Habits Predict Physical Activity on Days When Intentions Are Weak

Abstract

Physical activity is regulated by controlled processes, such as intentions, and automatic processes, such as habits. Intentions relate to physical activity more strongly for people with weak habits than for people with strong habits, but people's intentions vary day by day. Physical activity may be regulated by habits unless daily physical activity intentions are strong. University students (N = 128) self-reported their physical activity habit strength and subsequently self-reported daily physical activity intentions and wore an accelerometer for 14 days. On days when people had intentions that were weaker than typical for them, habit strength was positively related to physical activity, but on days when people had typical or stronger intentions than was typical for them, habit strength was unrelated to daily physical activity. Efforts to promote physical activity may need to account for habits and the dynamics of intentions.

Full text

157
Journal of Sport & Exercise Psychology, 2014, 36, 157-165
http://dx.doi.org/10.1123/jsep.2013-0173
© 2014 Human Kinetics, Inc.
Amanda L. Rebar is now with the School of Human, Health,
and Social Sciences, Central Queensland University, Rock-
hampton, QLD, Australia. Steriani Elavsky and Jaclyn P. Maher
are with the Department of Kinesiology, Pennsylvania State
University, University Park, PA. Shawna E. Doerksen is with
the Department of Recreation, Park, and Tourism Management,
Pennsylvania State University, University Park, PA. David E.
Conroy is with the Departments of Kinesiology and Human
Development & Family Studies, Pennsylvania State University,
University Park, PA.
Habits Predict Physical Activity
on Days When Intentions Are Weak
Amanda L. Rebar,1,2 Steriani Elavsky,2 Jaclyn P. Maher,2
Shawna E. Doerksen,2 and David E. Conroy2
1Central Queensland University; 2Pennsylvania State University
Physical activity is regulated by controlled processes, such as intentions, and automatic processes, such as
habits. Intentions relate to physical activity more strongly for people with weak habits than for people with
strong habits, but people’s intentions vary day by day. Physical activity may be regulated by habits unless
daily physical activity intentions are strong. University students (N = 128) self-reported their physical activ-
ity habit strength and subsequently self-reported daily physical activity intentions and wore an accelerometer
for 14 days. On days when people had intentions that were weaker than typical for them, habit strength was
positively related to physical activity, but on days when people had typical or stronger intentions than was
typical for them, habit strength was unrelated to daily physical activity. Efforts to promote physical activity
may need to account for habits and the dynamics of intentions.
Keywords: motivation, automatic regulation, dual-process model of physical activity
Physical activity lowers the risk for developing
coronary heart disease, high blood pressure, bone dis-
ease, stroke, type 2 diabetes, some forms of cancer, and
depression, resulting in a 30% lower chance of early
mortality (Physical Activity Guidelines Advisory Com-
mittee, 2008). Unfortunately, the majority of Americans
are not sufciently active enough to obtain signicant
health benets. In a recent study of objectively measured
physical activity, it was found that less than 10% of
Americans were participating in the recommended 150
min of moderate or 75 min of vigorous weekly physical
activity (Tucker, Welk, & Beyler, 2011). One way to
increase the number of Americans who engage in regu-
lar physical activity is to enhance people’s intentions to
engage in physical activity; however intention-focused
interventions have had limited success in changing
behavior (Rhodes & Dickau, 2012; Webb & Sheeran,
2006). Recent proposals have called for supplementing
intention-based theories of health behavior motivation
with automatic regulatory constructs, such as habits
(Dimmock & Banting, 2009; Marteau, Hollands, &
Fletcher, 2012; Sheeran, Gollwitzer, & Bargh, 2012). At
this point, however, little is known about how habits and
intentions interact at a within-person level to regulate
physical activity.
Dual-Process Model of Daily Physical
Activity Regulation
Maintaining a physically active lifestyle involves more
than a single decision or action; it requires repeated
successful regulation of physical activity. According to
dual-process models of behavior regulation (Chaiken &
Trope, 1999; Evans, 2008, 2009; Hofmann, Schmeichel,
& Baddeley, 2012), successful regulation of physical
activity can be the result of rational, goal-directed con-
trolled efforts (e.g., intentions), immediate, effortless
automatic processes (e.g., habits), or a combination of
both. A physically active lifestyle involves a combination
of these regulatory processes on a daily basis.
Physical Activity Habit Strength. It is likely that one
type of automatic process that regulates physical activ-
ity is a person’s habits—the automatic tendencies to
participate in physical activity in response to certain
cues or triggers such as settings, preceding events or
actions (Aarts, Paulussen, & Schaalma, 1997; Verplan-
ken, 2010; Verplanken & Orbell, 2003; Wood & Neal,
2007). Humans follow predictable patterns in daily
life, routinely being in the same places during the same
times (González, Hidalgo, & Barabási, 2008; Song, Qu,
Blumm, & Barabási, 2010). For some people, these daily
JOURNAL OF
SPORT EXERCISE
PSYCHOLOGY
Official Journal of NASPSPA
www.JSEP-Journal.com
ORIGINAL RESEARCH

158 Rebar et al.
life routines include regular physical activity participa-
tion. People with strong physical activity habits have
made the decision to be physically active so often in
the past that they now routinely participate in physical
activity (Gardner, de Bruijn, & Lally, 2011), and this
successful behavior regulation no longer requires delib-
eration about the benets of physical activity or reection
about attitudes or intentions (Ouellette & Wood, 1998).
In a meta-analysis, Gardner and colleagues (2011) found
a medium-strong relation between physical activity and
habit strength.
Physical Activity Intentions. One controlled process
that likely regulates daily physical activity is the inten-
tion to participate in physical activity (Ajzen, 1991).
Typically, intentions have been considered stable indi-
vidual differences, and the majority of research focuses
on between-person comparisons, showing that people
with strong intentions to participate in physical activ-
ity tend to be more physically active than those with
weaker intentions (Downs & Hausenblas, 2005; Hagger,
Chatzisarantis, & Biddle, 2002; McEachan, Conner,
Taylor, & Lawton, 2011). In a recent review, McEachan
and colleagues (2011) found a medium-to-strong relation
(ρ = 0.48) between intentions and prospective measures
of physical activity. Intention strength can differ at a
between-person level, but people’s intentions can also
uctuate at a within-person level depending on daily
social and physical constraints. Recent research has
revealed that intentions can vary considerably over
time, uctuating on at least daily, weekly, biweekly,
and monthly time scales and that these uctuations can
predict physical activity (Conroy, Doerksen, Elavsky, &
Maher, 2013; Conroy, Elavsky, Hyde, & Doerksen, 2011;
Scholz, Keller, & Perren, 2009; Scholz, Nagy, Schüz, &
Ziegelmann, 2008). It remains unclear, however, how
these daily uctuations in intentions interact with habits
to regulate daily physical activity.
Habits and Intentions Regulating Daily
Physical Activity
Habits will regulate daily behavior unless a person nds
himself in an unfamiliar context or other regulatory pro-
cesses are strong enough to interfere with this automatic
regulatory process (Neal, Wood, Wu, & Kurlander, 2011;
Ouellette & Wood, 1998). Based on the consistency of
human mobility patterns (González et al., 2008; Neal,
Wood, & Quinn, 2006; Song et al., 2010), it may be
assumed that people’s contexts are relatively stable; there-
fore, the current study focuses on intention strength as a
means of overriding physical activity habitual behavior.
Typically, people’s physical activity intentions cor-
respond with their physical activity habits. Indeed, strong
physical activity habits may be developed and maintained
as a result of consistently strong intentions (Neal, Wood,
Labrecque, & Lally, 2012; Wood & Neal, 2007). At a
between-person level, therefore, people’s intentions
should be positively associated with their habit strength.
Physical activity intentions can also differ from physi-
cal activity habits and create a conict in the regulation
of physical activity. For example, a person with weak
physical activity habits may have strong physical activity
intentions on some days. Particularly strong controlled
regulatory processes can override automatic regulatory
processes (Hofmann, Friese, & Strack, 2009; Hofmann
et al., 2012); therefore, strong daily intentions may inter-
fere with the habitual regulation of physical activity. In
contrast, weak daily intentions likely will not interfere
with habitual regulation of physical activity.
Previous research supports the theory that strong
intentions can override habits. For example, it has been
shown that behaviors that are repeated infrequently and in
a wide variety of contexts (and therefore less likely to be
habitual) are more intentionally regulated than behaviors
that are repeated more frequently and consistently in the
same context (Ouellette & Wood, 1998; Webb & Sheeran,
2006). In addition, it has consistently been shown that
behavior is regulated more by intentions for people with
weak habits than for people with strong habits (de Bruijn,
Kremers, Singh, van den Putte, & van Mechelen, 2009;
Ji & Wood, 2007; Knussen, Yule, MacKenzie, & Wells,
2004; Verplanken, Aarts, Knippenberg, & Moonen,
1998).
Gardner and colleagues (2011) reviewed studies
that tested habit-intention moderation effects within the
context of physical activity. They concluded that inten-
tions were more strongly related to physical activity for
people with weaker habits, but found some contradictory
evidence as well. Some evidence suggests habits and
intentions have an additive, but not interactive, effect on
physical activity (Rhodes, de Bruijn, & Matheson, 2010).
Beyond those covered in the review, studies have shown
that the relation between exercise intentions and behav-
ior is around three times stronger for those with weaker
habits than for those with stronger habits (de Bruijn &
Rhodes, 2011) and that habits and intentions both directly
predicted moderate and vigorous physical activity and
had interactive effects on these behaviors (Rhodes & de
Bruijn, 2010). These studies demonstrate that intentional
and habitual regulation differs between types of behaviors
and between people, but it remains unclear how they
interact on a within-person level. Previous investigations
have shown that patterns of physical activity motivation
present at the between-person level do not necessarily
translate to the same processes at a within-person level
(e.g., Conroy et al., 2011). By investigating how these
regulatory processes interact at a within-person level,
more can be understood about what underlies change in
physical activity behavior.
The Present Study
The present study used a within-person design to inves-
tigate if the link between a person’s habits and physical
activity varied as a function of between-person differ-
ences or within-person uctuations in physical activity

Physical Activity Habit Strength 159
intention strength. We tested the relation of objectively
measured physical activity to physical activity habit
strength, a person’s average daily physical activity
intentions across 2 weeks, and day-to-day uctuations
in intentions. We also tested if the relation between
physical activity habit strength and daily physical activity
was moderated by average daily intentions or daily uc-
tuations in intentions. Physical activity differs between
men and women (Caspersen, Pereira, & Curran, 2000;
McArthur & Raedeke, 2009), between weekdays and
weekends (Conroy et al., 2011; Matthews, Ainsworth,
Thompson, & Bassett, 2002; Trost, Pate, Freedson, Sallis,
& Taylor, 2000), and across study duration in response to
self-monitoring (Motl, McAuley, Dlugonski, 2012), so we
controlled for these covariates in the model. We hypoth-
esized that habit strength and average daily intentions
would positively relate to physical activity, and that both
average daily physical activity intentions and daily uc-
tuations in physical activity intentions would moderate
the relation between habit strength and physical activity.
Specically, it was expected that physical activity would
be more strongly regulated by habit strength for people
with weaker average intentions then for people with
stronger average intentions and on days when intentions
for physical activity were weaker than typical.
Method
Participants and Procedures
University students (N = 128, 76 women, Mage = 21 years)
participated in the study for course credit. Participants
were mostly White (95%), non-Hispanic (98%), and in
their senior year at the university (80%). The study proto-
col was approved by the local institutional review board.
All participants gave informed consent and permission
for their data to be used for research purposes.
During an initial laboratory visit, participants self-
reported their physical activity habit strength and were
provided with an ActiGraph accelerometer (ActiGraph,
Pensacola, FL). Participants were instructed to wear the
accelerometers on their right hip for all waking hours,
except when coming in contact with water (e.g., shower-
ing, swimming). For the next 13 days, participants wore
the accelerometers and logged in to a password-secured
website between the hours of 7:00 pm and 4:00 am each
night and self-reported daily intentions for physical
activity (for the next day). After the 13 days, participants
returned the accelerometers to the laboratory.
Measures
Physical Activity Habit Strength. The previously
validated automaticity four-item subscale of the Self-
Report Habit Index was used to measure physical
activity habit strength during the initial laboratory
visit (Gardner, Abraham, Lally, & de Bruijn, 2012;
Verplanken & Melkevik, 2008; Verplanken & Orbell,
2003). Participants reported how much they agreed with
the statements “Being physically active is something I
do automatically, I do without thinking, I do without
having to consciously remember, and I start doing before
I realize I’m doing it” on scales ranging from 1 (disagree
completely) to 7 (agree completely). The internal
consistency for the scale was .89. Previous research has
shown that this subscale is reliable, relates to prospective
behavior, and moderates between-person intention-
behavior relations as theorized (Gardner et al., 2012).
Physical Activity Intentions. Intentions for physical
activity were assessed using two custom items that
participants completed each evening in reference to the
next day. Participants reported how much they “intended
to engage in at least 30 min of moderate aerobic activity
tomorrow,” and “intended to engage in at least 15 min of
vigorous physical activity tomorrow” using slider scales
ranging from 0 (do not intend at all) to 100 (strongly
intend). Without accounting for within-person nesting,
the items were correlated .74. On average, the daily
items were correlated .73 within person across the 13
days. Intentions were calculated as the average of the
two scores.
Daily Physical Activity. Daily physical activity was
objectively assessed as average hourly activity counts
measured by the accelerometers. Days when the
accelerometer was worn less than 10 hr were treated
as missing. Based on conventional scoring procedures
(Choi, Liu, Matthews, & Buchowski, 2011), periods of
monitoring of 90 or more minutes of consecutive zero
activity counts (i.e., accelerometer was not worn) were
replaced with the average activity count of that individual
for that day.
Data Analyses
Physical activity intentions were separated into between-
and within-person components (Schwartz & Stone,
1998). The between-person level intentions variable was
calculated as the person’s average intentions across the
13 days. Daily intentions were calculated as the person’s
daily deviations from their average daily intentions. To
test the study hypotheses, a hierarchical linear model was
estimated. Physical activity was regressed onto average
daily intentions, daily intentions, habit strength, the
interaction between average daily intentions and habit
strength, and the interaction between daily intentions and
habit strength. We included sex (0 = female, 1 = male),
a dummy code of weekend (0 = Monday–Friday, 1 =
Saturday and Sunday), and study day (1–13) as covariates
in the model. We tested a hierarchical linear model that
optimized restricted maximum likelihood criterion in the
lmer function of the lme4 package in R version 2.14.1
(Bates, Maechler, & Bolker, 2011; R Development Core
Team, 2011). Condence intervals estimated from 100
posterior simulations using the sim function of the arm
package in R were used to test if the model coefcient
estimates signicantly varied from zero (Gelman & Hill,
2006; Gelman et al., 2012). This method of signicance

160 Rebar et al.
testing accounts for predictive and inferential uncertainty
in parameter estimations created by non-normal vari-
able distributions. The model, which consists of both
within-person (Level 1) and between-person (Level 2)
components, is represented by Equations 1–5.
Level 1:
Daily Physical Activitydi = β0i + β1i(Weekend)di +
β2i(Study Day)d + β3i(Daily Intentions)di + edi (1)
Level 2:
β0i = γ00 + γ01(Sex)i + γ02(Average Daily Intentions)i +
γ03(Habit Strength)i + γ04(Average Daily Intentions ×
Habit Strength)i + u0i (2)
β1i = γ10 + u1i (3)
β2i = γ20 (4)
β3i = γ30 + γ31(Habit Strength)i + u3i (5)
The within-person component, represented by Equa-
tion 1, models each individual i’s physical activity on
day d as a function of an intercept, Weekend (a binary
variable), Study Day (a continuous variable 1–13), and
that person’s Daily Intentions for physical activity.
These within-person parameters are constrained by the
between-person components of the model represented in
Equations 2–5, in which the γ00 – γ30 represent the xed
intercepts. Equation 2 includes the grand mean (γ00) and
tests whether Sex (γ01), Average Daily Intentions (γ02),
and Habit Strength (γ03) relate to Daily Physical Activ-
ity. In addition, it tests whether Average Daily Intentions
moderate the relation between Habit Strength and Daily
Physical Activity (γ04). Equation 3 tests the average differ-
ence in Daily Physical Activity on weekends as opposed
to weekdays (γ10). Equation 4 tests the average relation
between Study Day and Daily Physical Activity (γ20) and
this effect is xed to be the same between people (i.e.,
no random effects). Equation 5 tests the average relation
between Daily Intentions and Daily Physical Activity (γ30)
and the additional parameter is a cross-level interaction
that tests whether the relation between Habit Strength
and Daily Physical Activity (i.e., the estimate of β3i) is
moderated by Daily Intentions (γ31). The variance in Daily
Physical Activity unexplained by the model parameters
are represented in Equation 1 by edi, and the variance in
the within-person model coefcients unexplained by the
between-person constraints are represented by u0,1, and 3i
in Equations 2, 3, and 5.
Significant interaction coefficients were probed
using the Johnson–Neyman technique (Bauer & Curran,
2005; Hayes & Matthes, 2009). This technique overcomes
biases of other pick-a-point probing methods by math-
ematically determining which values of the moderator
closest to the mean result in signicant relations between
the predictor and outcome. The test whether a Level-1
variable (daily intentions) moderates a Level-2 predictor
(habit strength) is acceptable, although uncommon, for
cross-level interactions in hierarchical linear modeling
(Preacher, Curran, & Bauer, 2006).
Results
On average, participants completed 12 of 13 possible daily
intention measures. Most participants (95%) completed
at least 11 daily measures, and no participants completed
less than 8 daily measures. Accelerometer-derived physi-
cal activity data were available for an average of 12 days
for each participant. At least 11 days of accelerometer-
derived physical activity data were available for 82% of
participants, but 2 participants had no physical activity
data due to insufcient wear time and were not included
in analyses.
Table 1 presents descriptive statistics and between-
person correlations of physical activity, habit strength,
and intentions (correlations were calculated using within-
person means across days of intentions and physical activ-
ity). Habit strength and average daily intentions showed a
medium-sized positive relation (r = .37, p < .01). Physical
activity showed small positive relations with average
daily intentions (r = .21, p = .01) and habit strength (r =
.18, p = .06). Intraclass correlations indicated that less
than half of the variability in daily intentions and physical
activity were attributable to between-person variability,
suggesting that people’s intentions and physical activity
were variable across days. The average within-person
relation between daily intentions and daily physical
activity was not signicant (r = .01, p = .49, SD = 0.34).
Table 2 shows the coefcient estimates and the
condence intervals of the xed effects from the hierar-
Table 1 Descriptive Statistics and Correlations of Physical Activity,
Habit Strength, and Intentions
ICC
Variable
M SD
1 2 3
1 Habit Strength 6.43 1.44 — .37* .18*
2 Intentions 57.61 32.18 .49 — .21*
3Physical Activity 28,487.96 13,896.66 .30 —
Note. Between-person correlations were calculated between the within-person mean of daily intentions
and daily physical activity.
*p < .05.

Physical Activity Habit Strength 161
chical linear model testing the study hypotheses. People
with stronger average daily physical activity intentions
participated in more physical activity than people with
weaker average daily intentions (γ02), but habit strength
was not signicantly related to physical activity (γ03).
Average daily intentions did not signicantly moderate
the relation between habit strength and physical activity
(γ04), daily physical activity intentions were not signi-
cantly related to daily physical activity (γ30), but daily
intentions signicantly moderated the relation between
habit strength and daily physical activity (γ31). Figure 1
depicts the relation between habit strength and physical
activity on days when intentions were weaker than typical
(1 SD below M), typical (M), and stronger than typical
(1 SD above M).The Johnson–Neyman technique used to
probe the signicant moderating effect (Bauer & Curran,
2005; Hayes & Matthes, 2009) revealed that on days
when people’s intentions were typical (M) or stronger
than typical (scores > M) habit strength was unrelated to
daily physical activity, but on days when people’s inten-
tions were weaker than typical (scores < M – 0.55 SD),
habit strength was positively related to physical activity.
The coefficients of the covariates in the model
revealed that there were no sex differences in physical
activity after accounting for the other predictors in the
model (γ01), people tended to be more physically active
on weekdays than weekends (γ10), and study day did
not impact physical activity (γ20). There was signicant
variability in the amount of daily physical activity that
was unexplained by the model (edi = 134,623,586, SD
= 12,266.19), the relation between daily intentions and
daily physical activity (u2i = 3,837, SD = 61.86), and
the difference between weekday and weekend physical
activity (u1i = 6,093,565, SD = 2,468.51).
Table 2 Coefficient Estimates of Fixed Effects of Hierarchical Linear Model Testing the Moderating
Influence of Physical Activity Intentions on Relations Between Physical Activity Habit Strength and
Daily Physical Activity
Fixed Effects Coefficient
SE
95% Confidence Intervals
Intercept, γ00 28,283.79* 1,214.85 26,556.06 – 30,805.31
Sex, γ01 2,007.57 1,416.82 –814.83 – 4,529.32
Average Daily Intentions, γ02 102.84* 31.48 34.44 – 159.46
Habit Strength, γ03 810.75 520.42 –186.30 – 1,879.01
Average Daily Intentions × Habit Strength, γ04 –11.07 20.63 –52.89 – 28.46
Weekend, γ10 –2,342.09* 802.14 –4,186.42 – –332.43
Study Day, γ20 –29.46 84.37 –195.52 – 144.58
Daily Intentions, γ30 –17.39 16.42 –47.31 – 12.89
Habit Strength × Daily Intentions, γ31 –19.68* 10.34 –37.23 – –1.60
Note. Condence intervals based on 100 posterior simulations.
*p < .05.
Figure 1 — Slopes of the relations between habit strength and physical activity on days of
weaker than typical (1 SD < M), typical (M), and stronger than typical (1 SD > M) uctua-
tions in physical activity intentions with signicant slope marked with an asterisk.

162 Rebar et al.
Discussion
This study investigated the habitual and intentional regu-
lation of physical activity on a daily, within-person level.
When tested simultaneously, a person’s average daily
physical activity intentions, but not habit strength, posi-
tively related to physical activity. Although average daily
physical activity intentions did not signicantly moderate
the relation between physical activity habit strength and
physical activity, daily intentions were found to moderate
the relation between habit strength and physical activity.
Physical activity was regulated by people’s habit strength
when daily physical activity intentions were weaker than
usual. These results suggest that the ongoing regulation
of physical activity is a combination of controlled and
automatic processes.
Habits and Intentions Regulating Daily
Physical Activity
Previous research has shown that people with stronger
physical activity habits are more physically active than
people with weaker habits (Gardner et al., 2011; Rhodes
& de Bruijn, 2010), but this between-person effect was
not replicated in this study. In previous research, habit
was measured using the entire Self-Report Habit Index,
including items assessing frequency of physical activity
participation. This may have led to an overestimation
of the inuence of habit on physical activity, given that
other regulatory processes inuence past and future fre-
quency of behavior (Ajzen, 2002; Gardner et al., 2012).
The present study used a subscale of the index that
excluded frequency-related items; therefore, variability
shared between past and future behavior unrelated to
automaticity was not represented. Importantly, the null
main effect is not necessarily an indication of a lack of
correspondence between habit strength and physical
activity. People with strong physical activity habits tended
to also have strong physical activity intentions, suggest-
ing that people with strong physical activity habits have
chronically activated physical activity goals, consistent
with the habit-goal interface outlined by Wood & Neal
(2007). There is correspondence between the automatic
regulatory process of habits and the controlled regulatory
process of intentions; however, it seems that the inuence
of the automatic component of physical activity habits
does not generally extend beyond that of between-person
differences in physical activity intentions.
There is a strong research base suggesting that
people’s physical activity is linked to their intentions
to engage in physical activity (Hagger et al., 2002;
McEachan et al., 2011), so the present result that people
with stronger intentions were more physically active
than people with weaker intentions was not unexpected.
Repeated behaviors are not necessarily habits (Aarts &
Dijksterhuis, 2000; Ajzen, 2002; Wood & Neal, 2007),
so it may be that some people’s regular participation in
physical activity is the result of controlled regulatory
processes, such as intentions. It has also been shown
that intentions can be inferred from behavior (e.g., I
participate in physical activity regularly; therefore, I
must have strong intentions to do so; Banks & Isham,
2009; Eagleman, 2004; Wegner & Wheatley, 1999), so
physical activity may also lead to stronger physical activ-
ity intentions. Although the direction of the effect could
not be extrapolated from this observational design, the
current study replicated the well-established link between
intentions and physical activity and extended beyond
these previous studies by investigating the interactive
regulation of intentions and habits at both between- and
within-person levels.
Average daily intentions and habit strength were not
found to interact to regulate physical activity. This result
is similar to the ndings of a previous study (Rhodes et
al., 2010), but contrasts with research that showed that
behavior was regulated more by intentions for people
with weaker habits than for people with strong habits (de
Bruijn et al., 2009; de Bruijn & Rhodes, 2011; Gardner
et al., 2011; Ji & Wood, 2007; Knussen et al., 2004;
Verplanken et al., 1998). It may be that the difference
in ndings between this and the previous studies is par-
tially due to how between-person level physical activity
intentions were dened. Physical activity intentions are
typically measured as general intentions across the next
several weeks (e.g., de Bruijn et al., 2009; de Bruijn &
Rhodes, 2011), but in this study, intentions were dened
at a between-person level as the trend of a person’s daily
physical activity intentions across 2 weeks (see Fleeson,
2001, 2004). The ability to self-regulate uctuates on a
daily basis (Baumeister & Heatherton, 1996; Baumeister,
Muraven, & Tice, 2000; Conroy et al., 2011), and the
ndings of the current study demonstrate that the habitual
regulation of physical activity interacts with daily uctua-
tions in physical activity intentions, but not trends in a
person’s daily intentions across 2 weeks.
Habit strength only predicted physical activity on
days when people had weak intentions. As proposed by
dual process theories (Evans, 2008, 2009), these results
suggest that physical activity is regulated by both auto-
matic and controlled processes that interact on a daily
level. These results support Hoffmann and colleagues’
(2009, 2012) postulation that self-regulation must be
especially strong to override automatic tendencies.
Implications for Physical Activity
Motivation Theories and Interventions
Most contemporary physical activity motivation theories
focus on controlled regulatory processes such as inten-
tions, goals, values, and beliefs (Biddle & Mutrie, 2008).
Based on the results of this and other studies of automatic
regulation of physical activity (e.g., Conroy, Hyde,
Doerksen, & Ribeiro, 2010; Dimmock & Banting, 2009;
Gardner et al., 2011; Sheeran et al., 2012), it is becoming
increasingly clear that automatic regulatory processes
merit roles in our conceptualizations of physical activ-
ity motivation. In addition to strengthening people’s
physical activity intentions, physical activity motivation

Physical Activity Habit Strength 163
interventions may also benet from enhancing automatic
regulatory processes, such as habits, to compensate for
when physical activity intentions are weak.
Interventions with education components have been
shown to be effective in increasing intentions, but to be
less effective at changing behavior (Rhodes & Dickau,
2012; Webb & Sheeran, 2006). These interventions do
not address automatic regulatory processes, such as
habits. Enhancements of physical activity habit strength
may be possible through changes in the environment or
the introduction of habit development techniques, such
as implementation intentions—a form of behavior plan-
ning that species the context cues that will cue specic
behavioral responses (Gollwitzer, 1999; Marteau et al.,
2012; Sheeran & Orbell, 1999; Verplanken & Faes,
1999). Future investigations of intervention techniques
for targeting automatic regulation of physical activity will
help broaden the scope of physical activity interventions,
potentially improving the effectiveness of such efforts.
Strengths, Limitations and Future
Directions
Strengths of this study include the within-person study
design, the objective measurement of physical activity,
and the high completion rate of the repeated physical
activity intentions and behavior assessments. These
aspects of the study strengthen our condence in the
conclusions of the study, but further research is needed
to address limitations of this study to increase the gener-
alizability of the conclusions. Our study sampled mostly
White, high-functioning young adults. Physical activity
has been shown to differ as a function of sociodemo-
graphic characteristics (Trost, Owen, Bauman, Sallis,
& Brown, 2002), so studies with more heterogeneous
samples in terms of age, race, ethnicity, and functional
ability are necessary before these results can be general-
ized to a broader population. Although objective measure-
ment of physical activity reduces the risk for reporting
biases such as social desirability, the accelerometers that
were used also have limitations. For example, the devices
are not waterproof and therefore could not account for
water-based physical activity, such as swimming.
Participants may have been reactive to the monitor-
ing of their intentions (Chandon, Morwitz, & Reinartz,
2005) and behavior (Motl et al., 2012), so it is important
to note that the levels reported in this study may not
represent those present in the absence of monitoring.
This study investigated physical activity intentions and
habits, but other controlled and automatic processes are
also likely to play instrumental roles in regulating physi-
cal activity (e.g., self-efcacy, outcome expectancies,
automatic evaluations, automatic self-schemas). Future
research investigating the interaction between these
processes will also be important for clarifying how these
dual processes simultaneously regulate physical activity.
In sum, this study revealed that physical activity
is regulated by a person’s habits when they have weak
physical activity intentions for that day. Future research
and physical activity promotion efforts that attend to
habit development as well as to the time-varying nature
of intentions may enhance the effectiveness of those
interventions over time (via increased behavioral main-
tenance), increase the number of people who engage in
regular physical activity, and reduce the health burden
of physical inactivity.
References
Aarts, H., & Dijksterhuis, A. (2000). Habits as knowledge
structures: Automaticity in goal-directed behavior. Journal
of Personality and Social Psychology, 78, 53–63. PubMed
doi:10.1037/0022-3514.78.1.53
Aarts, H., Paulussen, T., & Schaalma, H. (1997). Physical
exercise habit: On the conceptualization and formation of
habitual health behaviours. Health Education Research,
12, 363–374. PubMed doi:10.1093/her/12.3.363
Ajzen, I. (1991). The theory of planned behavior. Organi-
zational Behavior and Human Decision Processes, 50,
179–211. doi:10.1016/0749-5978(91)90020-T
Ajzen, I. (2002). Residual effects of past on later behavior:
Habituation and reasoned action perspectives. Personality
and Social Psychology Review, 6, 107–122. doi:10.1207/
S15327957PSPR0602_02
Banks, W.P., & Isham, E.A. (2009). We infer rather than perceive
the moment we decided to act. Psychological Science, 20,
17–21. PubMed doi:10.1111/j.1467-9280.2008.02254.x
Bates, D., Maechler, M., & Bolker, B. (2011). lme4: Linear
mixed-effects models using S4 classes. R package version
0.999375-42. http://CRAN.Rproject.org/package=lme4.
Bauer, D.J., & Curran, P.J. (2005). Probing interactions in xed
and multilevel regression: Inferential and graphical tech-
niques. Multivariate Behavioral Research, 40, 373–400.
doi:10.1207/s15327906mbr4003_5
Baumeister, R.F., & Heatherton, T.F. (1996). Self-regulation
failure: An overview. Psychological Inquiry, 7, 1–15.
doi:10.1207/s15327965pli0701_1
Baumeister, R.F., Muraven, M., & Tice, D.M. (2000). Ego
depletion: A resource model of volition, self-regulation,
and controlled processing. Social Cognition, 18, 130–150.
doi:10.1521/soco.2000.18.2.130
Biddle, S., & Mutrie, N. (2008). Psychology of physical activity:
Determinants, well-being, and interventions. New York:
Taylor & Francis.
Caspersen, C.J., Pereira, M.A., & Curran, K.M. (2000). Changes
in physical activity patterns in the United States, by sex and
cross-sectional age. Medicine and Science in Sports and
Exercise, 32, 1601–1609. PubMed doi:10.1097/00005768-
200009000-00013
Chaiken, S., & Trope, Y. (1999). Dual-process theories in social
psychology. New York: Guilford Press.
Chandon, P., Morwitz, V.G., & Reinartz, W.J. (2005). Do
intentions really predict behavior? Self-generated validity
effects in survey research. Journal of Marketing, 69, 1–14.
doi:10.1509/jmkg.69.2.1.60755
Choi, L., Liu, Z., Matthews, C.E., & Buchowski, M.S.
(2011). Validation of accelerometer wear and nonwear
time classication algorithm. Medicine and Science in

164 Rebar et al.
Sports and Exercise, 43, 357–364. PubMed do i:10.1249 /
MSS.0b013e3181ed61a3
Conroy, D.E., Doerksen, S.E., Elavsky, S., & Maher, J.P. (2013).
A daily process a nalysis of intentions and physical activity
in college students. Journal of Sport & Exercise Psychol-
ogy, 35, 493 –502. PubMed
Conroy, D.E., Elavsk y, S., Hyde, A.L., & Doerksen, S.E. (2011).
The dynamic nature of physical activity intentions: A
within-person perspective on intention-behavior coupling.
Journal of Sport & Exercise Psychology, 33, 80 7– 82 7.
PubMed
Conroy, D.E., Hyde, A.L., Doerksen, S.E., & Ribeiro, N.F.
(2010). Implicit attitudes and explicit motivation prospec-
tively predict physical activity. Annals of Behavioral Medi-
cine, 39, 112–118. PubMed d oi:10.100 7/s1216 0- 010 -9161- 0
de Bruijn, G-J., Kremers, S.P.J., Singh, A., van den Putte, B.,
& van Mechelen, W. (2009). Adult active transportation:
Adding habit strength to the theory of planned behavior.
American Journal of Preventive Medicine, 36, 189–194.
PubMed doi:10.1016/j.amepre.2008.10.019
de Bruijn, G-J., & Rhodes, R.E. (2011). Exploring exercise
behavior, intention and habit strength relationships. Scan-
dinavian Journal of Medicine & Science in Sports, 21,
482–491. PubMed doi :10.1111 /j.160 0- 0838.200 9.010 64.x
Dimmock, J.A., & Banting, L.K. (2009). The inuence of
implicit cognitive processes on physical activity: How
the theory of planned behaviour and self-determination
theory can provide a platform for our understanding.
International Review of Sport and Exercise Psychology,
2, 3–22. doi:10.1080/17509840802657337
Downs, D.S., & Hausenblas, H.A. (2005). The theories of
reasoned action and planned behavior applied to exercise:
A meta-analytic update. Journal of Physical Activity and
Health, 2, 76–9 7.
Eagleman, D.M. (2004). The where and when of intention.
Science, 303, 114 4–1146 . PubMed doi:10.1126/sci-
en ce.1095331
Evans, J.S.B.T. (2008). Dual-processing accounts of reason-
ing, judgment, and social cognition. Annual Review of
Psychology, 59, 255–278. PubMed do i:10.114 6/an nu re v.
psych.59.103006.093629
Evans, J.S.B.T. (2009). How many dual-process theories do we
need? One, two, or many? In two minds: Dual processes
and beyond. New York: Oxford University Press.
Fleeson, W. (2001). Towards a structure- a nd process-integrated
view of personality: Traits as density distributions of
states. Journal of Personality and Social Psychology,
80, 1011–1027. PubMed doi:10.1037/0022-3514.80.6.1011
Fleeson, W. (2004). Moving personality beyond the person-
situation debate: The challenge and the opportunity
of within-person variability. Current Directions in
Psychological Science, 13(2), 83–87. doi:10.1111/j .0963 -
7214.2004.00280.x
Gardner, B., Abraham, C., Lally, P., & de Bruijn, G.-J.. (2012).
Towards parsimony in habit measurement: Testing the con-
vergent and predictive validity of an automaticity subscale
of the Self-Report Habit Index. The International Journal
of Behavioral Nutrition and Physical Activity, 9, 102 –113.
PubMed d oi:10.1186/1479 -5868-9-10 2
Gardner, B., de Bruijn, G.-J., & Lally, P. (2011). A systematic
review and meta-analysis of applications of the Self-Report
Habit Index to nutrition and physical activity behaviors.
Annals of Behavioral Medicine, 42, 174–187. PubMed
doi:10.1007/s12160-011-9282-0
Gelman, A., & Hill, J. (2006). Data analysis using regression
and multilevel/hierarchical models (1st ed.). New York:
Cambridge University Press.
Gelman, A., Su, Y-S., Yajima, M., Hill, J., Pittau, M.G., Kerman,
J., & Zheng, T. (2012). arm: Data analysis using regression
and multilevel/hierarchical models. R package version 1.5-
03. http://CRAN.R-project.org/package=arm.
Gollwitzer, P.M. (1999). Implementation intentions: Strong
effects of simple plans. The American Psychologist, 54,
493–503. doi:10.1037/0003-066X.54.7.493
González, M.C., Hidalgo, C.A., & Barabási, A-L. (2008). Under-
standing individual human mobility patterns. Nature, 453,
779–782. PubMed doi:10.1038/nature06958
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 valid-
ity and the contribution of additional variables. Journal of
Sport & Exercise Psychology, 24, 3–32.
Hayes, A.F., & Matthes, J. (2009). Computational procedures for
probing interactions in OLS and logistic regression: SPSS
and SAS implementations. Behavior Research Methods,
41, 924–936. PubMed doi:10.3758/BRM.41.3.924
Hofmann, W., Friese, M., & Strack, F. (2009). Impulse and
self-control from a dual-systems perspective. Perspectives
on Psychological Science, 4, 162–176. doi:10.1111/j.1745-
6924.2009.01116.x
Hofmann, W., Schmeichel, B.J., & Baddeley, A.D. (2012).
Executive functions and self-regulation. Trends in Cog-
nitive Sciences, 16, 174–180. PubMed doi:10.1016/j.
tics.2012.01.006
Ji, M.F., & Wood, W. (2007). Purchase and consumption
habits: Not necessarily what you intend. Journal of
Consumer Psychology, 17, 261–276. doi:10.1016/S1057-
7408(07)70037-2
Knussen, C., Yule, F., MacKenzie, J., & Wells, M. (2004). An
analysis of intentions to recycle household waste: The
roles of past behaviour, perceived habit, and perceived
lack of facilities. Journal of Environmental Psychology,
24, 237–246. doi:10.1016/j.jenvp.2003.12.001
Marteau, T.M., Hollands, G.J., & Fletcher, P.C. (2012). Chang-
ing human behavior to prevent disease: The importance of
targeting automatic processes. Science, 337, 1492–1495.
PubMed doi:10.1126/science.1226918
Matthews, C.E., Ainsworth, B.E., Thompson, R.W., & Bassett,
D.R. (2002). Sources of variance in daily physical activ-
ity levels as measured by an accelerometer. Medicine and
Science in Sports and Exercise, 34, 1376–1381. PubMed
doi:10.1097/00005768-200208000-00021
McArthur, L.H., & Raedeke, T.D. (2009). Race and sex dif-
ferences in college student physical activity correlates.
American Journal of Health Behavior, 33, 80–90. PubMed
doi:10.5993/AJHB.33.1.8
McEachan, R.R.C., Conner, M., Taylor, N.J., & Lawton, R.J.
(2011). Prospective prediction of health-related behaviours

Physical Activity Habit Strength 165
with the Theory of Planned Behaviour: a meta-analysis.
Health Psychology Review, 5, 97–144. doi:10.1080/1743
7199.2010.521684
Motl, R.W., McAuley, E., & Dlugonski, D. (2012). Reactivity in
baseline accelerometer data from a physical activity behav-
ioral intervention. Health Psychology, 31(2), 172–175.
PubMed doi:10.1037/a0025965
Neal, D.T., Wood, W., Labrecque, J.S., & Lally, P. (2012). Do
habits depend on goals? Perceived versus actual role of
goals in habit performance. Journal of Experimental Social
Psychology, 48, 492–498. doi:10.1016/j.jesp.2011.10.011
Neal, D.T., Wood, W., Wu, M., & Kurlander, D. (2011). The
pull of the past: When do habits persist despite conict
with motives? Personality and Social Psychology Bulletin,
37, 1428–1437. PubMed doi:10.1177/0146167211419863
Neal, D.T., Wood, W., & Quinn, J.M. (2006). Habits: A repeat
performance. Current Directions in Psychological Science,
15, 198–202. doi:10.1111/j.1467-8721.2006.00435.x
Ouellette, J.A., & Wood, W. (1998). Habit and intention in
everyday life: The multiple processes by which past behav-
ior predicts future behavior. Psychological Bulletin, 124,
54–74. doi:10.1037/0033-2909.124.1.54
Physical Activity Guidelines Advisory Committee. (2008).
Physical activity guidelines advisory committee report.
Washington, DC: U.S. Department of Health and Human
Services.
Preacher, K.J., Curran, P.J., & Bauer, D.J. (2006). Computational
tools for probing interactions in multiple linear regression,
multilevel modeling, and latent curve analysis. Journal of
Educational and Behavioral Statistics, 31(4), 437–448.
doi:10.3102/10769986031004437
R Development Core Team. (2011). R: A language and environ-
ment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria. ISBN 3-900051-07-0, URL
http://www.R-project.org/.
Rhodes, R.E., & de Bruijn, G-J. (2010). Automatic and motiva-
tional correlates of physical activity: Does intensity moderate
the relationship? Behavioral Medicine (Washington, D.C.),
36, 44–52. PubMed doi:10.1080/08964281003774901
Rhodes, R.E., de Bruijn, G-J., & Matheson, D.H. (2010). Habit
in the physical activity domain: Integration with intention
temporal stability and action control. Journal of Sport &
Exercise Psychology, 32, 84–98. PubMed
Rhodes, R.E., & Dickau, L. (2012). Experimental evidence for
the intention–behavior relationship in the physical activity
domain: A meta-analysis. Health Psychology, 31, 724–727.
PubMed doi:10.1037/a0027290
Scholz, U., Keller, R., & Perren, S. (2009). Predicting behav-
ioral intentions and physical exercise: A test of the health
action process approach at the intrapersonal level. Health
Psychology, 28, 702–708. PubMed doi:10.1037/a0016088
Scholz, U., Nagy, G., Schüz, B., & Ziegelmann, J.P. (2008). The
role of motivational and volitional factors for self‐regulated
running training: Associations on the between‐ and within‐
person level. The British Journal of Social Psychology,
47, 421–439. PubMed doi:10.1348/014466607X266606
Schwartz, J.E., & Stone, A.A. (1998). Strategies for analyzing
ecological momentary assessment data. Health Psychology,
17, 6–16. PubMed doi:10.1037/0278-6133.17.1.6
Sheeran, P., Gollwitzer, P.M., & Bargh, J.A. (2013). Nonconscious
processes and health. Health Psychology, 32, 460–473.
doi:10.1037/a0029203
Sheeran, P., & Orbell, S. (1999). Implementation intentions and
repeated behaviour: Augmenting the predictive validity
of the theory of planned behaviour. European Journal of
Social Psychology, 29, 349–369. doi:10.1002/(SICI)1099-
0992(199903/05)29:2/3<349::AID-EJSP931>3.0.CO;2-Y
Song, C., Qu, Z., Blumm, N., & Barabási, A-L. (2010). Limits of
predictability in human mobility. Science, 327, 1018–1021.
PubMed doi:10.1126/science.1177170
Trost, S.G., Owen, N., Bauman, A.E., Sallis, J.F., & Brown,
W. (2002). Correlates of adults’ participation in physi-
cal activity: Review and update. Medicine and Science
in Sports and Exercise, 34(12), 1996–2001. PubMed
doi:10.1097/00005768-200212000-00020
Trost, S.G., Pate, R.R., Freedson, P.S., Sallis, J.F., & Taylor, W.C.
(2000). Using objective physical activity measures with
youth: How many days of monitoring are needed? Medicine
and Science in Sports and Exercise, 32, 426–431. PubMed
doi:10.1097/00005768-200002000-00025
Tucker, J.M., Welk, G.J., & Beyler, N.K. (2011). Physical
Activity in U.S. adults: Compliance with the physical
activity guidelines for Americans. American Journal of
Preventive Medicine, 40, 454–461. PubMed doi:10.1016/j.
amepre.2010.12.016
Verplanken, B. (2010). Habit: From overt action to mental events.
Then a miracle occurs: Focusing on behavior in social
psychological theory and research (pp. 68–88). New York:
Oxford University Press.
Verplanken, B., Aarts, H., Knippenberg, A., & Moonen, A. (1998).
Habit versus planned behaviour: A eld experiment. The
British Journal of Social Psychology, 37, 111–128. PubMed
doi:10.1111/j.2044-8309.1998.tb01160.x
Verplanken, B., & Faes, S. (1999). Good intentions, bad
habits, and effects of forming implementation inten-
tions on healthy eating. European Journal of Social
Psychology, 29, 591–604. doi:10.1002/(SICI)1099-
0992(199908/09)29:5/6<591::AID-EJSP948>3.0.CO;2-H
Verplanken, B., & Melkevik, O. (2008). Predicting habit: The case
of physical exercise. Psychology of Sport and Exercise, 9,
15–26. doi:10.1016/j.psychsport.2007.01.002
Verplanken, B., & Orbell, S. (2003). Reections on past behavior:
A self‐report index of habit strength. Journal of Applied Social
Psychology, 33, 1313–1330. doi:10.1111/j.1559-1816.2003.
tb01951.x
Webb, T.L., & Sheeran, P. (2006). Does changing behavioral
intentions engender behavior change? A meta-analysis of
the experimental evidence. Psychological Bulletin, 132,
249–268. PubMed doi:10.1037/0033-2909.132.2.249
Wegner, D.M., & Wheatley, T. (1999). Apparent mental causation:
Sources of the experience of will. The American Psychologist,
54, 480–492. PubMed doi:10.1037/0003-066X.54.7.480
Wood, W., & Neal, D.T. (2007). A new look at habits and the
habit-goal interface. Psychological Review, 114, 843–863.
PubMed doi:10.1037/0033-295X.114.4.843
Manuscript submitted: August 6, 2013
Revision accepted: December 11, 2013

CopyrightofJournalofSport&ExercisePsychologyisthepropertyofHumanKinetics
Publishers,Inc.anditscontentmaynotbecopiedoremailedtomultiplesitesorpostedtoa
listservwithoutthecopyrightholder'sexpresswrittenpermission.However,usersmayprint,
download,oremailarticlesforindividualuse.