
t to
afrom
tare
improve
ioPage 5How-
constraints
constructs.
n inversely
effects
pa
between
In
variable
ushorter
rpmany
oof
the
se to
s7)
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

What is needed is for investigators to take advantage of the tionists cooperate in the testing and evaluation of theoret-
measurement and testing opportunities when they do ical principles, individuals in both camps need to not only
arise. Although what can be concluded from any single recognize the goals and values of each group, but also
assessment effort may be limited, the cumulative impact trust each other's ability to advance our understanding of
of well designed tests of a theoretical principle can be sub- both theory and practice.
stantial. If investigators consistently wait for another time
or another investigator to conduct the relevant assess- Competing Interests
ments, innovations in theory and practice will continue to None declared.
be slow.
Acknowledgments
As interventionists specify the degree to which a given The preparation of this paper has been supported in part by National Insti-
study can test all or a facet of a given theory, they are more tute of Neurological Disorders and Stroke Grant 1R01-NS38441-01.

likely to articulate the contribution a proposed study


could make to the empirical literature. This process not References
1. Health people 2010: National health promotion and disease
only makes the justification for the intervention clear, but prevention U.S.
objectives.
Department of Health and Human Serv-
also improves the likelihood that investigators will recog- Washington,
ices D.C.: U.S. Government Printing Office; 2001.
nize when their and their colleagues' efforts have focused 2. Jeffery RW, Drewnowski A, Epstein LH, Stunkard AJ, Wilson GT,
Wing RR, Long-term
Hill DR: maintenance of weight loss: Cur-
consistently on a single or limited aspect of a given theory. rent status. Health Psychol 2000, 19: 5-16.
Research activities motivated by the Transtheoretical 3. Kumanyika SK, Van Horn L, Bowen D, Perri MG, Rolls BJ, Czajkowski
SM, SchronMaintenance
E: of dietary behavior change. Health
Model [28] provide an excellent example of a domain
Psychol2000, 19: 42-56.
where researchers have consistently relied on a limited 4. Marcus BH, Dubbert PM, Forsyth LH, McKenzie TL, Stone EJ, Dunn
number of methodological strategies and thus, despite an AL,Physical
Blair SN: activity behavior change: Issues in adop-
tion and maintenance. Health Psychol 32-41. 2000,
19:
enormous amount of research activity, provided a very 5. Ockene JK, Emmons KM, Mermelstein RJ, Perkins KA, Bonollo DS,
narrow test of the theory. [8] Vorhees CC, Hollis Relapse
JF: and maintenance issues for
smoking cessation. Health Psychol 2000,17-31. 19:
6. Salovey P, Rothman AJ, Rodin J: HealthHandbook
behavior. of
In
The commitment of time and effort to using interventions social psychology
. 4th edition.
Volume Edited
2 by: Gilbert D, Fiske S,
to test theoretical principles will in the end be for naught, Lindzey G. New York: McGraw-Hill; 1998:633-683.
7. Weinstein ND:Testing four competing theories of health-pro-
if there is not an equal commitment to the dissemination tective behavior. Health Psychol 1993,324-333. 12:
of the findings generated by these activities. In particular, 8. Weinstein ND, Rothman AJ, SuttonStage SR: theories of health
investigators who are engaged in the development of behavior. Health Psychol 1998, 17: 290-299.
9. Glanz K, Rimer BK,Health Lewis FM:behavior and health education:
health behavior theories must take advantage of the infor- Theory, research,3rd edition.
and practice
New York, N.Y.: Jossey-Bass;
mation afforded by intervention activities and demon- 2002.
10. Green LW, Kreuter Health MW:
promotion planning: An educational
strate that they are responsive to this information as they
and ecological 3rd approach
edition. Mountain View, CA: Mayfield;
refine and revise their theories. Enhanced communication 1999.
should also provide an opportunity for basic and applied 11. Bartholomew LK, Parcel GS, Kok G, Gottlieb Intervention
NH: map-
ping: Designing theory- and evidence-based health promotion programs
behavioral scientists to recognize the strengths and weak- Mountain View, CA: Mayfield; 2001.
ness of current theories of health behavior and thus help 12. Prentice DA, Miller DT: When small effects are impressive. Psy-
formulate a fuller understanding of what needs to be done chol Bull 1992, 112: 160-164.
13. Glasgow RE, Vogt TM, Boles Evaluating
SM: the public health
to improve the quality of our theories. impact of health promotion interventions: the RE-AIM
framework. Am J Public Health 1999,1322-1327. 89:
14. Baranowski T, Anderson , Carmack MediatingC: variable frame-
Summary work in physical activity interventions: How are we doing?
With an eye toward the future How might we do better? Am J Prev 266-297.
Med
1998,
15:
Although Lewin may have been right that there is "noth- 15. Baranowski T, Lin LS, Wetter DW, Resnicow K, Hearn Theory
MD:
as mediating variables: Why aren't community interventions
ing more practical than a good theory" (p.169; [24]), his working as desired? Ann Epidemiol 1997, S89-S95.
S7:
dictum rests on the assumption that good theories are 16. Hansen WB, McNeal RB The Jr: law of maximum expected
available
of
and
on-going
ioral
are
between
able to
collaboration
If
an critical
"good"
iterative address
applicable
scientists.
to
these
collaboration
advances
theories
foster
process
and,
two practical
Research
has
a in
stronger
groups
in
by
that
been
turn,
health
which
between
is,
and problems.
hindered
aare
sense
theories
opportunity
behavior
professional
theoreticians
likely
basic
ofby
to
that The development
interdependence
and
theory
aprovide
breakdown
for
are
applied
activities
and
innovation.
both
depend
ainterven-
base
accurate
behav-
that
inon
for
the potential effect: Constraints placed on program Page effective-
6behav-
of
K,
ior21.
17.
18.
19.
20.
Baumeister
change:
Ajzen
Mook
Rothman
Understanding
nessI,
11:
38:
Process
Englewood
maintenance.
ral I:
DG:
maintenance.
Disentangling
501-507.
379-387.
by
Fishbein
R.
AJ:
AJ,
Toward
(pmediator
1991,
New
aThe
Baldwin
g Cliffs,
e York,
M:
theory
attitudes
In
nua50:
mtheory-based
defense
relationships.
NJ:
behavioral
bAS,
NY:
Health
e179-211.
of
rPrentice
The
Hertel
In
nand
Guilford
planned
o of
Self-regulation
thandbook
Psychol
predicting
fexternal
Organ
oAW:
initiation
Hall;
ranalysis
Press;
behavior.
ci t aBehav
2000,
1980.
of
Health
tinvalidity.
Edited
social
ioself-regulation
2004:130-148.
of
nand
Am
and
Hum
behavioral
p 19:64-69.
uEduc
behavio-
behavior
by:
Psychol
rp Deciss7)
o1983,
1996,
Vohs
Res
se
International Journal of Behavioral Nutrition and Physical Activity 2004, :11
1 http://www.ijbnpa.org/content/1/1/11

22. Bandura A: Self-efficacy: The exercise of control New York, NY: W.H.
Freeman and Company; 1997.
23. Suls J, Rothman AJ: Evolution of the psychosocial model: Impli-
cations for the future of health psychology. Health Psychol 2004,
23: 119-125.
24. Lewin K: Field theory in social science: Selected theoretical
papers. New York, NY: Harper & Row; 1951.
25. McGuire WJ: A perspectivist approach to the strategic plan-
ning of programmatic scientific research. In Psychology of sci-
ence: Contributions to metascience Edited by: Gholson B, Shadish Jr WR,
Neimeyer RA, Houts AC. New York: Cambridge University Press;
1989:214-245.
26. Cummings KM, Jette AA, Rosenstock IM: Construct validation of
the health belief model. Health Educ Monogr 1978, 6:394-405.
27. Harrison JA, Mullen PD, Green LW: A meta-analysis of the health
belief model with adults. Health Educ Res 1992, 7: 107-116.
28. Prochaska JO, DiClemente CC, Norcross JC: In search of how
people change: Applications to addictive behaviors. Am Psychol
1992, 47: 1102-1114.

Publish with Bio Med Central every


and
scientist can read your work free of charge
disseminating
"BioMed
Suavailable
htt peer
bm
p:/ itwYour
/wCentral
reviewed
the
yocited
yours
w.bufree
Si
ior(p
rm research
results
will
aof
man
P in

gand
au
edcenecharge
uPubMed
be
you
of
published
l tral
sc
nu
N ri.cur
m spapers
the
biomedical
keep
pt
ombto
h/iemost
e, and
the
er
rCthe
nfone:/pan will
immediately
oentire
archived
significant
tcopyright
ubl be:
c efisresearch
orhirbiomedical
cie on
Rng_ se ioP
aatPubMed
tadv upon age
development
in
rc
.asphnour 7oMed
communityof
central
pacceptance
UK s7)
Bio
ulifetime."
Central
rp sefor