Journal of Sport & Exercise Psychology, 2011, 33, 175-197
© 2011 Human Kinetics, Inc.
Simon J. Sebire is now with the Centre for Exercise, Nutrition & Health Sciences, University of Bris-
tol, Bristol, U.K. Martyn Standage is with the Department of Health, University of Bath, Bath, U.K.
Maarten Vansteenkiste is with the Department of Psychology, University of Ghent, Ghent, Belgium.
Predicting Objectively Assessed Physical
Activity From the Content
and Regulation of Exercise Goals:
Evidence for a Mediational Model
Simon J. Sebire,1 Martyn Standage,1
and Maarten Vansteenkiste2
1University of Bath and 2Ghent University
Grounded in self-determination theory (Deci & Ryan, 2000), the purpose of this
work was to examine effects of the content and motivation of adults’ exercise goals
on objectively assessed moderate-to-vigorous physical activity (MVPA). After
reporting the content and motivation of their exercise goals, 101 adult participants
(Mage = 38.79 years; SD = 11.5) wore an ActiGraph (GT1M) accelerometer for
seven days. Accelerometer data were analyzed to provide estimates of engagement
in MVPA and bouts of physical activity. Goal content did not directly predict
behavioral engagement; however, mediation analysis revealed that goal content
predicted behavior via autonomous exercise motivation. Specifically, intrinsic
versus extrinsic goals for exercise had a positive indirect effect on average daily
MVPA, average daily MVPA accumulated in 10-min bouts and the number of
days on which participants performed 30 or more minutes of MVPA through
autonomous motivation. These results support a motivational sequence in which
intrinsic versus extrinsic exercise goals influence physical activity behavior because
such goals are associated with more autonomous forms of exercise motivation.
Keywords: self-determination theory, exercise goals, motivation, accelerometer
Owing to epidemic levels of physical inactivity, overweight and obesity (World
Health Organization, 2002) and the comorbidities associated with such factors
(Hardman & Stensel, 2003), it is recommended that adults perform five 30-min
bouts of moderate-to-vigorous physical activity (MVPA) per week accumulated
in bouts of activity at least 10 min in duration (Physical Activity Guidelines Advi-
sory Committee, 2008). As such, understanding motivation for physical activity
behavior, including exercise, is a priority.1 A conceptual framework frequently
used to study motivation for health-enhancing behaviors such as physical activity
176 Sebire, Standage, and Vansteenkiste
is self-determination theory (SDT; Deci & Ryan, 2000; Ryan & Deci, 2007; Ryan,
Patrick, Deci, & Williams, 2008).
The goals that individuals pursue are important when studying motivation
using SDT. Austin and Vancouver (1996) define goals as “internal representations of
desired states, where states are broadly construed as outcomes, events or processes”
(p. 338), and within SDT both the content and the behavioral regulation of goals
are important considerations. Goal content refers to the “what” of motivation, or
a person’s specific aspiration (e.g., to be healthier), whereas behavioral regulation
refers to the “why” of motivation, or a person’s reasoning behind their goal (e.g.,
because a physician has told them to exercise) (Deci & Ryan, 2000). The purpose
of the current study was to investigate the prediction of objectively assessed MVPA
behavior from the “what” and the “why” of exercise goal pursuit.
Goal Content and Motivational Regulations
Self-determination theory is based on organismic (i.e., humans are hypothesized
to be growth oriented) and dialectic (i.e., growth occurs via environmental interac-
tions) foundations (Deci & Ryan, 2000). Central to this perspective is the proposal
of three psychological needs that are purported to underlie optimal human growth,
effective functioning and adaptive environmental interactions, namely, autonomy,
competence and relatedness. Autonomy is the need to be the origin of one’s behavior,
competence is the need to interact effectively with and to master one’s environment
and relatedness refers to the need to feel a sense of mutual connectedness in one’s
social surroundings (Deci & Ryan, 2000). Deci and Ryan (2000) contend that
where goal pursuit (through its content and motivation) facilitates need satisfaction,
adaptive cognitive, affective and behavioral outcomes will occur.
Within the broader SDT framework, goal contents theory (GCT; Ryan, Wil-
liams, Patrick, & Deci, 2009; Vansteenkiste et al., 2010) holds that goals can be
defined as intrinsic (i.e., are inwardly focused on self-development and conducive
to the satisfying of psychological needs) or extrinsic (i.e., are outwardly focused,
related to self-evaluative concerns and unsatisfying of psychological needs) (Deci
& Ryan, 2000; Vansteenkiste, Lens, & Deci, 2006). Research from this perspective
has shown that where intrinsic aspirations (e.g., community contribution, physical
fitness or social affiliation) are more central to people’s lives than extrinsic aspi-
rations (e.g., wealth, fame or appearance) they experience greater psychological
well-being, less depression and anxiety, and reduced physical symptoms (see Kasser,
2002; Kasser, Cohn, Kanner, & Ryan, 2007 and Vansteenkiste, Soenens, & Duriez,
2008, for reviews). Conversely, extrinsic, relative to intrinsic goal endorsement is
negatively associated with well-being (Sheldon, Ryan, Deci, & Kasser, 2004) and
positively associated with body image concerns (Thøgersen-Ntoumani, Ntoumanis,
& Nikitaras, 2010), antisocial attitudes (Duriez, Vansteenkiste, Soenens, & De Witte,
2007; McHoskey, 1999) and behaviors (Sheldon & McGregor, 2000). In addition,
positive associations between relative intrinsic life aspirations and health outcomes
such as tobacco abstinence and number of days not smoking (Niemiec, Ryan, Deci
& Williams, 2009) and between relative extrinsic aspirations and multiple health
risk behaviors (Williams, Cox, Hedberg & Deci, 2000) have been reported.
Research examining goal contents from an SDT perspective has recently been
extended to the exercise domain (Sebire, Standage, & Vansteenkiste, 2008; see
Goal Content and Physical Activity 177
Standage & Ryan, in press, for a review). In line with GCT, Sebire, Standage, and
Vansteenkiste (2009) identified that the pursuit of intrinsic (i.e., health manage-
ment, skill development and social affiliation) relative to extrinsic (i.e., image and
social recognition) exercise goals was positively associated with psychological
need satisfaction in exercise, physical self-worth, psychological well-being, and
self-reported exercise behavior and negatively associated with exercise-based
anxiety. These observations support previous findings delineating positive associa-
tions between intrinsic (e.g., health) and extrinsic (e.g., image) exercise goals with
adaptive and maladaptive psychological outcomes respectively (e.g., Crawford &
Eklund, 1994; Maltby & Day 2001).
In the present work we aimed to build upon previous research (e.g., Sebire,
Standage, & Vansteenkiste, 2008; 2009) by (a) focusing on the prediction of behav-
ioral outcomes (i.e., objectively assessed bouts of MVPA) and (b) by exploring
in more detail the motivational dynamics between goal content and motivational
With regards to motivational regulation, in SDT, motivation is classified as
autonomous or controlled based on the degree to which a goal is pursued with a
sense of self-determination (Deci & Ryan, 2000). Autonomous motivation com-
prises intrinsic motivation and identified regulation.2 Intrinsic motivation stems
from the inherent satisfaction, interest or challenge of participation in an activity,
whereas identified regulation derives from a sense of value and personal endorse-
ment of an activity. Controlled motivation reflects less self-determined reasons for
goal pursuit and is underpinned by introjected regulation in which motivation stems
from self-imposed sanctions such as guilt, shame or pride and external regulation
where behaviors are enacted to comply with external demands or to obtain exter-
nally based rewards. When an individual’s actions are underpinned by autonomous
versus controlled motivation it is hypothesized that they will experience adaptive
psychological functioning and greater behavioral performance and persistence
(Deci & Ryan, 2000). Consistent with this hypothesis, studies of adults’ motiva-
tion in the exercise domain, have largely identified positive associations between
autonomous motivation toward exercise and self-reported exercise (Edmunds,
Ntoumanis, & Duda, 2006; Wilson, Rodgers, Blanchard, & Gessell, 2003; Wilson,
Rodgers, Fraser, & Murray, 2004) and objectively assessed (Standage, Sebire, &
Loney, 2008) bouts of MVPA.
Despite representing theoretically distinct constructs, goal content and motiva-
tional regulation tend to covary as they share associations with psychological need
satisfaction (Deci & Ryan, 2000). Evidence from previous studies in the exercise
domain supports this assertion, identifying an average bivariate correlation between
the two concepts of r = .44 (Sebire, Standage, & Vansteenkiste, 2008; 2009).
Although correlated, the average correlation between goal content and motivational
regulation is not indicative of redundancy and it is hypothesized that scores on these
two sets of variables predict independent variance or have independent effects (e.g.,
the initial association between goal content as a predictor and an outcome remains
significant after entering motivational regulation as a predictor) on outcomes such
as well-being (Deci & Ryan, 2000; Sheldon et al., 2004). This hypothesis has been
challenged by some researchers (e.g., Carver & Baird, 1998; Srivastava, Locke, &
Bartol, 2001) who contend that the observed associations between relative intrinsic
goal content and well-being (e.g., Kasser & Ryan, 1996) are due to autonomous and
178 Sebire, Standage, and Vansteenkiste
controlled motivation which underlie the pursuit of intrinsic and extrinsic goals,
respectively. However, studies of the constructs at the general life level (Sheldon
et al., 2004) and in exercise (Sebire et al., 2009) have provided support for their
independent effects on cognitive and affective outcomes (e.g., psychological well-
being, physical self-worth and exercise-based anxiety), although other researchers
(Gardarsdottir, Dittmar, & Aspinall, 2009; Srivastava et al., 2001), using measures
that are not directly grounded in SDT, failed to provide support for the independent
effects of goal contents.
Goal Content, Behavioral Regulation and the Prediction
of Exercise Behavior
Although the independent effects of goal content and motivation on cognitive and
affective outcomes has been reasonably supported (Sebire et al., 2009; Sheldon et
al., 2004), similar findings have not been observed when testing behavioral indices
in the physical activity domain. Indeed, work across a number of life domains using
experimental and cross-sectional research designs has yielded equivocal findings.
For example, Vansteenkiste, Simons, Lens, Sheldon, and Deci (2004) experi-
mentally manipulated goals for learning tae-bo exercises in a student sample in terms
of attaining intrinsic (e.g., physical health) and extrinsic (e.g., appearing attractive
to others) outcomes. Participants who received the intrinsic versus extrinsic goal
framing manipulation performed better on tests related to tae-bo performance,
persisted longer in free-choice tae-bo activities and displayed greater autonomous
motivation. Mediation analysis revealed that test performance was independently
predicted by the intrinsic goal content and autonomous motivation; however, the
effect of the intrinsic goal content manipulation on behavioral persistence was
mediated by autonomous motivation.
In a cross-sectional study of adolescent British schoolchildren, Gillison,
Standage and Skevington (2006) found individual differences in exercise goal
content to have weak direct effects on self-reported leisure-time exercise, as well as
indirect effects via exercise motivational regulation. Recent work in large samples
of British adults (e.g., Ingledew & Markland, 2008; Sebire et al. 2009) has, how-
ever, found that goal content no longer yielded an independent association with
self-reported exercise behavior once motivation was taken into account. Sebire
et al. (2009) suggested that it could be the case that one’s behavioral motivation,
relative to exercise goal content, may be more proximal to behavioral engagement
as these motivations are more present-oriented. In contrast, the cognitive foci on
the content of one’s exercise goals (e.g., the promise of social recognition or the
ideal appearance) may be rather distal as goals are typically future-oriented, thereby
being situated at mid- to even long-term distance from one’s current behavior. As
a result, the chance that goal content predicts recently enacted and/or proximal
behavioral engagement above and beyond the reasons by which one is motivated
to act are considerably reduced. In line with previous work (Gillison et al., 2006;
Ingledew & Markland, 2008), we suggest that intrinsic and extrinsic goal content
precedes the motivational regulation of that goal: a focus on intrinsic, relative to
extrinsic, goals would be conducive to physical activity because one would derive
more enjoyment and pleasure from the activity (autonomous motivation) and feel
less pressured to do it (controlled motivation).
Goal Content and Physical Activity 179
Taken collectively, the findings of extant work examining whether the effect
of exercise goal content on exercise behavior occur above and beyond or via
exercise behavioral motivation (e.g., Vansteenkiste, Simons, Lens et al. 2004;
Gillison et al., 2006; Ingledew & Markland, 2008; Sebire et al., 2009) paint an
unclear picture. However, apart from the evidence from the Vansteenkiste, Simons,
Lens, et al. (2004) study, evidence of the independent effects of goal content and
motivational regulations on physical activity behavior is derived from studies that
used cross-sectional study design. Further, only two of these studies (i.e., Ingledew
& Markland, 2008; Sebire et al., 2009) were conducted among adult populations.
In addition, although the predictive effects of autonomous and controlled motiva-
tion on physical activity have been assessed using objectively measured behavior
(e.g., Standage et al., 2008), studies of the relationship between goal content and
physical activity have relied on behavioral self-reports, with the exception of the
experimental goal framing studies by Vansteenkiste and colleagues (Vansteen-
kiste, Simons, Lens, et al., 2004; Vansteenkiste, Simons, Soenens, & Lens, 2004).
However, in these experimental investigations, participants’ voluntary engage-
ment in single exercise sessions rather than their daily engagement was assessed.
As such, further studies in adult populations using robust and daily behavioral
assessments are warranted.
Self-report assessments of activity behavior are cost and time effective but
are vulnerable to error from factors such as social desirability and recall biases
(see Dale, Welk, & Matthews, 2002 for a review of physical activity assessment).
Accelerometers are small electronic motion sensors that objectively detect bodily
acceleration, which can be used to quantify physical activity. Using accelerometers
in a field that relies on questionnaire inventories for the measurement of most cogni-
tive constructs may reduce common method variance: a source of systematic error
where observed relationships between variables may be artificially inflated by the
similarity of measurement methods (Podsakoff, MacKenzie, Lee, & Podsakoff,
2003). In addition, accelerometers eliminate issues pertaining to activity recall
biases and accelerometer data can be used to more accurately quantify physical
activity duration, activity in continuous bouts that may reflect purposeful exercise
and the meeting of public health recommendations at an individual level (Loney,
Standage, Thompson, Sebire, & Cumming, in press). Knowing whether theoreti-
cal constructs such as goal content and motivation predict meaningful behavioral
engagement, such as the meeting of public health recommendations is important
for psychologists and health practitioners alike.
The Present Research
To summarize, the purpose of the present work was to examine the prediction of
MVPA and continuous bouts of MVPA by the content and motivation of exercise
goals in an adult sample. Extending previous work, accelerometers were used to
quantify meaningful behavioral engagement (i.e., that recommended for health).
180 Sebire, Standage, and Vansteenkiste
Participants were volunteers from southwest England. Participants were eligible
for inclusion if they (a) were aged 18–65 years, (b) reported performing at least
one exercise session per week on average (to ensure the relevance of questions
pertaining to exercise motivation), (c) were not a solely water-based exerciser (e.g.,
a swimmer) and (d) were free from conditions that restrict physical activity.
The study protocol was completed by 107 participants (racial identity; White
= 99%, Asian = 1%). Owing to device malfunction, behavioral data were incom-
plete (i.e., <5 valid days of data) for six participants. The final sample comprised
101 individuals (33 men, 68 women; Mage = 38.79 years; SD = 11.50, age range =
21.39–63.39 years). Of the participants, 61.7% were university employees, 21.5%
members of the public, and 16.8% postgraduate students. Of the university employ-
ees, 54.5% worked in administration, 18.3% teaching, 13.6% research and 13.6%
technical employees. Secondary-level and tertiary-level education were reported
by 91.1% and 8.9% of participants, respectively.
The average body mass index (BMI) of the sample (M = 24.33 kg/m2; SD =
3.52; range = 19.17–37.89 kg/m2) resided near the upper threshold of the “normal”
range and average waist circumference (WC) values (males: M = 85.76 cm; SD =
10.80, females: M = 76.03 cm; SD = 8.56) were within gender-appropriate “low
risk” ranges (see American College of Sports Medicine, 2006).
Following institutional ethical approval, participants were recruited by an adver-
tisement on a university Web page and posters located around a university campus.
The protocol of the study followed a three time-point design, which lasted 16 days.
At Time 1, participants were given information about the study, asked to provide
informed consent and demographics, and responses to exercise goal content and
motivation self-report measures were obtained. Seven days later (Time 2) par-
ticipants’ anthropomorphic measurements (height, weight and WC) were taken.
Participants were given an ActiGraph GT1M accelerometer (ActiGraph, LLC,
Pensacola, FL) to wear for the next seven days and a daily physical activity log
to complete. Eight days after Time 2, participants returned the accelerometer and
activity log and were debriefed as to the purpose of the study (Time 3).
Exercise Goal Content. The goal content for exercise questionnaire (GCEQ;
Sebire, Standage, & Vansteenkiste, 2008) was used to assess the importance
participants placed on intrinsic exercise goals for health management, skill
development and social affiliation and extrinsic goals for image and social
recognition. The GCEQ consists of 20 items (four per goal subscale) which are
rated on a 7-point Likert scale ranging from 1 (not at all important) to 7 (extremely
important). Previous research showed that scores derived from the GCEQ are
internally consistent and display construct validity, temporal stability and gender
invariance (Sebire, Standage, & Vansteenkiste, 2008; 2009). In the current study,
Goal Content and Physical Activity 181
internal consistency of the five subscales was as follows; health management (α =
.82), skill development (α = .89), social affiliation (α = .84), image (α = .89) and
social recognition (α = .89). In line with the SDT literature (Deci & Ryan, 2000)
and previous work (Sebire et al., 2009; Sheldon et al., 2004) a relative intrinsic goal
variable was calculated by subtracting the average of scores on extrinsic goal items
(α = .87) from the average of scores on intrinsic goal items (α = .85).
Exercise Motivation. Participants completed Mullan, Markland and Ingledew’s
(1997) 15-item behavioral regulations in exercise questionnaire (BREQ). The BREQ
assesses intrinsic motivation and identified, introjected and external exercise-based
motivational regulations. Participants respond to the stem “why do you exercise?”
and rated each item on a 5-point Likert scale ranging from 0 (not true for me) to
4 (very true for me). In the current study the internal consistency of the BREQ
subscales was as follows: intrinsic motivation (α = .92), identified regulation (α =
.82), introjected regulation (α = .82) and external regulation (α = .77). Bivariate
correlations between the subscales conformed to a simplex pattern (Guttman,
1954). As such, and aligned with past work (Koestner, Otis, Powers, Pelletier, &
Gagnon, 2008; Ntoumanis & Standage, 2009) scores on intrinsic motivation and
identified regulation were averaged to form an autonomous motivation variable (α
= .88) and introjected regulation and external regulation scores were averaged to
form a controlled motivation variable (α = .66).
Physical Activity. Participants wore an ActiGraph GT1M accelerometer
(ActiGraph, LLC, Pensacola, FL). The GT1M is a small (3.8 × 3.7 × 1.8 cm)
biaxial accelerometer that detects time-varying acceleration (approximate range
= 0.05–2.0 g) in the vertical plane 30 times per second. These measurements are
then summed to provide a number of “counts” over a user-defined epoch (e.g.,
60 s). Previous research supports the reliability (Brage, Brage, Wedderkopp, &
Froberg, 2003) and validity (Plasqui & Westerterp, 2007) of scores derived from
the Actigraph unit.
Accelerometers were secured to a nylon belt and worn tightly around partici-
pants’ waists with the accelerometer positioned on the midaxillary line of the right
hip (Trost, McIver, & Pate, 2005). Participants were instructed verbally and via
standardized written instructions to start wearing the accelerometer on waking
each day, to wear it throughout the day and to remove it on going to bed at night.
Participants also removed the accelerometer during participation in water sports,
showering and bathing. To address potential reactivity to wearing the accelerometer,
participants were instructed that it was measuring the time of day they moved and
care was taken not to divulge its true purpose (i.e., to measure level of physical
activity). In addition, participants wore the accelerometer for the remainder of the
day of their Time 2 appointment; however, Actigraphs were programmed to start
from 00:01 the following morning. Accelerometer data were collected in 60-s
epochs for seven consecutive days.
During the behavioral monitoring, participants completed a daily log of the
major activities (both active and inactive) they performed. Participants recorded a
description, the duration and perceived intensity (i.e., light, moderate or vigorous)
of the activities in addition to times and details of activities performed when they
did not wear the monitor during waking hours (e.g., for showering or swimming).
182 Sebire, Standage, and Vansteenkiste
Accelerometer Data. Accelerometer data were downloaded to a PC and imported
to a Microsoft Access database for analysis. The number of “valid days” of data for
each participant was defined using the 70/80 rule (Catellier et al., 2005) combined
with activity log data (Trost et al., 2005). The 70/80 rule indicated that a valid
weekday and weekend day should consist of 732 min and 664 min of recorded
data respectively. Data were screened using a Microsoft Access macro that ignored
periods in the data file at which the accelerometer count was zero for ≥31 continuous
minutes, which may indicate that the accelerometer was not being worn. There is
no recommended threshold of continuous zero counts for adult samples. Owing
to the age and sedentary occupations of the participants in the current study,
additional analyses were conducted to determine a feasible length of continuous
zeros of accelerometer data ≥20 min that occur when participants reported wearing
the accelerometer in the activity log. Data from 20 randomly sampled participants
(5 males, 15 females, Mage = 41.19 years, SD = 10.36; MBMI = 23.51 kg/m2, SD =
1.87) were analyzed. The average length of continuous zero sequences ≥20 min
was 31.35 min. Valid days were classified in a two-step process. In Step 1, weekday
and weekend day minutes per day (using the 70/80 rule while ignoring sequences
of ≥31 zero counts) were analyzed. If a day was valid according to these criteria,
no further analysis was required. In Step 2 (i.e., where total minutes/day was less
than the 70/80 rule criteria), activity log data and graphical accelerometer data
were screened to identify whether the day was invalid due to nonwear, or valid but
reflective of a largely sedentary day. A total of 83 additional days were classified
as valid at Step 2. It is suggested that 3–6 days of accelerometer data can provide
reliable estimates of physical activity (Trost et al. 2005). To balance sample size and
data quality, participants with ≥5 valid days of accelerometer data were included
in the final sample. Data were screened to quantify participation in MVPA and
bouts of MVPA. The threshold for moderate intensity activity (i.e., ≥3 metabolic
equivalents3) was set at ≥1952 counts∙min-1 (Freedson, Melanson, & Sirard, 1998).4
In line with public health physical activity recommendations (Physical Activity
Guidelines Advisory Committee, 2008) behavioral variables were quantified to
reflect (a) average minutes of MVPA per day, (b) average minutes of MVPA in
bouts ≥10 min per day and (c) number of days during monitoring on which ≥30
min of MVPA was accumulated in bouts of ≥10 min. We reasoned that the variable
comprising continuous bouts of MVPA were more reflective of purposeful exercise
In line with recommendations (Pettee, Storti, Ainsworth, & Kriska, 2009),
the variable reflecting average minutes per day of MVPA was adjusted to account
for moderate-to-vigorous activities that are not accurately quantified by the accel-
erometer (e.g., swimming, cycling, rowing and resistance exercise). Data (i.e.,
mode, duration and intensity) for activities reported in participants’ activity log as
moderate or vigorous were compared with Ainsworth’s compendium of physical
activities (Ainsworth et al., 2000) to confirm that the intensity was of a MET level
considered to be at least moderate. Visual inspection of graphical accelerometer
data was conducted to ensure that this activity was not measured by the Actigraph
and thus counted twice. A range of activities were reported in the activity logs
including walking (e.g., for exercise, transport, shopping, exercising pets and office
Goal Content and Physical Activity 183
based), structured exercise (e.g., walking, running, cycling, swimming and exercise
classes), organized sports (e.g., racket sports, football, hockey, rugby and golf) and
domestic activities (e.g., DIY, gardening and cleaning).
Preliminary Statistical Analyses. Mean replacement was used to replace missing
values (n = 2) and standardized skewness and kurtosis values were analyzed to
determine normality. The variables representing autonomous motivation, average
daily time spent in MVPA, adjusted average daily time spent in MVPA, and average
time spent in MVPA in bouts ≥10 min per day displayed non-normality. These
variables were transformed according to the nature of their individual distributions
(Tabachnick & Fidell, 2007). A reflect and square root transformation was applied
to autonomous motivation scores. For interpretation purposes this variable was
re-reflected following analysis (Tabachnick & Fidell, 2007). A square-root
transformation was applied to each of the behavioral variables.5 Finally, analysis
of residuals for assumptions pertaining to regression analyses (i.e., linearity,
homoscedasticity and independence and normality of residuals) revealed no
particular problems. Bivariate correlations were used to assess associations between
variables and t tests were employed to test mean differences between variables and
within variables between males and females.
Primary Statistical Analyses. A single-step multiple mediation model (Preacher
& Hayes, 2008b) was hypothesized and tested (Figure 1). A series of hierarchical
linear regression analyses were used to establish the value of the c, a1, a2, b1, b2
and c′ paths while controlling for participant gender, age and BMI/WC. When
calculating the values of the b1 and b2 paths, the effect of the other mediator on
the dependent variable was controlled (Preacher & Hayes, 2008a). Mediation of
the effect of relative intrinsic goal content on physical activity via motivation
was tested using the macro developed by Preacher and Hayes (2008a). This is an
add-on program that users can download (http://www.afhayes.com/spss-sas-and-
mplus-macros-and-code.html) and execute using SPSS, which allows advanced
mediation analyses to be conducted, such as the nonparametric testing of multiple
mediator models while controlling for covariates. Indirect effects of relative
intrinsic goal content on physical activity (i.e., the product of a1b1 and a2b2) were
analyzed to establish mediation. As the product of ab is normally distributed only
in large samples, it is recommended that bootstrapping (see Efron & Tibshirani,
1993; MacKinnon, 2007, for an overview specific to mediation analysis) is used
to construct confidence intervals around a point estimate of the indirect effect
(MacKinnon, Lockwood, & Williams, 2004; Preacher & Hayes, 2008a). Briefly,
bootstrapping is a resampling method in which statistics (in this case the mediated
effect) are calculated in multiple samples generated from the original sample.
When applied to mediation models, estimates from the multiple bootstrap samples
are used to produce a distribution of the mediated effect from which confidence
intervals can be obtained (MacKinnon, 2007). In the present work, we used
bootstrapping to construct bias-corrected and accelerated 95% confidence intervals
(BCa 95% CI) of the indirect effect. Five thousand bootstrapped samples with
replacement and of the same size as the original sample were requested (Preacher
& Hayes, 2008a).
184 Sebire, Standage, and Vansteenkiste
Seven valid days of accelerometer data were collected for 82 participants, 15 par-
ticipants provided six valid days and 4 participants provided five valid days. On
average, a valid day consisted of 834 min (13.90 hr) of data. Bivariate correlations
between the serial number of the ActiGraph worn with each behavioral variable
(Trost et al., 2005) revealed that interinstrument variability did not account for
variance in accelerometer-derived scores (i.e., all correlations were very small and
nonsignificant); therefore, ActiGraph unit was not specified as a covariate.
Table 1 presents the mean scores, standard deviations, range values and bivariate
correlations between the study variables. On average, the participants rated intrinsic
exercise goals (M = 4.14, SD = 0.89) as more important than extrinsic exercise
goals (M = 3.49, SD = 1.09) and endorsed autonomous motivation (M = 3.31, SD
= 0.65) more strongly than controlled motivation (M = 1.02, SD = 0.60). Accord-
ing to the unadjusted MVPA variable, participants completed approximately 59
min of MVPA per day. Adjustment of this value with data from the activity logs
increased average daily MVPA participation to approximately 68 min. The aver-
age minutes of unadjusted MVPA (M = 58.85, SD = 25.80) was significantly less
than the average minutes of activity-log-adjusted MVPA (M = 68.07, SD = 30.81)
t(100) = –6.86, p < .01, (Hedges’s g = –.32). As such, both variables were included
in further analyses. Despite accumulating about an hour of MVPA daily, on aver-
age, participants achieved approximately 25 min of MVPA in bouts ≥10 min and
achieved the guidelines of a daily accumulation of 30 min of MVPA accumulated
in bouts of ≥10 min on 2.56 days. T-tests revealed that females spent less time
than males in MVPA (M = 55.41 min, SD = 26.28 & M = 65.94 min, SD = 23.60
respectively) t(99) = –1.96, p < .05 (Hedges’s g = –.41) and diary-corrected MVPA
(M = 61.68 min, SD = 27.97 and M = 81.24 min, SD = 32.61 respectively) t(99) =
–3.12, p < .01, (Hedges’s g = –.66).
Figure 1 — Hypothesized multiple mediation model. Note. Symbols in parentheses indicate
the hypothesized direction of effects.
Table 1 Descriptive Statistics and Bivariate Correlations Among Study Variables
23.39 to 63.39
Relative intrinsic goals
−3.17 to 4.33
1.50 to 4.00
.00 to 2.67
Average daily time in MVPA
15.43 to 143.14
Adjusted daily time in MVPA
16.67 to 160.67
Average daily time in MVPA
in bouts >10 min (min)
.00 to 116.43
Number of days meeting ACSM/AHA guidelines
.00 to 7.00
Note. MVPA = moderate-to-vigorous physical activity. BMI/WC = body mass index/waist circumference. ACSM/AHA = American College of Sports Medicine/American
Heart Association. * p <.05, ** p <.01.
186 Sebire, Standage, and Vansteenkiste
Bivariate correlations revealed that participant age was positively correlated
with adjusted daily MVPA participation. Risk stratification based on BMI/WC
(Zhu et al., 2004) measures correlated negatively with MVPA accumulated in bouts
≥10 min and number of days on which participants met the ACSM/AHA physical
activity recommendations. Owing to these relationships, participant age, gender
and BMI/WC risk stratification were entered as covariates in further analyses.
With regards to the motivational variables, relative intrinsic goal content was
correlated in the expected directions with autonomous and controlled motivation
and was uncorrelated with the behavioral variables. Autonomous motivation dis-
Figure 2 — Models showing effects of goal content on MVPA criterion via autonomous
and controlled motivation. Note. Values are standardized estimates. The b paths (as in Figure
1) are estimated controlling for the other mediator. Dashed lines represent nonsignificant
estimates. * p < .05, ** p < .01. The ACSM/AHA guidelines are for 30 min of MVPA
accumulated in bouts of ≥10 min.
Goal Content and Physical Activity 187
played positive correlations, whereas controlled motivation was uncorrelated with
all behavioral variables.
Models A to D in Figure 2 display the results of the regression analyses. In all
models there was no direct relationship between relative intrinsic goal content and
the behavioral variable. Relative intrinsic goal content was positively associated
with autonomous motivation and negatively associated with controlled motivation.
Figure 2, continued.
188 Sebire, Standage, and Vansteenkiste
Autonomous motivation was positively associated with each of the behavioral
variables indicating its possible mediating role, whereas controlled motivation was
unrelated to the behavioral variables suggesting that it could not be a mediator.
Although a significant c path, that is, the direct association between goal content
and behavior, was not identified, it is recommended that this is not a requirement
of mediation (MacKinnon, 2007; Shrout & Bolger, 2002). Failing to find a sig-
nificant direct association between an independent and dependent variable may be
due to the variables being too distal from one another to be meaningfully related
(Shrout & Bolger, 2002). Given their distal relationship, it is likely that the effects
of the predictor (e.g., goal content) on the outcome variable (e.g., physical activity
behavior) is transmitted through a mediator which is located more proximally to
both the independent and dependent variables (e.g., motivation) and bridges the gap
between them. In this situation, it is recommended that analysis continue without
an initial direct effect.
Table 2 presents the total and specific indirect effects transmitted through
autonomous and controlled motivation and their BCa 95% CIs. In each model, the
total indirect effects were nonsignificant suggesting that the two proposed media-
tors did not transmit the effect of goal content to behavior. However, it is possible
for the specific indirect effects to be significant, when the total indirect effect is
nonsignificant, indicating mediation (Preacher & Hayes, 2008b; Shrout & Bolger,
2002). This was true of the present data as significant specific indirect effects indi-
cated that (controlling for controlled motivation) autonomous motivation mediated
the effect of relative intrinsic goal content on each behavioral dependent variable
(i.e., the BCa 95% CI of the indirect effect did not include zero). Pairwise contrasts
of the specific indirect effects indicated that the mediated effect via autonomous
motivation differed significantly from the mediated effect via controlled motivation.
The findings of a nonsignificant total indirect effect despite a significant indirect
effect being observed via autonomous motivation may be indicative of inconsistent
mediation (MacKinnon, Fairchild, & Fritz, 2007) and suppression (Preacher &
Hayes, 2008b). Multiple mediation models are described as inconsistent where
indirect effects have opposing signs. In the present data, the positive specific indirect
effect through autonomous motivation may have been cancelled out by the negative
(albeit nonsignificant) specific indirect effect through controlled motivation. In this
case autonomous motivation acted as a mediator whereas controlled motivation
acted as a suppressor (Preacher & Hayes, 2008b).6 In summary, our analysis of
direct and mediated (indirect) effects revealed evidence for a meditational model.
An indirect effect was observed in which the association between relative intrinsic
goal content and behavior was transmitted through autonomous motivation. Goal
content was not directly associated with any of the behavioral variables.
The purpose of the present work was to advance the existing literature pertaining
to the prediction of MVPA and bouts of MVPA in adults from the content and
motivation of their exercise goal pursuit. This was achieved through the use of
theoretically aligned measures of goal content and motivation (i.e., the GCEQ and
BREQ, respectively) combined with the objective assessment of physical activity
(i.e., ActiGraph accelerometry).
Goal Content and Physical Activity 189
Table 2 Unstandardized Indirect Effects of Relative Intrinsic Goal
Content on MVPA Criterion Through Autonomous and Controlled
BCa 95% Confidence Interval
Point estimate LowerUpper
Model A: Average minutes per day in MVPA
Total indirect effect
Autonomous motivation (a1b1)
Controlled motivation (a2b2)
Contrast (a1b1) vs. (a2b2)
Model B: Adjusted average minutes per day in MVPA
Total indirect effect
Autonomous motivation (a1b1)
Controlled motivation (a2b2)
Contrast (a1b1) vs. (a2b2)
Model C: Average minutes in MVPA in bouts ≥ 10 min
Total indirect effect
Autonomous motivation (a1b1)
Controlled motivation (a2b2)
Contrast (a1b1) vs. (a2b2)
Model D: Number of days on which ACSM/AHA guidelines were achieved
Total indirect effect
Autonomous motivation (a1b1)
Controlled motivation (a2b2)
Contrast (a1b1) vs. (a2b2)
Note. BCa = Bias corrected and accelerated. MVPA = moderate-to-vigorous physical activity. ACSM/
AHA guidelines are for 30 min of MVPA accumulated in bouts of ≥ 10 min.
Our preliminary analysis did not reveal a bivariate correlation between rela-
tive intrinsic goal content and MVPA or bouts of MVPA. This finding stands in
contrast to those of Ingledew and Markland (2008) and Sebire et al. (2009), who
identified small but significant associations between (a) health and fitness goals
and self-reported exercise behavior and (b) relative intrinsic goals and self-reported
exercise, respectively. Aligned with the tenets of SDT, and supporting past work
using subjective (e.g., Edmunds et al., 2006; Wilson et al., 2004) and objective
(e.g., Standage et al., 2008) behavioral measures, we observed a positive relation-
ship between autonomous motivation and MVPA and MVPA bouts. These findings
190 Sebire, Standage, and Vansteenkiste
provide support for the adaptive behavioral effects of being autonomous in one’s
motivation for exercise. The relationship between controlled motivational regula-
tion and behavior was consistently nonsignificant across the entire behavioral cri-
terion. The low internal consistency reliability coefficient of controlled motivation
replicates previous findings (Standage et al., 2008) and may partially explain the
observed lack of association between controlled motivation and physical activity.7
In line with previous findings (e.g., Kasser & Ryan 1993; Wilson et al., 2004),
participants endorsed intrinsic goals and autonomous motivational regulations
more strongly than extrinsic goals and controlled motivational regulations. This is
encouraging given the aforementioned evidence for the adaptive correlates of intrin-
sic goals and autonomous self-regulation. Greater endorsement of these constructs
could reflect the inherent trajectory toward growth posited in SDT (Deci & Ryan,
2000) or alternatively socially desirable responding given the more positive nature
of intrinsic goals and autonomous motivational regulations. Researchers may want
to develop implicit measures (i.e., tapping goal endorsement or motivation which
is not under conscious guidance or control) to examine whether intrinsic goals
and autonomous motivation are also highly endorsed at the implicit level and to
investigate whether these implicit measures can account for independent variance
in physical activity above and beyond the self-reported constructs.
Effects of “What” and “Why” on Physical Activity
Within SDT, it is hypothesized that goal content and motivation are associated
processes that can have distinct impact on outcomes (Deci & Ryan, 2000; Sheldon
et al., 2004). In the current study goal content and motivational regulations were
moderately correlated. Mediation analysis showed that after controlling for par-
ticipants’ age, gender and BMI/WC, goal content had a positive indirect effect on
the behavioral outcomes through autonomous motivation. We observed no direct
independent effect of goal content on behavior. These findings support observa-
tions made previously (e.g., Ingledew & Markland, 2008; Sebire et al., 2009) of
a motivational sequence, in which intrinsic goal content engenders autonomous
motivation that then positively predicts MVPA and bouts of MVPA.
A direct effect of goal content on behavior was not supported in our analyses
of physical activity outcomes. A critical interpretation of this finding might suggest
that goal-content does not play an important role in predicting physical activity or
that no evidence was obtained for GCT. However, the observation that goal contents
influences behavior only through the motivational types they engender does not
reduce the study of goal contents to a meaningless endeavor. Indeed, building on the
theoretical and empirical contributions of Kasser and Ryan (1993, 1996) coupled
with the accumulating evidence supporting the consideration of goal content in
numerous life domains, the current findings add support to the recent addition of
GCT as a fifth mini-theory within SDT (Ryan et al., 2009). That is, that the specific
content of goals is worthy of consideration, in the present case due to the impact
that people’s exercise goals appear to have on their motivational regulations. This
perspective aligns with the notion that exercise goals are more future oriented
and distally related to people’s current behavior than their more present-oriented
motivational regulations. The concept of basic need satisfaction within SDT allows
researchers to study goals from a quality, not just quantity, perspective as not all
Goal Content and Physical Activity 191
pursued or contextually promoted goals are equally conducive to need satisfaction.
Practically, many people have a specific goal in mind when starting to exercise and,
hence, it is interesting to gain insight in the processes (e.g., need satisfaction and
motivational regulations) that different goal contents engender.
The Role of Goal Content in Practice
Understanding the pathways through which a person’s goals may influence their
behavior or cognitive-affective experiences is important because such pathways
may represent avenues for intervention with individuals seeking to change their
behavior. The results of our analyses suggest that the content of an exerciser’s
goals may give a practitioner insight into their likely motivation, which will predict
their activity behavior. However, and while past work supports moderate relations
between goal content and motivation (Sheldon et al., 2004; Sebire, Standage, &
Vansteenkiste, 2008; 2009), practitioners should remember that it is theoretically
possible that an individual could pursue intrinsic goals with controlled motivation
and extrinsic goals with autonomous motivation (Sheldon et al., 2004). A central
objective of SDT-based interventions is to encourage the internalization of externally
prescribed motivation to more personally valued and self-endorsed motivation, so
that they become self-emanating and autonomously enacted (cf. Ryan et al., 2008).
The present findings suggest that the content of an exerciser’s goals might be an
important consideration if practitioners are to optimize the internalization process.
Limitations and Future Directions
As the majority of the participants in this study were White and well educated, the
associations explored in this work should also be examined in more racially and
educationally diverse samples. Although at least five days of accelerometer data
were collected from 101 participants in the current study, the results of this work
should be examined using larger samples. This said, while large samples would be
advantageous, the greater precision in measurement when using objective behav-
ioral assessment tools may go some way to overcome concerns related to having
sufficient power to detect relationships in moderate sized samples such as ours.
A limitation is that physical activity was assessed over a single seven-day period
only rather than on multiple occasions and therefore the analysis is cross-sectional.
Previous SDT-based research suggests that longitudinal studies of goals, motiva-
tion and behavior in the physical domain may be important to capture the dynamic
nature of motivation (cf. Standage & Ryan, in press). For example, Vansteenkiste,
Simons, Soenens et al. (2004) found that extrinsic goal framing was associated
with greater short-term but not continued behavioral persistence relative to a no-
goal control group but that the type of persistence displayed by individuals in the
extrinsic goal framing condition was not authentic in nature. This poor quality
exercise engagement presumably contributed to its lack of persistence over time.
Given the associations between goal content and motivation identified in the current
study, future longitudinal research using objective assessment of exercise behavior
would do well to explore the temporal dynamics of both of these motivational con-
structs on behavioral engagement. Future longitudinal studies would also do well
to attend to the cognitive-affective quality, in addition to the quantity of people’s
192 Sebire, Standage, and Vansteenkiste
physical activity / exercise behavior. It would be useful to extend the findings of
Sebire et al. (2009) to determine whether over time the quality (i.e., flexible vs.
rigid engagement, anxiety, and social comparison processes; see Vansteenkiste,
et al., 2008) rather than the quantity of physical activity / exercise engagement is
predicted by one’s exercise goals and motivation.
In the present work only one part of the more complex theoretical model
posited in SDT was considered. Although this study adds to a number of strands
of evidence that together begin to support the larger theoretical framework per-
taining to goal content it is important that future work strives to test the complete
model. In SDT one’s level of psychological need satisfaction is proposed to lead
to the pursuit of goals with different content as well as be a function of one’s goal
content. Accordingly, it would be insightful to examine theoretically proposed
causal structures leading from social-contextual factors (e.g., autonomy support)
through psychological need satisfaction to relative intrinsic/extrinsic goal content
and motivation types and subsequently to the consequences of such goal pursuits.
The incorporation of cognitive-attentional mediational mechanisms proposed in
previous work (e.g., Sebire, Standage, Gillison, & Vansteenkiste, 2008; Vansteen-
kiste, Soenens, & Lens, 2007) within these models would be particularly insightful.
Further, although we examined goal content as a predictor of motivational regu-
lations, it is possible that effects are reciprocal and that motivational regulations
could lead to the pursuit of certain goals.
The objective assessment of behavior in the current study provided a number
of advancements over the use of self-report methods. However, the treatment of
accelerometer data is based on a number of assumptions and is vulnerable to some
limitations. For example, when considering bouts of MVPA of ≥10 min, our software
only allowed the extraction of bouts with no interruptions. It is likely, however, that
prolonged exercise is intermittent (e.g., waiting to cross a busy road while jogging)
and that ignoring these instances underestimates total MVPA performed in bouts
and therefore more purposeful exercise. Future work could, aligned with the rec-
ommendations of Ward, Evenson, Vaughn, Rodgers, and Troiano (2005), employ
software that allows detection of activity bouts while allowing minor interruptions.
Biaxial accelerometry is likely to be most accurate in measuring ambulatory
activities such as walking and running (Dale et al., 2002). As such, the intensity of
participating in other activities such as weight lifting or cycling will not be measured
or underestimated. In the current study, we sought to address this limitation by
triangulating the accelerometer data with concurrently collected activity log data;
however, the use of more advanced technology may address this issue further. For
example, Standage et al. (2008) have recently used the Actiheart device (Cambridge
Neurotechnology Ltd), which bases the prediction of energy expenditure on con-
currently assessed acceleration and heart rate data. Such methodologies account
for activities that induce a high heart rate without significant vertical acceleration
(e.g., cycling, rowing, weight lifting). Despite the intuitive appeal of such methods,
as pointed out by Dale et al. (2002) when weighing up the use of accelerometry
or combined accelerometry and heart rate devices, researchers should consider
issues such as the device cost, participant burden and the component of behavior
that they are seeking to measure. A related limitation is that although we quantified
continuous bouts (≥10 min) of MVPA to attempt to reflect more purposeful exercise
behavior it is possible that these variables also capture some continuous activity
Goal Content and Physical Activity 193
which is not driven by one’s exercise cognitions (e.g., transport or employment
related). As the goal content and motivation questionnaires used in the current study
pertained to exercise behaviors there may be some mismatch between the motivation
constructs and certain behaviors measured. As this would also seemingly attenuate
the specified associations, future research using objectively assessed behavior may
explore this issue through the simultaneous assessment of motivation and goals for
exercise and physical activity.
In line with the sequential motivational model proposed by Ingledew and Markland
(2008) and supported by Sebire et al. (2009), the current study identified that relative
intrinsic goal content was positively associated with autonomous motivation, which
in turn positively predicted objectively assessed MVPA, bouts of MVPA and the
meeting of public health physical activity recommendations. Building on previous
findings that have shown relative intrinsic goal content to predict cognitive-affective
outcomes in exercise (Sebire, Standage, & Vansteenkiste, 2008; Sebire et al., 2009),
the results of the current study suggest that the content of exercisers’ goals is an
important consideration when exploring optimal motivation for exercise behavior.
1. The term physical activity encompasses all movement produced by skeletal muscles that
confers energy expenditure above rest (Caspersen, Powell, & Christenson, 1985). Exercise is a
subcomponent of physical activity that is more “planned, structured, repetitive, and purposive in
the sense that improvement or maintenance of one or more components of fitness is an objective”
(Caspersen et al., 1985, p. 128). In the present research we used the term physical activity to refer
to the behavior captured by the accelerometer that will encompass exercise behavior. The MVPA
measured in bouts of ≥10 min may be more reflective of purposeful exercise and less incidental
2. Integrated regulation, the most autonomous form of extrinsic motivation is also posited as
part of the self-determination continuum representing where identifications have been assessed
and aligned with one’s values, goals and needs (Deci & Ryan, 2000). The measure of motivation
(viz., BREQ; Mullan et al., 1997) employed in this study does not include a scale to assess this
construct and so it is not considered further in this article.
3. The quantity 1 metabolic equivalent (MET) is equal to resting energy expenditure (≈ 3.5
mL O2 kg–1·min–1).
4. The software used to analyze the accelerometer data produced output in blocks of 200
counts (i.e., 0–199, 200–299) and time spent in MVPA was extracted from the block closest to
the threshold derived by Freedson et al. (1998) (i.e., ≥2000 counts∙min–1).
5. Analyses using the nontransformed data yielded an identical pattern of findings.
6. The mediation analyses were replicated using individual intrinsic and extrinsic goal content
scores as predictor variables (Results available from first author on request). Similar to the pres-
ent findings in each model, intrinsic goal content displayed specific positive indirect effect on
the behavioral outcome via autonomous motivation. Regression analyses indicated that extrinsic
goal content was unrelated to autonomous motivation (which was then significantly related to
the behavioral outcomes) and positively related to controlled motivation (which was unrelated
to the behavioral outcomes). As such, for extrinsic goals, mediation was not possible and these
194 Sebire, Standage, and Vansteenkiste
analyses were not pursued further. These analyses suggest that the effects observed of relative
intrinsic goal content on behavior were carried by intrinsic rather than extrinsic goal scores.
7. Given the low reliability of the combined controlled motivation score, we analyzed a
mediation model specifying individual introjected and external motivational regulation scores as
mediators between relative intrinsic goal content and behavior. In the model specifying average
daily time in MVPA as the behavioral outcome, relative intrinsic goal content was significantly
negatively related to both introjected (β = –.38, p < .001) and external motivation (β = –.07, p <
.05). Introjected motivation (β = .25, p < .05) and external motivation (β = –.15, p > .05) were
not associated with behavior. These results were consistent across the three other behavioral
Ainsworth, B.E., Haskell, W.L., Whitt, M.C., Irwin, M.L., Swartz, A.M., Strath, S.J., et
al. (2000). Compendium of Physical Activities: an update of activity codes and MET
intensities. Medicine and Science in Sports and Exercise, 32, S498–S516.
American College of Sports Medicine. (2006). ACSM’s guidelines for exercise testing and
prescription (7th ed.). Baltimore, MD: Lippincott Williams & Wilkins.
Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure, process,
and content. Psychological Bulletin, 120, 338–375.
Brage, S., Brage, N., Wedderkopp, N., & Froberg, K. (2003). Reliability and validity of the
Computer and Science Applications accelerometer in a mechanical setting. Measure-
ment in Physical Education and Exercise Science, 7, 101–119.
Carver, C.S., & Baird, E. (1998). The American dream revisited: Is it what you want or why
you want it that matters? Psychological Science, 9, 289–292.
Caspersen, C.J., Powell, K.E., & Christenson, G.M. (1985). Physical activity, exercise, and
physical fitness: Definitions and distinctions for health-related research. Public Health
Reports, 100, 126–131.
Catellier, D.J., Hannan, P.J., Murray, D.M., Addy, C.L., Conway, T.L., Yang, S., et al. (2005).
Imputation of missing data when measuring physical activity by accelerometry. Medi-
cine and Science in Sports and Exercise, 37, S555–S562.
Crawford, S., & Eklund, R.C. (1994). Social physique anxiety, reasons for exercise, and
attitudes toward exercise settings. Journal of Sport & Exercise Psychology, 16, 70–82.
Dale, D., Welk, G.J., & Matthews, C.E. (2002). Methods for assessing physical activity and
challenges for research. In G.J. Welk (Ed.), Physical activity assessments for health-
related research (pp. 19–34). Champaign, IL: Human Kinetics.
Deci, E.L., & Ryan, R.M. (2000). The “what” and “why” of goal pursuits: Human needs and
the self-determination of behavior. Psychological Inquiry, 11, 227–268.
Duriez, B., Vansteenkiste, M., Soenens, B., & De Witte, H. (2007). The social costs of
extrinsic relative to intrinsic goal pursuits: Their relation with social dominance and
racial and ethnic prejudice. Journal of Personality, 75, 757–782.
Edmunds, J., Ntoumanis, N., & Duda, J.L. (2006). A test of self-determination theory in the
exercise domain. Journal of Applied Social Psychology, 36, 2240–2265.
Efron, B., & Tibshirani, R. (1993). An introduction to the bootstrap. New York: Chapman
Freedson, P.S., Melanson, E., & Sirard, J. (1998). Calibration of the Computer Science and
Applications, Inc. accelerometer. Medicine and Science in Sports and Exercise, 30,
Gardarsdottir, R.B., Dittmar, H., & Aspinall, C. (2009). It’s not the money, it’s the quest
for a happier life: The role of happiness and success motives in the link between
financial goals and subjective well-being. Journal of Social and Clinical Psychology,
Goal Content and Physical Activity 195
Gillison, F., Standage, M., & Skevington, S.M. (2006). Relationships among adolescents’
weight perceptions, exercise goals, exercise motivation, quality of life and leisure-
time exercise behaviour: A self-determination theory approach. Health Education
Research, 21, 836–847.
Guttman, L. (1954). A new approach to factor analysis: The radex. In P. Lazarfel (Ed.),
Mathematical thinking in the social sciences (pp. 258–348). NY: Free Press of Glencoe.
Hardman, A.E., & Stensel, D.J. (2003). Physical activity and health: The evidence explained.
Ingledew, D.K., & Markland, D. (2008). The role of motives in exercise participation.
Psychology & Health, 23, 807–828.
Kasser, T. (2002). The high price of materialism. Cambridge, MA: MIT Press.
Kasser, T., Cohn, S., Kanner, A.D., & Ryan, R.M. (2007). Some costs of American corporate
capitalism: A psychological exploration of value and goal conflicts. Psychological
Inquiry, 18, 1–22.
Kasser, T., & Ryan, R.M. (1993). A dark side of the American dream: Correlates of financial
success as a central life aspiration. Journal of Personality and Social Psychology,
Kasser, T., & Ryan, R.M. (1996). Further examining the American dream: Differential
correlates of intrinsic and extrinsic goals. Personality and Social Psychology Bul-
letin, 22, 280–287.
Koestner, R., Otis, N., Powers, T.A., Pelletier, L., & Gagnon, H. (2008). Autonomous
motivation, controlled motivation, and goal progress. Journal of Personality, 76,
Loney, T., Standage, M., Thompson, D., Sebire, S.J., & Cumming, S.P. (2011).Self-report
vs. objectively assessed physical activity: Which is right for public health? Journal of
Physical Activity and Health. 8, 62–70.
MacKinnon, D.P. (2007). Introduction to mediation analysis. NY: Lawrence Erlbaum
MacKinnon, D.P., Fairchild, A.J., & Fritz, M.S. (2007). Mediation analysis. Annual Review
of Psychology, 58, 593–614.
MacKinnon, D.P., Lockwood, C.M., & Williams, J. (2004). Confidence limits for the indirect
effect: Distribution of the product and resampling methods. Multivariate Behavioral
Research, 39, 99–128.
Maltby, J., & Day, L. (2001). The relationship between exercise motives and psychological
well-being. The Journal of Psychology, 135, 651–660.
McHoskey, J.W. (1999). Machiavellianism, intrinsic versus extrinsic goals, and social
interest: A self-determination theory analysis. Motivation and Emotion, 23, 267–283.
Mullan, E., Markland, D., & Ingledew, D.K. (1997). A graded conceptualisation of self-
determination in the regulation of exercise behaviour: Development of a measure using
confirmatory factor analysis. Personality and Individual Differences, 23, 745–752.
Niemiec, C.P., Ryan, R.M., Deci, E.L., & Williams, G.C. (2009). Aspiring to physical
health: The role of aspirations for physical health in facilitating long-term tobacco
abstinence. Patient Education and Counseling, 74, 250–257.
Ntoumanis, N., & Standage, M. (2009). Morality in sport: A self-determination theory
perspective. Journal of Applied Sport Psychology, 21, 365–380.
Pettee, K.K., Storti, K.L., Ainsworth, B.E., & Kriska, A.M. (2009). Measurement of physical
activity and inactivity in epidemiologic studies. In I.M. Lee (Ed.), Epidemiological
methods in physical activity studies (pp. 15–33). Oxford: Oxford University Press.
Physical Activity Guidelines Advisory Committee. (2008). Physical Activity Guidelines
Advisory Committee Report. Washington, DC: U.S. Department of Health and Human
Plasqui, G., & Westerterp, K.R. (2007). Physical activity assessment with accelerometers: An
evaluation against doubly labeled water. Obesity (Silver Spring, Md.), 15, 2371–2379.
196 Sebire, Standage, and Vansteenkiste
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., & Podsakoff, N.P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended
remedies. The Journal of Applied Psychology, 88, 879–903.
Preacher, K.J., & Hayes, A.F. (2008a). Asymptotic and resampling strategies for assess-
ing and comparing indirect effects in multiple mediator models. Behavior Research
Methods, 40, 879–891.
Preacher, K.J., & Hayes, A.F. (2008b). Contemporary approaches to assessing mediation in
communication research. In A.F. Hayes, M.D. Slater, & L.B. Snyder (Eds.), The Sage
sourcebook of advanced data analysis methods in communications research (pp.
13–54). Thousand Oaks, CA: Sage.
Ryan, R.M., & Deci, E.L. (2007). Active human nature: Self-determination theory and the
promotion and maintenance of sport, exercise and health. In M.S. Hagger & N.L.D.
Chatzisarantis (Eds.), Intrinsic motivation and self-determination in exercise and
sport (pp. 1–19). Champaign, IL: Human Kinetics.
Ryan, R.M., Patrick, H., Deci, E.L., & Williams, G.C. (2008). Facilitating health behavior
change and its maintenance: Interventions based on Self-Determination Theory. The
European Health Psychologist, 10, 2–5.
Ryan, R.M., Williams, G.C., Patrick, H., & Deci, E.L. (2009). Self-determination theory and
physical activity: The dynamics of motivation in development and wellness. Hellenic
Journal of Psychology, 6, 107–124.
Sebire, S.J., Standage, M., Gillison, F., & Vansteenkiste, M. (2008). “Coveting thy neighbor’s
legs”: Qualitative analysis of the experiences of exercisers pursuing relative intrin-
sic and extrinsic goals. Poster session at the Conference of the British Psychological
Association Division of Sport and Exercise Psychology: London.
Sebire, S.J., Standage, M., & Vansteenkiste, M. (2008). Development and validation of the
Goal Content for Exercise Questionnaire. Journal of Sport & Exercise Psychology,
Sebire, S.J., Standage, M., & Vansteenkiste, M. (2009). Examining intrinsic versus extrinsic
exercise goals: Cognitive, affective, and behavioral outcomes. Journal of Sport &
Exercise Psychology, 31, 189–210.
Sheldon, K.M., & McGregor, H.A. (2000). Extrinsic value orientation and “The tragedy of
the commons”. Journal of Personality, 68, 383–411.
Sheldon, K.M., Ryan, R.M., Deci, E.L., & Kasser, T. (2004). The independent effects of goal
contents and motives on well-being: It’s both what you pursue and why you pursue it.
Personality and Social Psychology Bulletin, 30, 475–486.
Shrout, P.E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:
New procedures and recommendations. Psychological Methods, 7, 422–445.
Srivastava, A., Locke, E.A., & Bartol, K.M. (2001). Money and subjective well-being: It’s not
the money, it’s the motives. Journal of Personality and Social Psychology, 80, 959–971.
Standage, M., & Ryan, R.M. Self-determination theory and exercise motivation: Facilitat-
ing self-regulatory processes to support and maintain health and well-being (in press).
In G.C. Roberts & D.C. Treasure (Eds.), Motivation in sport and exercise (Vol. 3).
Champaign, IL: Human Kinetics.
Standage, M., Sebire, S.J., & Loney, T. (2008). Does exercise motivation predict engagement
in objectively-assessed bouts of moderate-intensity exercise behavior? A self-deter-
mination theory perspective. Journal of Sport & Exercise Psychology, 30, 337–352.
Tabachnick, B.G., & Fidell, L.S. (2007). Using multivariate statistics (5th ed.). USA:
Thøgersen-Ntoumani, C., Ntoumanis, N., & Nikitaras, N. (2010). Unhealthy weight control
behaviours in adolescent girls: A process model based on self-determination theory.
Psychology & Health, 25, 535–550.
Goal Content and Physical Activity 197 Download full-text
Trost, S.G., McIver, K.L., & Pate, R.R. (2005). Conducting accelerometer-based activity
assessments in field-based research. Medicine and Science in Sports and Exercise,
Vansteenkiste, M., Lens, W., & Deci, E.L. (2006). Intrinsic versus extrinsic goal contents
in self-determination theory: Another look at the quality of academic motivation.
Educational Psychologist, 41, 19–31.
Vansteenkiste, M., Niemiec, C., & Soenens, B. (2010). The development of the five mini-
theories of self-determination theory: An historical overview, emerging trends, and
future directions. In T. Urdan & S. Karabenick (Eds.). Advances in motivation and
achievement, vol. 16: The decade ahead. UK: Emerald Publishing.
Vansteenkiste, M., Simons, J., Lens, W., Sheldon, K.M., & Deci, E.L. (2004). Motivating
learning, performance, and persistence: The synergistic effects of intrinsic goal contents
and autonomy-supportive contexts. Journal of Personality and Social Psychology,
Vansteenkiste, M., Simons, J., Soenens, B., & Lens, W. (2004). How to become a persevering
exerciser? The importance of providing a clear, future intrinsic goal in an autonomy-
supportive manner. Journal of Sport & Exercise Psychology, 26, 232–249.
Vansteenkiste, M., Soenens, B., & Duriez, B. (2008). Presenting a positive alternative to
materialistic strivings and the thin-ideal: Understanding the effects of extrinsic rela-
tive to intrinsic goal pursuits. In S.J. Lopez (Ed.), Positive psychology: Exploring the
best in people (Vol. 4, pp. 57–86). Westport, CT: Greenwood Publishing Company.
Vansteenkiste, M., Soenens, B., & Lens, W. (2007). Intrinsic versus extrinsic goal promotion
in exercise and sport: Understanding the differential impacts on performance and per-
sistence. In M.S. Hagger & N.L.D. Chatzisarantis (Eds.), Intrinsic motivation and self-
determination in exercise and sport (pp. 167–180). Champaign, IL: Human Kinetics.
Ward, D.S., Evenson, K.R., Vaughn, A., Rodgers, A.B., & Troiano, R.P. (2005). Accelerom-
eter use in physical activity: Best practices and research recommendations. Medicine
and Science in Sports and Exercise, 37, S582–S588.
Williams, G.C., Cox, E.M., Hedberg, V.A., & Deci, E.L. (2000). Extrinsic life goals and
health-risk behaviors in adolescents. Journal of Applied Social Psychology, 30,
Wilson, P.M., Rodgers, W.M., Blanchard, C.M., & Gessell, J. (2003). The relationship
between psychological needs, self-determined motivation, exercise attitudes, and physi-
cal fitness. Journal of Applied Social Psychology, 33, 2373–2392.
Wilson, P.M., Rodgers, W.M., Fraser, S.N., & Murray, T.C. (2004). Relationships between
exercise regulations and motivational consequences in university students. Research
Quarterly for Exercise and Sport, 75, 81–91.
World Health Organization. (2002). The world health report 2002. Reducing risks, promot-
ing healthy life. Geneva: World Health Organization.
Zhu, S., Heshka, S., Wang, Z., Shen, W., Allison, D.B., Ross, R., et al. (2004). Combination
of BMI and waist circumference for identifying cardiovascular risk factors in whites.
Obesity Research, 12, 633–645.
Manuscript submitted: March 15, 2010
Revision accepted: November 11, 2010