Strategies Of Health Promotion Unit 4 Article Review Applied Sciences homework help

See attachment

Digitally Characterizing the Dynamics of Multiple Health Behavior Change

Bonnie Spring1, Tammy K. Stump1, Samuel L. Battalio1, H. Gene McFadden1, Angela Fidler Pfammatter1,
Nabil Alshurafa1, and Donald Hedeker2

1 Department of Preventive Medicine, Northwestern University Feinberg School of Medicine
2 Department of Public Health Sciences, University of Chicago

Objective: We applied the ORBIT model to digitally define dynamic treatment pathways whereby
intervention improves multiple risk behaviors. We hypothesized that effective intervention improves the
frequency and consistency of targeted health behaviors and that both correlate with automaticity (habit)
and self-efficacy (self-regulation). Method: Study 1: Via location scale mixed modeling we compared
effects when hybrid mobile intervention did versus did not target each behavior in the Make Better
Choices 1 (MBC1) trial (n � 204). Participants had all of four risk behaviors: low moderate-vigorous
physical activity (MVPA) and fruit and vegetable consumption (FV), and high saturated fat (FAT) and
sedentary leisure screen time (SED). Models estimated the mean (location), between-subjects variance,
and within-subject variance (scale). Results: Treatment by time interactions showed that location
increased for MVPA and FV (Bs � 1.68, .61; ps � .001) and decreased for SED and FAT (Bs � �2.
01, �.07; ps � .05) more when treatments targeted the behavior. Within-subject variance modeling
revealed group by time interactions for scale (taus � �.19, �.75, �.17, �.11; ps � .001), indicating that
all behaviors grew more consistent when targeted. Method: Study 2: In the MBC2 trial (n � 212) we
examined correlations between location, scale, self-efficacy, and automaticity for the three targeted
behaviors. Results: For SED, higher scale (less consistency) but not location correlated with lower
self-efficacy (r � �.22, p � .014) and automaticity (r � �.23, p � .013). For FV and MVPA, higher
location, but not scale, correlated with higher self-efficacy (rs � .38, .34, ps � .001) and greater
automaticity (rs � .46, .42, ps � .001). Conclusions: Location scale mixed modeling suggests that both
habit and self-regulation changes probably accompany acquisition of complex diet and activity behaviors.

Keywords: health promotion, location-scale model, mobile health, multiple health behavior change,
self-regulation

Poor-quality diet and physical inactivity are the most prevalent
risk factors for chronic diseases, including diabetes, cardiovascular
disease, and cancers (Adams et al., 2017; Arena et al., 2015; Bauer
et al., 2014; Baruth et al., 2011; Lloyd-Jones et al., 2010; Mendis
et al., 2015; Mozaffarian et al., 2015; Myint et al., 2009; Schuit et
al., 2002; Spring et al., 2014; Spring et al., 2013). In turn, chronic
diseases are the main causes of premature death and disability and
the leading drivers of the United States’ $3.3 trillion annual health
care costs (Centers for Disease Control & Prevention, 2019). Just
1 in 10 U.S. adults consumes the recommended intake of fruits and
vegetables (Lee-Kwan et al., 2017), and fewer than one-third meet
dietary guidelines to consume less than 10% of calories from

saturated fats (American Heart Association, 2015; Sacks et al.,
2017; U.S. Department of Health and Human Services and U.S.
Department of Agriculture, 2015). Only about half meet public
health recommendations for moderate–vigorous physical activity
(MVPA; Centers for Disease Control & Prevention, 2018), and
more than 50% exceed two sedentary hours per day watching TV
(Fedewa et al., 2015). Moreover, risk behaviors co-occur: the
average adult reports at least two, and 25% report three or more
(Baruth et al., 2011; Chou, 2008; Meader et al., 2017; Schuit et al.,
2002).

Public health guidelines for diet and physical activity advise
consumers to accumulate a total amount of a food commodity
(e.g., servings/cups of fruits/vegetables) or type of activity (total
minutes MVPA) on a regular basis. They also specify how eating
and activity behaviors should optimally be distributed over time:
for example, whether target levels need to be met daily or weekly
to achieve a health benefit (Dunton, 2018; U.S. Department of
Health and Human Services, 2018; U.S. Department of Health and
Human Services and U.S. Department of Agriculture, 2010, 2015).
What neither current guidelines nor behavioral theories specify
well is how to design interventions so that they produce the
temporal pattern of changes that leads to a sustainable modifica-
tion of behavior. As others have noted, existing psychological
theory and analytic methods are more adept at characterizing

Bonnie Spring X https://orcid.org/0000-0003-0692-9868
This study was funded by National Heart, Lung, and Blood Institute

Grant HL075451 and by National Institute of Diabetes and Digestive and
Kidney Diseases Grant DK108678 to Bonnie Spring. The work was also
supported, in part, by National Cancer Institute Grant T32 CA193193 (PI:
Bonnie Spring, providing salary support for Tammy K. Stump).

Correspondence concerning this article should be addressed to Bonnie
Spring, Department of Preventive Medicine, Northwestern University
Feinberg School of Medicine, 680 North Lakeshore Drive, Suite 1400,
Chicago, IL 60611, United States. Email: bspring@northwestern.edu

T
hi

s
do

cu
m

en
t

is
co

py
ri

gh
te

d
by

th
e

A
m

er
ic

an
P

sy
ch

ol
og

ic
al

A
ss

oc
ia

ti
on

or
on

e
of

it
s

al
li

ed
pu

bl
is

he
rs

.
T

hi
s

ar
ti

cl
e

is
in

te
nd

ed
so

le
ly

fo
r

th
e

pe
rs

on
al

us
e

of
th

e
in

di
vi

du
al

us
er

an
d

is
no

t
to

be
di

ss
em

in
at

ed
br

oa
dl

y.

Health Psychology
© 2021 American Psychological Association
ISSN: 0278-6133 https://doi.org/10.1037/hea0001057

2021, Vol. 40, No. 12, 897–90*8

897

This article was published Online First February 11, 2021.

differences between people than fluctuations within a person over
time (Dunton, 2018; Dunton & Atienza, 2009; Riley et al., 2011).

This is an era when digital sensing affords continuous, real-time
detection of changes in a person’s behavioral state and surrounding
context. New technologies and analytic methods allow us to mea-
sure and conceptualize the dynamics of behavior change in ways
that can inform our theories and interventions (Riley et al., 2011).
Whereas traditional examination of intervention effects has fo-
cused on overall changes to the mean level or frequency of a
behavior, we can now leverage technological advances to identify
more complex behavioral patterns. The problem being addressed is
how to identify behavioral features that are evident before the end
of treatment and that can successfully predict whether a healthy
behavioral change will be maintained after treatment is withdrawn.
Later, we elaborate how we plan to draw predictors of maintenance
from a broad array of behavioral features (e.g., changes in day-to-
day variability, rate of change in variability, responses to “lapses”
in behavior). The present study takes a first step toward that goal
by characterizing patterns of change in daily variability (scale) and
level (location) of targeted behaviors within a successful interven-
tion.

This article’s premise is that breakthrough advances for com-
plex health behavior change interventions will be facilitated by
introducing more dynamic concepts and tools into two domains of
the science of behavior change. The first needed development is an
expansion of the manner in which traditional statistical methods
characterize within-person change. The second is integration into
psychological theory of dynamic constructs that explain how an
intervention induces a pattern of transitions between psychological
and behavioral states that leads to the acquisition of durable
behavioral change. When stated dynamically in algorithmic form,
adaptive treatment decision rules (Murphy, 2005) can specify how
change in the patient’s response pattern and change in the inter-
vention should come to reciprocally determine each other over
time (Bandura, 1985).

Intervention Development Model—Orbit Phase 1

Step 1 in the ORBIT treatment development model (Czajkowski
et al., 2015) is to identify a clinically meaningful question. Our
question is how to maximize healthy change in multiple diet and
physical activity risk behaviors for the least possible resource
expenditure. Effective interventions for diet and PA risk behaviors
exist but are intensive: involving multiple treatment sessions, each
lasting from 10 –90 min (Curry et al., 2014, 2018). Patients
(Becker et al., 2017; Jensen et al., 2012) and payers (Arterburn et
al., 2008; Jones et al., 2015) consistently name high cost and
burden (long treatment duration, high time commitment) as top
barriers to uptake of intensive behavioral treatments. Yet, even
after intensive lifestyle intervention, behavioral improvements of-
ten are not maintained (Arterburn et al., 2008; Perri, 1998),
prompting recourse to the most common maintenance strategy:
continuing to offer behavioral treatment (Perri, 1998), which fur-
ther augments cost and participant burden.

Accordingly, a critical goal for the science of health behavior
change is to detect when, during treatment, an intervention has
produced durable behavioral improvement that might indicate
when intervention delivery can begin to be tapered and then
discontinued. Likewise, an ability to detect during the posttreat-

ment maintenance phase when a previously stable healthy behav-
ior pattern is beginning to “wobble,” could indicate, just in time,
when intervention should be reinstated. Such knowledge would
allow treatment dosing decisions to be made by assessing dynamic
processes that underlie the acquisition and maintenance of durable
healthy eating and PA patterns.

Our specific intervention development goal is to optimize the
existing Make Better Choices (MBC) intervention for multiple diet
and physical activity change (Spring et al., 2012, 2018) to achieve
and maintain the maximal healthy behavior improvement that is
attainable for the least resource utilization (cf., Spring et al., 2020).
All of the work shown here represents activity undertaken in either
Orbit Phase 1a (Define: during which the scientific foundation of
the behavioral treatment is defined), or ORBIT Phase 1b (Refine:
when candidate targets for treatment components are specified and
the hypothesized pathway by which treatment produces benefit is
formulated; Czajkowski et al., 2015). Our Phase 1 treatment de-
velopment process involves performing secondary analysis of two
prior MBC clinical trials (Spring et al., 2012, 2018). In ORBIT
Phase 1a, we use these data to define what it means to achieve
healthful behavior change. We apply location-scale modeling to
identify two behavioral features (level and temporal consistency)
that change as different interventions produce acquisition of
guidelines-concordant diet and activity behaviors. These second-
ary analyses were prompted by the premise that the acquisition of
healthy behavior change is more likely to be durable when inter-
vention causes the behavior to both reach its targeted level and be
enacted consistently. In Phase 1b, we sought to refine our under-
standing of this behavior change patterning by learning how what
is measured as changes in behavioral level and consistency fits into
the nomological network of psychological constructs (Cronbach &
Meehl, 1955) that are thought to explain the successful acquisition
of sustained healthy behavior change. Two constructs in this
network pertain to the acquisition of self-regulation skills and the
acquisition of habits. For behavioral consistency to be useful as a
new construct, it should converge with these related constructs but
not overlap entirely.

Orbit Phase 1A: Acquisition of Behavioral Consistency

In this phase, we perform secondary analysis of prior clinical
trial data to define what it means to achieve multiple health
behavior change. We proposed in the introductory section of this
article that boosting the growth of dynamic behavior change in-
terventions requires expanding the manner in which traditional
statistical methods measure within-person change. Most treatment
evaluations have inferred healthy habit acquisition from changes in
the frequency or rate of a behavior: increases for healthy behaviors
and decreases for unhealthy ones. Many statistical analysis tech-
niques for intensive longitudinal data are well-suited to model this
type of improvement in behavioral rate. In contrast, another aspect
of intraindividual change (improvement in behavioral consistency)
has not been measured routinely. Yet, for many medical and public
health guidelines, maintaining a stable homeostatic range of a
biomarker or behavior is at least as important as attaining a mean
target level (American Diabetes Association, 2020; Dunton, 2018;
Riley et al., 2011). Consider, for example, two individuals with
diabetes and acute coronary syndrome who have the same average
blood glucose levels across the month, one whose glucose is within

T
hi

s
do

cu
m

en
t

is
co

py
ri

gh
te

d
by

th
e

A
m

er
ic

an
P

sy
ch

ol
og

ic
al

A
ss

oc
ia

ti
on

or
on

e
of

it
s

al
li

ed
pu

bl
is

he
rs

.
T

hi
s

ar
ti

cl
e

is
in

te
nd

ed
so

le
ly

fo
r

th
e

pe
rs

on
al

us
e

of
th

e
in

di
vi

du
al

us
er

an
d

is
no

t
to

be
di

ss
em

in
at

ed
br

oa
dl

y.

898 SPRING ET AL.

a healthy range on most days and the other whose glucose fluctuates
between healthy, hyperglycemic, and hypoglycemic. Fewer adverse
cardiac events are expected for the patient who maintains a stable
glucose level than the one whose glucose shows great within-
person variability (Gerbaud et al., 2019). Thus, we sought an
analytic technique that could measure within-person variability to
study how it is impacted by behavioral intervention. Such an
analytic technique could, for instance, let us differentiate interven-
tions and individuals that not only improve fruit and vegetable
consumption to a targeted level on average but also improve
consumption to a similar, high level of fruits and vegetables on
most days.

Mixed-effects location scale modeling (a statistical analysis
technique for intensive longitudinal data; Hedeker et al., 2008,
2012) appeared to offer the needed capabilities. Location scale
modeling quantifies change in within-person variability of behav-
ior (scale), unlike most analytic techniques which only assess
changes in absolute level (location). The availability of this
method let us evaluate the hypothesis that effective interventions
produce both improvement in location (increases for healthy be-
haviors; decreases for unhealthy ones) and decrease in scale (i.e.,
increased behavioral consistency) of targeted behaviors. We stud-
ied this by applying location scale modeling to data from a previ-
ously published trial of behavioral interventions to improve mul-
tiple diet and PA behaviors (Spring et al., 2010, 2012).

The Make Better Choices 1 Trial

The MBC1 study enrolled 204 adults with unhealthy levels of
sedentary leisure screen time (SED), fruit/vegetable intake (FV),
saturated fat intake (FAT), and moderate vigorous physical activity
(MVPA). All participants had all four risk behaviors and were
randomized to one of four hybrid interventions involving an app
plus remote coaching: (a) increase FV and MVPA, (b) decrease
FAT and SED, (c) decrease FAT and increase MVPA, (d) increase
FV and decrease SED (Spring et al., 2010, 2012). The aim was to
determine which combination of one diet and one PA intervention
target produced the maximum improvement in all four risk behav-
iors. Based on behavioral choice theory (Bickel & Vuchinich,
2000), we hypothesized that the increase FV and decrease SED
intervention condition would yield maximum improvement in a
composite measure combining all four behaviors because of sub-
stitute and complementary relationships between FV, SED, and the
other behaviors. Specifically, we predicted and found that in
addition to directly increasing FV, this intervention indirectly
decreased FAT: FV partially substituted for (crowded out) FAT,
probably as a result of increased satiety owing to heightened fiber
intake. We also observed a complementary relationship between
SED and FAT, such that decreasing leisure screen time was
accompanied by decreased FAT, at least partially because reducing
TV viewing also decreased the hand-to-mouth snacking with
which participants paired it. By the end of treatment, the group that
was asked to increase FV and decrease SED improved on a
standardized composite score reflecting all behaviors more than
the groups in other treatment conditions that were asked to change
different pairs of diet and PA behaviors (p � .001), and the
difference was maintained through the 6-month follow-up period
(Spring et al., 2012).

Method

All participants were asked to wear an accelerometer and use a
custom-designed app to self-monitor dietary intake, MVPA, and
SED daily during a 2-week Baseline phase, early phase treatment
[Rx1: 1 week] when goals were set to 50% of final level, and later
phase treatment [Rx2: 2 weeks] when goals were set to 100%. The
intervention for all participants included three weeks of telephone
coaching as well as a mobile application designed based on Con-
trol Systems Theory (Carver & Scheier, 1982) to help participants
set diet and PA goals, self-monitor their behavior and receive
feedback about progress, and earn financial incentives for self-
monitoring and goal attainment. In the 6-month follow-up phase,
participants were only incentivized to self-monitor on the follow-
ing schedule: daily for one week at week 4, 3 consecutive days for
weeks 5 and 6, biweekly for 6 weeks, and monthly until 6-month
follow-up. Goal attainment was no longer incentivized. Study
procedures were approved by the Institutional Review Boards at
the University of Illinois at Chicago and Northwestern University.

Analyses were run separately for each of the four behaviors,
with participants divided into one of two groups (Targeted or Not
Targeted) with regard to whether their assigned intervention tar-
geted that behavior or targeted other behaviors. For instance, for
analysis of FV, the Targeted group included those in the increase
FV and increase MVPA intervention condition as well as those in
the increase FV and decrease SED condition. For FAT, the Tar-
geted group included those in the decrease FAT and SED inter-
vention condition, and those in the decrease FAT, increase MVPA
condition. For MVPA, the Targeted group included those in the
increase FV and MVPA intervention condition, and those in the
decrease FAT increase MVPA condition. For SED, the Targeted
group included those in the decrease FAT and SED intervention
condition, and those in the increase FV and decrease SED condi-
tion. Here we tested the hypothesis that each behavior would show
an interaction between treatment group and time, such that im-
provement in the behavior’s rate and consistency would be greater
for the group whose intervention targeted that behavior than for
those for whom the behavior was not targeted, demonstrating
treatment differentiation for the enactment of specific diet and PA
behavioral improvements.

Each of the four behavioral outcomes was analyzed using a
mixed-effects location scale model, as implemented in the
MIXREGLS software program (Hedeker & Nordgren, 2013). To
better satisfy the normality assumption of the model, a square root
transformation was used for the count outcomes: fruit and vege-
table consumption (FV; servings), MVPA (minutes), sedentary
leisure behavior (SED; minutes), and arc sin transformation was
used for the percentage outcome (FAT; % daily calories attribut-
able to saturated fat). The mean, BS variance, and WS variance
models included terms for group (0/1), time (0 � baseline, 1 �
Rx1, 2 � Rx2) and group by time interaction. The mean model
corresponds to a regression model, whereas the variance models
(BS and WS) are log-linear regression models so that the resulting
variances are always positive (Hedeker et al., 2008). Group was
dummy-coded separately for the four outcomes, such that group �
1 for participants in the conditions that targeted that outcome. For
example, in the analysis of MVPA, group � 1 for those in the
conditions that were targeted to increase MVPA, and 0 otherwise.
For SED, group � 1 for the conditions that were targeted to

T
hi

s
do

cu
m

en
t

is
co

py
ri

gh
te

d
by

th
e

A
m

er
ic

an
P

sy
ch

ol
og

ic
al

A
ss

oc
ia

ti
on

or
on

e
of

it
s

al
li

ed
pu

bl
is

he
rs

.
T

hi
s

ar
ti

cl
e

is
in

te
nd

ed
so

le
ly

fo
r

th
e

pe
rs

on
al

us
e

of
th

e
in

di
vi

du
al

us
er

an
d

is
no

t
to

be
di

ss
em

in
at

ed
br

oa
dl

y.

899DYNAMICS OF DIGITAL MULTIPLE BEHAVIOR CHANGE

decrease sedentary behavior, and 0 otherwise. For FV, group � 1
for the conditions that were targeted to increase FV, and 0 other-
wise. For FAT, group � 1 for the conditions that were targeted to
decrease saturated fat consumption, and zero otherwise. With this
coding of the outcomes, the main effect of group represents the
group difference at baseline, the time effect represents the time
trend for the group � 0 (i.e., not targeted) conditions, and the
group by time interaction represents the difference in the time
trend for the group � 1 (i.e., targeted) conditions, relative to
group � 0 condition. Thus, the group by time interaction is of
greatest interest here, as it indicates the effect of targeting behav-
iors on the outcomes across the time intervals.

Results

For MVPA (see Figure 1), there were significant effects of time
(� � 0.25, p � .008) and the group by time interaction (� � 1.68,
p � .001) in the mean model. This indicates that all subjects
increased MVPA levels across the three time intervals (baseline,
Rx1, Rx2), but that the conditions that were designed to target
increasing MVPA did so to a much greater degree (cf. Figure 1).
In terms of the BS variance modeling, there was a significant effect
of time (� � .22, p � .001), indicating that subjects became more
heterogeneous across the three time intervals. For the WS variance
modeling, there was only a significant group by time interaction
(� � �0.19, p � .001). This indicates that the consistency of
MVPA within subjects did not change across days for the condi-
tions in which MVPA was not targeted, but MVPA became more
consistent for the conditions in which it was targeted (i.e., the WS
variance was reduced across the time intervals for the targeted
MVPA conditions). Exponentiating this estimate yields a variance
ratio effect of exp (�0.19) � 0.83. This indicates that the WS
variance was reduced by 17% (100% � 83% � 17%), such that
the level of PA became more consistent with each successive time
interval after baseline for the targeted MVPA group, relative to the
group for which MVPA was not targeted.

Similar findings were also observed for the other three behav-
iors, such that the mean level of targeted healthy behaviors (FV)
increased over time and the mean level of targeted unhealthy
behaviors (SED and FAT) decreased over time. In the conditions
in which they were targeted, behaviors also became more consis-
tent over time. Table 1 presents results for the location scale mixed
models run separately for each behavioral outcome, in terms of the

effects on the location (mean) and scale (WS variance). As pre-
dicted, these results revealed increases in mean levels of targeted
healthy behaviors (PA and FV), decreases in mean levels of
targeted unhealthy behaviors (SED and FAT; i.e., improved loca-
tion), and decreases in WS variance (i.e., increased consistency,
improved scale) for all targeted behaviors.

Discussion

These findings provide preliminary support for the use of loca-
tion scale mixed modeling to evaluate improved behavioral fre-
quency and increased consistency during the acquisition of healthy
diet and activity changes induced by an effective intervention. It is
noteworthy that both location and scale improved for each of the
quite different health behaviors that the interventions targeted.
Location improvements characterized both increasing behavioral
levels for low-rate healthy behaviors (FV, MVPA) and decreasing
behavioral levels for high-rate unhealthy ones (FAT, SED). Re-
gardless of the directionality of the targeted behavior change,
interventions that effectively improved the behavior’s location also
lowered its scale, diminishing within-person variability across
time. Stated differently, effective interventions that achieved
guidelines-recommended levels of dietary intake or PA also
achieved guidelines-recommended consistency of the behavior
across occasions (i.e., in this case, days). It might be asked why we
consider consistent nonaction or minimal action for an unhealthy
behavior to be a good thing, once effective intervention has re-
duced behavioral rate (location) to low or zero. It is because, just
as single or rare slips to an unhealthy behavior increase the odds of
a full relapse, continued, consistent lapse-free intervals build self-
efficacy that is protective against relapse (Kirchner et al., 2012;
Larimer et al., 1999). In sum, the introduction of location-scale
modeling illustrates the first scientific development that we found
needed to support more dynamic behavioral interventions: im-
proved modeling of within-person change.

ORBIT Phase 1B: How Is Behavioral Consistency
Related to Psychological Constructs Thought to

Underlie Healthful Behavior Change?

The second scientific development that we found needed is the
integration into psychological theory of dynamic constructs that
explain how a pattern of transitions between psychological and

Figure 1
Changes in mean and WS variance for square root transformed minutes of
MVPA over time between groups whose intervention targeted vs. did not target
increasing MVPA

T
hi

s
do

cu
m

en
t

is
co

py
ri

gh
te

d
by

th
e

A
m

er
ic

an
P

sy
ch

ol
og

ic
al

A
ss

oc
ia

ti
on

or
on

e
of

it
s

al
li

ed
pu

bl
is

he
rs

.
T

hi
s

ar
ti

cl
e

is
in

te
nd

ed
so

le
ly

fo
r

th
e

pe
rs

on
al

us
e

of
th

e
in

di
vi

du
al

us
er

an
d

is
no

t
to

be
di

ss
em

in
at

ed
br

oa
dl

y.

900 SPRING ET AL.

behavioral states could lead to the acquisition of behavioral
changes (particularly those that might be maintained). In this
instance, we reanalyze prior clinical trial data to learn whether
changes to the absolute level and consistency of various diet and
PA behaviors co-occur with changes to psychological variables
previously identified as important accompaniments of behavior
change. When behavioral consistency emerges, does self-
regulation improve? Do habits form and strengthen? Both? In
other words, we refine understanding of the MBC intervention’s
treatment pathway by identifying which psychological variables
are associated with improved behavioral level and consistency. In
doing so, we sought to provide evidence for behavioral consiste-
ncy’s recognition as a distinct construct and to explore its place-
ment within the nomological network of related constructs (Cron-
bach & Meehl, 1955).

Two main classes of psychological theory are especially rele-
vant to achieving changes to healthy behavior that become con-
sistently enacted at the daily level: self-regulation theory (Bandura,
1991; Baumeister et al., 2007; Kanfer & Goldstein, 1991) and
habit theory (Aarts & Dijksterhuis, 2000; Gardner, 2015). Self-
regulation is the process of volitionally exerting control over the
self, by inhibiting competing responses, to change or sustain a
pattern of thought, feeling, or behavior (Baumeister et al., 1994).
An intervention to improve self-regulation might teach skills to
execute generalizable behavior change strategies or techniques that
have utility across many different risk behaviors and contexts.
Such behavior change techniques (BCTs) have been characterized
in comprehensive taxonomies put forward by Michie and col-
leagues (Abraham & Michie, 2008; Michie et al., 2011, 2013). Our
recent metareview of meta-analyses examining interventions to
foster diet and physical activity changes for health or weight loss
found that goal-setting and self-monitoring were by far the most
commonly evaluated BCTs (Spring, 2019). None of the specific 14
BCTs analyzed was consistently related to diet and activity
improvements, however (Spring, 2019). On the other hand, in
multiple studies, increased self-efficacy has been a consistently
observed consequence of effective interventions that train self-
regulatory skills. Self-efficacy often mediates the beneficial effects
of self-regulatory interventions on diet and PA behavioral out-
comes (Schneider et al., 2016; Darker et al., 2010). Hence, we

consider greater self-efficacy to be an indicator of better self-
regulation.

Self-regulation is a goal-driven and effortful undertaking, particu-
larly when it involves pursuit of multiple different behavior change
goals that require complex new action patterns. Given the consider-
able burden of executing several System 2 deliberative, conscious,
goal-directed cognitive processes at once (Hagger

Looking for this or a Similar Assignment? Click below to Place your Order