ArticlePDF Available

Can You Have Your Vigorous Exercise and Enjoy It Too? Ramping Intensity Down Increases Postexercise, Remembered, and Forecasted Pleasure


Abstract and Figures

There is a paucity of methods for improving the affective experience of exercise. We tested a novel method based on discoveries about the relation between exercise intensity and pleasure, and lessons from behavioral economics. We examined the effect of reversing the slope of pleasure during exercise from negative to positive on pleasure and enjoyment, remembered pleasure, and forecasted pleasure. Forty-six adults were randomly assigned to a 15-min bout of recumbent cycling of either increasing intensity (0%-120% of Watts corresponding to the ventilatory threshold) or decreasing intensity (120%-0%). Ramping intensity down, thereby eliciting a positive slope of pleasure during exercise, improved postexercise pleasure and enjoyment, remembered pleasure, and forecasted pleasure. The slope of pleasure accounted for 35%-46% of the variance in remembered and forecasted pleasure from 15 min to 7 days postexercise. Ramping intensity down makes it possible to combine exposure to vigorous and moderate intensities with a pleasant affective experience.
Content may be subject to copyright.
Journal of Sport & Exercise Psychology, 2016, 38, 149 -159
© 2016 Human Kinetics, Inc.
Zachary Zenko and Panteleimon Ekkekakis are with the
Department of Kinesiology, Iowa State University, Ames, IA.
Dan Ariely is with Duke University, Durham, NC. Address
author correspondence to Panteleimon Ekkekakis at ekkekaki@
Can You Have Your Vigorous Exercise and Enjoy It
Too? Ramping Intensity Down Increases Postexercise,
Remembered, and Forecasted Pleasure
Zachary Zenko,1 Panteleimon Ekkekakis,1 and Dan Ariely2
1Iowa State University; 2Duke University
There is a paucity of methods for improving the affective experience of exercise. We tested a novel method
based on discoveries about the relation between exercise intensity and pleasure, and lessons from behavioral
economics. We examined the effect of reversing the slope of pleasure during exercise from negative to positive
on pleasure and enjoyment, remembered pleasure, and forecasted pleasure. Forty-six adults were randomly
assigned to a 15-min bout of recumbent cycling of either increasing intensity (0–120% of watts corresponding
to the ventilatory threshold) or decreasing intensity (120–0%). Ramping intensity down, thereby eliciting a
positive slope of pleasure during exercise, improved postexercise pleasure and enjoyment, remembered plea-
sure, and forecasted pleasure. The slope of pleasure accounted for 35–46% of the variance in remembered and
forecasted pleasure from 15 min to 7 days postexercise. Ramping intensity down makes it possible to combine
exposure to vigorous and moderate intensities with a pleasant affective experience.
Keywords: affective forecasting, behavioral economics, remembered utility, predicted utility
Most theories of behavior change used in the domain
of public health rest on the assumption that, once provided
with appropriate, accurate, and adequate information,
most individuals will act to alter their behavior. When
judged from the standpoint of such theories, the case of
exercise appears paradoxical. While nearly everyone in
western societies reports awareness of the health benets
of exercise (Martin, Morrow, Jackson, & Dunn, 2000;
O’Donovan & Shave, 2007), the rates of participation
are extremely low. A nationwide study in the United
States, using objective assessment of physical activity
with accelerometers, showed that only 3.2% of adults
are active at levels recommended for health promotion
(Tudor-Locke, Brashear, Johnson, & Katzmarzyk, 2010).
Moreover, dropout from short-term interventions aver-
ages 45% (Marcus et al., 2006) and is even higher in eld
settings (Edmunds, Ntoumanis, & Duda, 2007).
Interventions designed to increase exercise and
physical activity (PA) have mainly targeted cognitive
constructs, such as appraisals of self-efcacy, outcome
expectations, perceptions of social support, and antici-
pated benets versus costs. The results of these inter-
ventions so far have been modest (Marcus et al., 2006).
Thus, it seems reasonable to suggest that an expansion
of the theoretical perspective through which the problem
of activity promotion is approached may be warranted.
In particular, the theories that are commonly used
to explain, predict, and change PA and exercise behavior
rely on the assumption that, in making behavioral deci-
sions, people act as rational decision makers: they seek,
collect, and analyze relevant information; methodically
weigh pros and cons; and make probabilistic predictions
about the future consequences of their actions or inac-
tions. However, research from behavioral economics
has cast doubt on the assumption that decision making
is based solely on the rational evaluation of information
(Stanovich & West, 2000). Explaining the notion of
bounded rationality, Simon (1983) argued that “human
beings have neither the facts nor the consistent struc-
ture of values nor the reasoning power at their disposal
that would be required [to behave rationally]” (p. 17).
Building on this idea, Kahneman (2003) proposed that,
rather than always relying on rationality, humans use a
set of heuristics and biases, which, although occasion-
ally disadvantageous, bring the complexity of problems
down to a manageable scale and help people navigate
their world. One such device, called the affect heuristic
(e.g., Finucane, Alhakami, Slovic, & Johnson, 2000),
has been singled out as “probably the most important
development in the study of judgment heuristics in the
past few decades” (Kahneman, 2003, p. 710). The simple
but powerful idea behind the affect heuristic is that judg-
ments and decisions are inuenced by affective responses
(Finucane et al., 2000).
150 Zenko, Ekkekakis, and Ariely
JSEP Vol. 38, No. 2, 2016
Prompted by the desire to uncover sources of behav-
ioral variation not captured by cognitive appraisals,
researchers have started focusing on the role of affective
constructs, such as pleasure and enjoyment. Despite
considerable heterogeneity in their methodologies, early
studies have found positive associations between affective
responses during exercise bouts and subsequent PA (for
reviews, see Ekkekakis & Dafermos, 2012; Rhodes &
Kates, 2015). Likewise, a meta-analysis on the relation-
ship of enjoyment and related variables (e.g., affective
component of attitude, intrinsic motivation) with PA
found an average correlation of 0.42, which is larger than
those for self-efcacy, social and sociodemographic vari-
ables, personality factors, and attributes of the built envi-
ronment (Rhodes, Fiala, & Conner, 2009). Other studies
have shown that affective associations (e.g., “When I
think of exercise, I feel . . .”; Kiviniemi, Voss-Humke, &
Seifert, 2007) and anticipated affective responses (e.g.,
“I will feel regret if I do not exercise over the next four
weeks”; Conner, McEachan, Taylor, O’Hara, & Lawton,
2015; Dunton & Vaughan, 2008) are also signicantly
associated with PA participation.
This preliminary evidence demonstrates that target-
ing affective constructs in interventions aimed at promot-
ing PA may hold promise. However, there is presently
a surprising dearth of information on how to make PA
and exercise more pleasant. Many researchers consider
reduced pleasure and enjoyment during the early stages
of exercise interventions as more-or-less unavoidable. For
example, Wilson, Rodgers, Blanchard, and Gessell (2003)
wrote that, at the initial stages, exercise is “unlikely to
be construed as inherently pleasurable or enjoyable” (p.
2375). Indeed, research with chronically sedentary and/or
low-tness participants shows declines in pleasure over
most of the range of exercise intensity (e.g., Ekkekakis,
Lind, & Vazou, 2010; Sheppard & Partt, 2008; Welch,
Hulley, Ferguson, & Beauchamp, 2007). Likewise,
interventions with formerly sedentary adults have found
reductions in enjoyment (Castro, Sallis, Hickmann, Lee,
& Chen, 1999; Stevens, Lemmink, van Heuvelen, de
Jong, & Rispens, 2003).
The present study was designed to examine the
effect of manipulating the slope of pleasure–displeasure
during an exercise bout on how pleasant or unpleasant
the bout is later remembered (“remembered utility” in
behavioral-economic terms) and how pleasant or unpleas-
ant future bouts are expected to be (“predicted utility”
in behavioral-economic terms, typically referred to as
affective forecasting in psychology). In behavioral eco-
nomics, it has been theorized that both the remembered
utility and the predicted utility of an experience predict
whether a behavior will be repeated (Ariely & Carmon,
2000; Kahneman, Wakker, & Sarin, 1997). Accumulating
evidence indicates that anticipated affect and enjoyment
are indeed associated with PA intentions and behavior
(Conner et al., 2015; Dunton & Vaughan, 2008; Helfer,
Elhai, & Geers, 2015; Loehr & Baldwin, 2014).
It is generally assumed that predictions of how
pleasant or unpleasant an experience will feel depends on
how similar past experiences have registered in memory.
From the standpoint of intervention, the question, then,
is how the affective memories of and the affective fore-
casts for exercise can be improved. Intuition perhaps
suggests that the entire exercise session should be made
more pleasant. However, research shows that not all
aspects of an episode are equally inuential in shaping
the memory of that episode. Evidence from behavioral
economics has shown that people prefer experiences
during which pleasure increases over time to those that
involve decreasing pleasure, even if the total amount of
derived pleasure is the same (Ariely & Carmon, 2000;
Ariely & Zauberman, 2003; Zauberman, Diehl, & Ariely,
2006). This nding is relevant to exercise since, among
individuals with a low level of cardiorespiratory tness,
even relatively low workloads can result in the inability
to maintain a physiological “steady state.” In such cases,
physiological variables associated with metabolic strain,
including heart rate, oxygen uptake, and blood lactate,
exhibit a continuous upward “drift.” In turn, this trend is
associated with declining levels of pleasure (Ekkekakis,
Partt, & Petruzzello, 2011) and the rising desire to stop.
Thus, exercise bouts with nonathletic participants typi-
cally culminate with the highest level of physiological
strain and perceived exertion, as well as the lowest level
of pleasure (e.g., Lind, Joens-Matre, & Ekkekakis, 2005).
Thus, the purpose of the current study was to explore
the psychological implications of changing the pattern of
exercise intensity during a bout, from the typical increas-
ing slope to a decreasing slope. In a between-subject
design, we compared two bouts of similar physiological
demands but opposite intensity slopes, one increasing
(intended as a simulation of a typical bout) and the other
decreasing. Based on evidence from exercise-psycho-
logical research (Ekkekakis et al., 2011), we expected
that this manipulation would result in opposite slopes
of pleasure ratings, with continuous during-exercise
decline in the increasing-intensity group but continuous
improvement in the decreasing-intensity group. In turn,
based on evidence from behavioral economics (Ariely
& Carmon, 2000), we hypothesized that participants
in the decreasing-intensity (increasing-pleasure) group
would report more postexercise pleasure and enjoyment,
would remember the exercise bout as having been more
pleasant, and would forecast that a future bout would be
more pleasant than participants in the increasing-intensity
(decreasing-pleasure) group.
Power calculations for a between-within interaction in
a 2 (groups) by 5 (time points) design, anticipating a
“small” to “medium” effect (f = 0.15), α = .05, 1 – β
= 0.80, correlated dependent variables (r = .70), and a
violation of the assumption of sphericity (ε = 0.70) indi-
cated a required total sample size of 44. Participants were
deemed eligible if they (a) were between the ages of 18
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
Exercise as Affective Experience 151
JSEP Vol. 38, No. 2, 2016
and 40 years, (b) were not pregnant, (c) had no history
of cardiometabolic disease, (d) experienced no pain or
dizziness during exercise, (e) did not use supplemental
oxygen for breathing, and (f) had no metal allergies
or implanted electromagnetic devices. No restriction
was placed on gender or habitual PA. We opted for a
heterogeneous sample and thus a more stringent test of
our hypotheses (i.e., under conditions of unrestricted
within-group variance).
Following Institutional Review Board approval, 54
members of a university community (21 women and 33
men, mainly staff members) responded to e-mail solicita-
tions and satised the inclusion/exclusion criteria. After
the initial exercise test (as described below), the ventila-
tory threshold could not be determined unambiguously
in seven cases, so these individuals were not scheduled
for additional sessions. One additional individual suffered
an injury unrelated to the study. Thus, 46 individuals
(15 women, 31 men) completed all sessions and were
included in the analyses. Of them, 6 women and 16 men (n
= 22) were randomly allocated to the increasing-intensity
group, whereas 9 women and 15 men (n = 24) were
allocated to the decreasing-intensity group. Additional
characteristics are presented in Table 1.
Pleasure during and after exercise. The core affective
dimension of valence is conceptualized as bipolar, rang-
ing from pleasure to displeasure (Russell, 1980). It was
assessed with the Feeling Scale (FS; Hardy & Rejeski,
1989), a single-item, 11-point bipolar rating scale rang-
ing from +5 (I feel very good) to –5 (I feel very bad)
and verbal anchors at zero (neutral) and odd numbers.
Concurrent validity data have been reported by Hardy and
Rejeski (1989). The use of a single-item rating scale was
deemed necessary, as it allowed the collection of data with
adequate temporal resolution during the exercise bouts,
while minimizing respondent burden.
Remembered pleasure. To minimize common-method
variance, remembered pleasure–displeasure was assessed
using a scale with a different format than the FS. Speci-
cally, a visual analog scale (VAS) was used, in response
to the question, “How did the exercise session in the
laboratory make you feel?” The scale ranged from very
pleasant (+100) to very unpleasant (–100) in intervals of
1. Participants responded using a computer by moving an
on-screen slider. The slider was initially positioned in the
middle (0). The verbal descriptors and slider were visible
to participants but the numbers were not.
Forecasted pleasure. Again, to minimize common-
method variance, forecasted pleasure was assessed using
a scale with a different format than the scales used during
exercise (i.e., FS) and for the assessment of remembered
pleasure (i.e., VAS). Specically, participants responded
to the question, “If you repeated the exercise session
again, how do you think it would make you feel?” using
the Empirical Valence Scale (EVS; Lishner, Cooter, &
Zald, 2008) presented on a computer screen. Respondents
chose from 15 empirically spaced verbal anchors, rang-
ing from most unpleasant imaginable (–100) to most
pleasant imaginable (+100). The value associated with
each verbal anchor corresponded to the values specied
by Lishner et al. (2008).
Postexercise enjoyment. Enjoyment of the exercise
session was measured with the Physical Activity Enjoy-
ment Scale (PACES; Kendzierski & DeCarlo, 1991). In
accordance with the standard instructions, respondents
were asked to “rate how you feel at the moment about
the physical activity you have been doing.” The PACES
Table 1 Participant Characteristics (M ± SD)
Intensity-Pattern Group
Increasing Intensity
= 22) Decreasing Intensity
= 24)
Men/Women 16/6 15/9
Age (years) 28 ± 5 27 ± 4
Height (cm), Men 176 ± 6 178 ± 6
Height (cm), Women 164 ± 8 167 ± 6
Body Mass (kg), Men 82 ± 17 78 ± 9
Body Mass (kg), Women 69 ± 13 67 ± 16
Body Mass Index (kg·m–2), Men 26 ± 5 25 ± 2
Body Mass Index (kg·m–2), Women 25 ± 3 24 ± 4
Planned Exercise (min·week–1) 328 ± 247 356 ± 245
Incidental PA (min·week–1) 790 ± 739 707 ± 591
VO2peak (ml·kg–1·min–1), Men 27 ± 8 32 ± 5
VO2peak (ml·kg–1·min–1), Women 29 ± 12 26 ± 4
VT (%VO2peak) 62 ± 11 56 ± 9
Note. PA = physical activity; VT = ventilatory threshold. For all group comparisons, p > .05.
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
152 Zenko, Ekkekakis, and Ariely
JSEP Vol. 38, No. 2, 2016
consists of 18 bipolar items (e.g., I enjoy it versus I
hate it), with the polar opposites separated by a 7-point
scale (“4” being the midpoint). Validation studies by
Kendzierski and DeCarlo (1991) have shown a negative
correlation with boredom and a signicant prediction
of choice between different activities. In the present
sample, the PACES exhibited high internal consistency
(Cronbach’s α = .95).
Perceived exertion. As a manipulation check, per-
ceptions of exertion were assessed with the Rating of
Perceived Exertion (RPE; Borg, 1998), which ranges
from 6 (No exertion at all) to 20 (Maximal exertion).
The validity of this scale has been established through
correlations with physiological indices, including venti-
lation, oxygen update, and lactate accumulation (Chen,
Fan, & Moe, 2002).
Physical activity. For descriptive purposes, the Inci-
dental and Planned Exercise Questionnaire (IPEQ-WA;
Delbaere, Hauer, & Lord, 2010) was used as a measure
of habitual PA. The IPEQ-WA includes questions inquir-
ing about both planned exercise (e.g., exercise classes)
and weekly incidental PA behavior (e.g., walking for
Using a computer algorithm, participants were randomly
assigned to either an increasing-intensity or a decreasing-
intensity group. Participation consisted of three visits to
the laboratory: (a) an initial session for maximal exercise
testing, (b) an experimental exercise session, and (c) a
follow-up outcome-assessment session. The three visits
were separated by 1 week and were scheduled at the same
time of day for each participant, to control for possible
diurnal variation.
Session 1: Maximal exercise test. The purposes of
this session were to (a) collect anthropometric data; (b)
determine peak aerobic capacity and ventilatory thresh-
old, used in setting the workload for the subsequent
experimental session; and (c) familiarize participants with
the self-report measures. After completing the informed
consent process and the IPEQ-WA, participants had their
height (wall-mounted stadiometer) and weight (BF-626,
Tanita, Tokyo, Japan) measured. Participants were then
tted with a heart rate monitor (Polar, Kempele, Finland)
and a nose-and-mouth facemask for the collection of
expired gases (Hans Rudolph, Kansas City, MO) and had
the standard instructions for FS and RPE read to them.
Testing was conducted with a computer-controlled,
electronically braked recumbent cycle ergometer (Corival
Recumbent, Lode BV, Groningen, Netherlands). Oxygen
uptake and carbon dioxide production were measured
with a metabolic cart (TrueOne 2400, ParvoMedics, Salt
Lake City, UT), which was calibrated before each use.
After a 5-min warm-up (0 W), the workload increased in
a ramp fashion (1 W every 4 s). To gain experience with
the FS and RPE, participants reported their responses
to these scales each minute during the test by pointing
on laminated poster-size versions of the scales (kept
out of the eld of vision at other times). Upon reaching
volitional exhaustion, the facemask was removed and
participants did a 5-min cool-down (0 W), followed by a
5-min rest period. Thereafter, the participants responded
to the VAS and EVS, as an opportunity to familiarize
themselves with these scales as well.
The ventilatory threshold was later determined by
consensus by two judges who worked independently,
analyzing the gas exchange data ofine with the aid of
a software program (WinBreak 3.7, Epistemic Mind-
works, Ames, IA). The software combines three methods
(V-slope, ventilatory equivalents, excess CO2), as rec-
ommended by Gaskill, Ruby, Walker, Sanchez, Serfass,
and Leon (2001). The ventilatory threshold could not
be determined unambiguously (due to excessive noise
in the gas exchange data) for seven individuals. Rather
than subjecting these individuals to additional testing,
they were excluded from the rest of the study.
Session 2: Experimental session. Participants returned
1 week later for the experimental session. Preparatory
procedures identical to those of Session 1 were followed
(i.e., tting of heart rate monitor, rereading of instruc-
tions for FS and RPE), but no facemask was used (and
no expired gases were collected), to enhance external
validity. The exercise consisted of recumbent station-
ary cycling on the same ergometer used for maximal
testing. The increasing-intensity group started at 0 W
and progressed to 120% of the watts corresponding to
the ventilatory threshold over 15 min. Conversely, the
decreasing-intensity group started at 120% of the watts
corresponding to the ventilatory threshold and progressed
to 0 W over 15 min. The workload and time passed were
not visible to participants. Participants responded to the
FS immediately before exercise; during the last 15 s of
Minutes 3, 6, 9, 12, and 15 of the 15-min bout; and 2, 5,
and 10 min after exercise. The RPE data were collected
at the during-exercise time points.
After exercise, participants rested in a recliner.
Fifteen minutes postexercise, after being left alone in
the room, participants used a computer to respond to the
VAS and EVS, for the assessment of remembered and
forecasted pleasure, respectively. Next, participants com-
pleted the PACES. Finally, participants were informed
that they would receive a follow-up e-mail message, 24
hr later, with an Internet link, for one more administration
of the VAS and EVS.
Session 3: Follow-up outcome assessment. Partici-
pants returned to the laboratory 1 week later, to respond
to the VAS and EVS. They were then thanked, debriefed,
and released.
Statistical Analysis
As a manipulation check, an intensity-pattern (between)
by time (within) MANOVA was used to investigate the
effect of intensity pattern on the percentages of peak
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
Exercise as Affective Experience 153
JSEP Vol. 38, No. 2, 2016
heart rate (%HRpeak), RPE, and FS at Minutes 3, 6, 9,
12, and 15 during exercise. To determine the effect of
the experimental manipulation on postexercise pleasure,
an intensity-pattern group (between) by time (within)
ANOVA was used, with FS ratings obtained 2, 5, and 10
min after exercise as the dependent variable. For both the
MANOVA and ANOVA, if the sphericity assumption was
violated, the Greenhouse–Geisser adjustment was applied
to the degrees of freedom. The individual slopes of FS
during exercise (Minutes 3, 6, 9, 12, and 15) were calcu-
lated by linear regression. A series of regression analyses
were then conducted to determine (a) the association
of the slope of pleasure during exercise with ratings of
remembered pleasure, (b) the association of remembered
pleasure with forecasted pleasure, and (c) the association
of the slope of pleasure during exercise with forecasted
pleasure. Separate analyses were conducted for assess-
ments completed 15 min, 24 hr, and 7 days postexercise.
Likewise, a regression analysis was used to determine the
association of the slope of pleasure during exercise with
postexercise enjoyment. Finally, correlations were used
to assess the association of the average pleasure ratings
during exercise with remembered pleasure, forecasted
pleasure, and postexercise enjoyment.
Participant Characteristics
The two groups did not differ with respect to the assessed
demographic, anthropometric, behavioral, and physiolog-
ical characteristics (see Table 1). Half of the participants
(48%) were overweight or obese. Although the average
levels of planned exercise and incidental PA suggest
an active sample, there was considerable heterogeneity
and apparent discordance between self-reported PA and
objectively assessed cardiorespiratory tness. Based on
self-reports (IPEQ-WA), 77% of participants reported
never attending exercise classes. Only 3 of 46 reported
attending exercise classes on three or more days per week.
Most (61%) reported never doing any exercise at home.
Only 4 of 46 reported doing some exercise at home on
ve or more days per week. Half (53%) reported never
walking for exercise, while 31% reported some walking
for exercise but on fewer than 5 days per week. Although
40% reported some walking for transportation on a daily
basis, 83% walked for less than 30 min per day. The
average level of cardiorespiratory tness for both men
(29.33 ± 7.30 ml·kg–1·min–1) and women (27.16 ± 8.03
ml·kg–1·min–1) was “poor” (bottom 25%) by normative
standards for cycle ergometry (American College of
Sports Medicine, 2013, p. 84).
Manipulation Checks
The MANOVA showed no signicant effect of group,
Pillai’s V = .15, F (3, 38) = 2.18, p = .11, η2 = .15, which
suggests that the two groups were exposed to similar
overall levels of exercise intensity (%HRpeak), perceived
exertion (RPE), and during-exercise pleasure (FS). On
the other hand, there was a signicant group-by-time
interaction, Pillai’s V = .83, F (12, 480) = 15.32, p <
.001, η2 = .28. Univariate tests revealed that the interac-
tion was signicant for all three dependent variables: (a)
%HRpeak, F (1.37, 54.79) = 129.78, p < .001, η2 = .76;
(b) RPE, F (1.32, 52.89) = 55.63, p < .001, η2 = .58;
and (c) FS, F (1.83, 72.99) = 31.85, p < .001, η2 = .44.
As expected, in the increasing-intensity group, HR and
RPE increased over time, whereas FS ratings declined
over time. The opposite trends were evident in the
decreasing-intensity group (see Figure 1). Importantly,
the average FS ratings in the decreasing-intensity (1.91
± 1.45) and increasing-intensity groups (2.04 ± 1.34) did
not differ signicantly, t (44) = –0.31, p = .76, d = 0.09,
indicating that the total amount of reported pleasure was
similar in the two groups. Moreover, baseline FS levels
did not differ signicantly (increasing intensity: 2.23 ±
1.95; decreasing intensity: 2.83 ± 1.88, t (44) = 1.07, p
= .29, d = –0.32), suggesting that subsequent changes
in FS were unlikely to reect the law of initial values or
regression to the mean.
On the other hand, the individual slopes of during-
exercise FS ratings differed signicantly between the two
groups, t (44) = 7.24, p < .001, d = –2.14. As expected, in
the increasing-intensity group, the slope of FS was nega-
tive (–0.16 ± 0.17), whereas, in the decreasing-intensity
group, the slope was positive (+0.22 ± 0.18). Of the 22
participants allocated to the increasing-intensity group,
17 (77%) showed negative FS slopes (2 had slopes of
zero and 3 had near-zero positive slopes, from 0.03 to
0.10). Conversely, of the 24 participants allocated to the
decreasing-intensity group, 20 (83%) showed positive FS
slopes (1 had a slope of zero and 3 had near-zero negative
slopes, from –0.03 to –0.07). Following the intention-to-
treat principle, all participants were analyzed in the group
to which they were originally allocated.
Postexercise Pleasure
A 2 (groups) by 3 (time points: 2, 5, 10 min postexer-
cise) ANOVA on postexercise FS ratings revealed no
main effect of time, F (1.19, 52.45) = 0.06, p = .84, η2 =
.001, and no group-by-time interaction, F (1.19, 52.45)
= 0.06, p = .84, η2 = .001. There was, however, a sig-
nicant main effect of group, F (1, 44) = 4.39, p = .04,
η2 = .09. The mean postexercise FS rating was 3.38 ±
1.25 for the decreasing-intensity group and 2.56 ± 1.39
for the increasing-intensity group, t (44) = 2.09, p = .04,
d = –0.62.
Postexercise Enjoyment
An independent-sample t test on postexercise PACES
scores showed a signicant difference between groups,
t (43) = 3.32, p = .002, d = –0.99. Specically, the
decreasing-intensity group averaged 100.39 ± 11.46,
whereas the increasing-intensity group averaged 86.64
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
154 Zenko, Ekkekakis, and Ariely
JSEP Vol. 38, No. 2, 2016
± 16.04, on a scale on which 72 is the midpoint and 126
represents maximum enjoyment.
Remembered and Forecasted Pleasure
A 2 (groups) by 3 (time points: 15 min, 24 hr, 7 days)
MANOVA on remembered (VAS) and forecasted pleasure
(EVS) showed no signicant main effect of time, Pillai’s
V = .06, F (4, 168) = 1.31, p = .27, η2 = .03, but a signi-
cant main effect of group, Pillai’s V = .333, F (2, 41) =
10.22, p < .001, η2 = .33. The decreasing-intensity group
averaged signicantly higher levels of both remembered
pleasure (55.51 ± 23.45 vs. 25.05 ± 27.95), t (44) = 4.02, p
< .001, d = –1.19, and forecasted pleasure (51.75 ± 22.67
vs. 31.47 ± 26.05), t (44) = 2.82, p = .007, d = –0.83.
There was also a signicant group-by-time interaction,
Pillai’s V = .14, F (4, 168) = 3.23, p = .014, η2 = .07.
Follow-up univariate ANOVAs showed that the interac-
tion was signicant only for remembered pleasure, F (2,
84) = 6.62, p = .002, η2 = .14. The interaction was driven
by a gradual decline in the decreasing-intensity group
from 65.09 ± 21.81 at 15 min postexercise to 56.77 ±
21.88 a day later, t (21) = 3.00, p = .006, and to 51.42 ±
25.40 a week later, t (23) = 4.24, p < .001. Conversely,
the increasing-intensity group showed a smaller, non-
signicant improvement from 22.41 ± 30.02 at 15 min
postexercise to 27.41 ± 26.09 a week later. Despite this
convergence, even a week after the experimental session,
the difference between groups was still signicant and
large, t (44) = 3.16, p = .003, d = –0.93.
Slope of Pleasure, Postexercise
Enjoyment, Remembered,
and Forecasted Pleasure
The slope of FS ratings during exercise signicantly
predicted postexercise PACES scores, r = .58, r2 = .33, b
= 33.86, F (1, 43) = 21.21, p < .001. Similarly, in a series
of simple regressions, the slope of FS ratings predicted
all VAS ratings of remembered pleasure: (a) 15 min: r =
.68, r2 = .46, b = 88.04, t = 6.17, p < .001; (b) 24 hr: r =
.64, r2 = .40, b = 77.31, t = 5.33, p < .001; (c) 7 days: r
= .59, r2 = .35, b = 64.57, t = 4.87, p < .001. Likewise,
the slope of FS ratings predicted all EVS ratings of fore-
casted pleasure: (a) 15 min: r = .63, r2 = .40, b = 72.89,
t = 5.43, p < .001; (b) 24 hr: r = .61, r2 =.37, b = 61.84, t
= 4.94, p < .001; (c) 7 days: r = .61, r2 = .37, b = 62.01,
t = 5.04, p < .001 (see Figures 2 and 3). Because the
“24-hr” VAS and EVS ratings were not entered exactly
24 hr after exercise, hierarchical regressions were also
conducted, with the actual time (in minutes) since the
exercise session entered as the rst step. Variation in the
time of data entry was unrelated to predicted (p = .30) and
forecasted (p = .16) pleasure, leaving the 24-hr VAS and
EVS results essentially unchanged: b = 76.50, t = 5.27,
p < .001, and b = 60.85, t = 4.93, p < .001, respectively.
Despite the use of different ratings scales, remembered
pleasure and forecasted pleasure at each time point were
strongly interrelated: (a) 15 min: r = .84, p < .001; (b) 24
hr: r = .85, p < .001; (c) 7 days: r = .88, p < .001.
Figure 1 — Percentages of peak heart rate (Panel a), ratings of perceived exertion (Panel b), and Feeling Scale scores (Panel c) in the
decreasing-intensity (lled squares) and increasing-intensity (lled circles) groups during exercise. The error bars indicate standard errors.
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
Exercise as Affective Experience 155
JSEP Vol. 38, No. 2, 2016
Mean Pleasure and Postexercise
Enjoyment, Remembered,
and Forecasted Pleasure
Unlike the slope of during-exercise FS ratings, the aver-
age FS rating during exercise was unrelated to enjoyment,
remembered pleasure, and forecasted pleasure. The cor-
relation coefcients ranged from r = .14, p = .37, to r =
.26, p = .09.
The exercise promotion literature seems to reect a res-
ignation to the idea that, for individuals who are chroni-
cally sedentary and/or have low cardiorespiratory tness,
exercise is unlikely to be experienced as pleasant. This
is accepted as the unavoidable “price of admission,” a
“necessary evil,” that new exercisers must endure. Most
techniques being used to improve the exercise experience
(i.e., mainly cognitive interventions, including manipula-
tions of attentional focus) aim to attenuate the degree of
unpleasantness, thus making the exercise more tolerable,
though not necessarily pleasant. However, under condi-
tions of self-determination, for exercise to be sustainable
in the long term, it should be not just tolerable but more
pleasant than sedentary alternatives competing for a por-
tion of discretionary time (Ekkekakis & Dafermos, 2012).
The present study tested an innovative method,
derived from a cross-disciplinary evidence base, com-
bining the elds of exercise psychology and behavioral
economics. Specically, research on the relation between
exercise intensity and pleasure informed the intensity
manipulations used in this study. This evidence indicates
that an intensity that exceeds the ventilatory threshold
results in declining pleasure ratings, whereas the cessa-
tion of suprathreshold exercise leads to a robust rebound
in pleasure (Ekkekakis et al., 2011), reminiscent of the
affective contrast effect described by Solomon (1980). On
the other hand, evidence from behavioral economics has
shown that the slope of change of pleasure–displeasure
during an episode weighs heavily on subsequent retro-
spective evaluations of the experience (Ariely, 1998).
Based on such ndings, Ariely and Carmon (2000) sug-
gested that “summary evaluations may benet from an
(unneeded) initial low point in the experience prole,
since this allows for greater improvement over the dura-
tion of the experience” (p. 199). Ariely and Carmon
(2000) anticipated reactions to the seemingly counter-
intuitive nature of their suggestion: “if we ask decision
makers directly if they prefer to add an undesirable start
to their experience, they will most likely say no” (p. 199).
Yet, they predicted that “such an addition may be ‘better
for them’ in terms of their global evaluations” (p. 199).
The present findings are consistent with this
suggestion, as we observed robust benefits in the
Figure 2 — Cross-lagged correlations of during-exercise slope of pleasure with remembered and forecasted pleasure, 15 min, 24 h, and
7 days after the exercise bout (for all correlation coefcients, p < .001). Comparisons of remembered and forecasted pleasure between the
decreasing-intensity (gray bars) and increasing-intensity (white bars) groups at each of the three time points are shown in the bar graphs.
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
156 JSEP Vol. 38, No. 2, 2016
Figure 3 — Scatterplots illustrating the association of the slope of pleasure during the exercise bout with (a) postexercise pleasure (average of
ratings at 2, 5, and 10 min postexercise), (b) postexercise enjoyment, (c) remembered pleasure (15 min postexercise), (d) forecasted pleasure
(15 min postexercise), (e) remembered pleasure (24 hr postexercise), (f) forecasted pleasure (24 hr postexercise), (g) remembered pleasure
(7 days postexercise), and (h) forecasted pleasure (7 days postexercise). Pearson product–moment correlation coefcients and associated
probability values are shown for each graph.
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
Exercise as Affective Experience 157
JSEP Vol. 38, No. 2, 2016
decreasing-intensity (i.e., increasing-pleasure) group
compared with the increasing-intensity (i.e., decreasing-
pleasure) group in postexercise pleasure and enjoyment,
as well as in remembered and forecasted pleasure for up
to a week later. Moreover, consistent with the theoretical
prediction (Ariely & Carmon, 2000), unlike the slope of
pleasure, the average pleasure (and, therefore, also the
total amount of derived pleasure) reported during the
exercise bout was unrelated to postexercise enjoyment,
remembered pleasure, and forecasted pleasure.
These ndings have considerable potential impli-
cations for exercise promotion. These become readily
apparent when the study is placed within the context of the
ongoing heated debate about the role of exercise intensity
in exercise adoption and adherence. Since the mid-1990s,
PA recommendations issued by scientic organizations
and governmental agencies have focused primarily on
moderate-intensity activity (approximately correspond-
ing to “brisk” walking). This approach has been based
on “the belief that the promotion of moderate-intensity
exercise would lead to greater adoption and adherence
to exercise [compared to vigorous-intensity exercise]”
(Garber et al., 2011, pp. 1346–1347). However, with
increased recognition of the added health benets that can
be achieved with higher levels of intensity (e.g., Swain
& Franklin, 2006), since 2007 the American College of
Sports Medicine and the American Heart Association
have explicitly endorsed vigorous-intensity activity as an
option, alongside moderate-intensity activity (Haskell et
al., 2007). Moreover, for individuals with diabetes and
other cardiovascular risk factors, the American Heart
Association, citing evidence of robust cardiometabolic
adaptations, has noted that “vigorous intensities should
be targeted if tolerated and with consideration of contra-
indications” (Marwick et al., 2009, p. 3253).
To improve tolerability for nonathletic participants,
high-intensity sessions are structured as a sequence of
brief intense spurts interspersed with periods of rest or
active recovery. This pattern has become known as high-
intensity interval training (HIIT). Proponents argue that
HIIT offers the opportunity for the accrual of cardio-
metabolic adaptations in a time-efcient manner because
HIIT sessions can be of shorter total duration than typi-
cal moderate-intensity sessions (Jung, Bourne, & Little,
2014). In contrast, skeptics counter that HIIT might be
of limited value as a public-health intervention because
the inherent displeasure of high-intensity exercise could
discourage most participants and adversely inuence
adherence (Biddle & Batterham, 2015; Hardcastle, Ray,
Beale, & Hagger, 2014; Lunt et al., 2014).
The present study represents an attempt to advance
the discourse beyond the seemingly stagnant and polar-
izing debate on moderate-versus-vigorous intensity. The
pattern of intensity tested in this study (i.e., continuous
ramp-down) is the rst known attempt to respond to the
challenge issued by Dishman (1982) over three decades
ago, namely to devise a hitherto-elusive “compromise
between the ideal physiological prescription and a
manageable behavioral prescription” (p. 248). It is also
the rst attempt to develop an exercise prescription that
targets the promotion of pleasure as a central consider-
ation alongside effectiveness and safety (as proposed
by Ekkekakis et al., 2011) and the rst prescription to
directly incorporate theorizing from a discipline outside
the exercise sciences (i.e., behavioral economics).
Participants in the decreasing-intensity group
received, on average, approximately 5 min (i.e., 33%)
of exercise within the “vigorous” range, dened by the
American College of Sports Medicine (2013, p. 5) as
extending from 77% to < 94% of maximal heart rate (i.e.,
Minute 2: 78% ± 10%; Minute 3: 78% ± 10%; Minute
4: 77% ± 10%; Minute 5: 76% ± 10%). Yet their ratings
of affective valence, after the initial drop (pre: 2.83 ±
1.88; Minute 3: 0.67 ± 1.90), started an upward trend
(Minute 6: 1.08 ± 1.95), responding to the progressively
decreasing intensity, consistent with the affective contrast
effect (Solomon, 1980). The participants remained, on
average, within the range of “moderate” intensity, dened
as extending from 64% to < 77% of maximal heart rate,
until Minute 13 (Minute 12: 65% ± 8%; Minute 13:
64% ± 9%; Minute 14: 63% ± 8%). Overall, the bout of
decreasing intensity provided a reasonable combination
of vigorous and moderate intensity.
At the same time, the pattern of intensity observed
in the increasing-intensity group represents a realistic
simulation of a typical exercise bout exhibiting a con-
tinuous upward “drift” of physiological parameters. The
percentage of peak heart rate increased from 63% ± 7%
(Minute 3) to 86% ± 8% (Minute 15) and RPE increased
from 8.04 ± 1.53 (Minute 3) to 14.00 ± 2.18 (Minute 15).
For example, low-tness middle-aged women (VO2peak:
22.98 ± 5.69 ml·kg–1·min–1) who were asked to exercise
on a treadmill at their self-selected pace for 20 min started
from 67% ± 13% of peak heart rate after the warm-up
and progressed to 83% ± 13% at Minute 20. Their RPE
rose from 8.87 ± 1.77 to 13.78 ± 1.95 (Lind et al., 2005).
Limitations of the current study that future investiga-
tions should address include the following. First, while
the sample consisted of community volunteers with
heterogeneous PA proles and “poor” average cardiore-
spiratory tness, it remains to be seen whether the results
can be replicated in samples with different characteristics
(e.g., participants with higher cardiorespiratory tness
or participants with obesity, diabetes, or coronary artery
disease). Second, even though this study demonstrated
robust differences in during-exercise pleasure and imme-
diate postexercise pleasure and enjoyment, as well as
more distal remembered and forecasted pleasure, future
investigations should examine the effects of repeated
sessions involving intensity ramp-downs on PA behavior
and exercise adherence.
In conclusion, we developed and tested an innova-
tive and practical exercise bout that combines exposure
to meaningful doses of vigorous and moderate intensity
with signicantly improved postexercise pleasure and
enjoyment, remembered pleasure, and forecasted plea-
sure. Exercise resulting in more pleasure during the bout
(Ekkekakis & Dafermos, 2012; Rhodes & Kates, 2015),
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
158 Zenko, Ekkekakis, and Ariely
JSEP Vol. 38, No. 2, 2016
more enjoyment (Rhodes et al., 2009), more remembered
pleasure (Brewer, Manos, McDevitt, Cornelius, & van
Raalte, 2000), and more forecasted pleasure (Conner et
al., 2015; Dunton & Vaughan, 2008; Helfer et al., 2015;
Loehr & Baldwin, 2014) should lead to higher levels of
subsequent PA and exercise participation.
American College of Sports Medicine. (2013). ACSM’s
guidelines for exercise testing and prescription (9th ed.).
Philadelphia, PA: Lippincott Williams & Wilkins.
Ariely, D. (1998). Combining experiences over time: The effects
of duration, intensity changes and on-line measurements
on retrospective pain evaluations. Journal of Behavioral
Decision Making, 11, 19–45. doi:10.1002/(SICI)1099-
Ariely, D., & Carmon, Z. (2000). Gestalt characteristics of experi-
ences: The dening features of summarized events. Journal
of Behavioral Decision Making, 13, 191–201. doi:10.1002/
Ariely, D., & Zauberman, G. (2003). Differential partitioning
of extended experiences. Organizational Behavior and
Human Decision Processes, 91, 128–139. doi:10.1016/
Biddle, S.J., & Batterham, A.M. (2015). High-intensity interval
exercise training for public health: A big HIT or shall we
HIT it on the head? The International Journal of Behav-
ioral Nutrition and Physical Activity, 12:95. PubMed
Borg, G. (1998). Borg’s perceived exertion and pain scales.
Champaign, IL: Human Kinetics.
Brewer, B.W., Manos, T.M., McDevitt, A.V., Cornelius, A.E., &
van Raalte, J.L. (2000). The effect of adding lower inten-
sity work on perceived aversiveness of exercise. Journal
of Sport & Exercise Psychology, 22, 119–130.
Castro, C.M., Sallis, J.F., Hickmann, S.A., Lee, R.E., & Chen,
A.H. (1999). A prospective study of psychosocial correlates
of physical activity for ethnic minority women. Psychology
& Health, 14, 277–293. doi:10.1080/08870449908407328
Chen, M.J., Fan, X., & Moe, S.T. (2002). Criterion-related valid-
ity of the Borg ratings of perceived exertion scale in healthy
individuals: A meta-analysis. Journal of Sports Sciences,
20, 873–899. PubMed doi:10.1080/026404102320761787
Conner, M., McEachan, R., Taylor, N., O’Hara, J., & Lawton, R.
(2015). Role of affective attitudes and anticipated affective
reactions in predicting health behaviors. Health Psychol-
ogy, 34, 642–652. PubMed doi:10.1037/hea0000143
Delbaere, K., Hauer, K., & Lord, S.R. (2010). Evaluation of
the incidental and planned activity questionnaire (IPEQ)
for older people. British Journal of Sports Medicine, 44,
1029–1034. PubMed doi:10.1136/bjsm.2009.060350
Dishman, R.K. (1982). Compliance/adherence in health-
related exercise. Health Psychology, 1, 237–267.
Dunton, G.F., & Vaughan, E. (2008). Anticipated affective
consequences of physical activity adoption and main-
tenance. Health Psychology, 27, 703–710. PubMed
Edmunds, J., Ntoumanis, N., & Duda, J.L. (2007). Adherence
and well-being in overweight and obese patients referred
to an exercise on prescription scheme: A self-determination
theory perspective. Psychology of Sport and Exercise, 8,
722–740. doi:10.1016/j.psychsport.2006.07.006
Ekkekakis, P., & Dafermos, M. (2012). Exercise is a many-
splendored thing, but for some it does not feel so splended:
Staging a resurgence of hedonistic ideas in the quest to
understand exercise behavior. In E.O. Acevedo (Ed.),
Oxford handbook of exercise psychology (pp. 295–333).
New York: Oxford University Press.
Ekkekakis, P., Lind, E., & Vazou, S. (2010). Affective responses
to increasing levels of exercise intensity in normal-weight,
overweight, and obese middle-aged women. Obesity
(Silver Spring, Md.), 18, 79–85. PubMed doi:10.1038/
Ekkekakis, P., Partt, G., & Petruzzello, S.J. (2011). The plea-
sure and displeasure people feel when they exercise at dif-
ferent intensities: Decennial update and progress towards
a tripartite rationale for exercise intensity prescription.
Sports Medicine (Auckland, N.Z.), 41, 641–671. PubMed
Finucane, M.L., Alhakami, A.L.I., Slovic, P., & Johnson, S.M.
(2000). The affect heuristic in judgments of risks and ben-
ets. Journal of Behavioral Decision Making, 13, 1–17.
Garber, C.E., Blissmer, B., Deschenes, M.R., Franklin,
B.A., Lamonte, M.J., Lee, I.M., . . . Swain, D.P. (2011).
Quantity and quality of exercise for developing and
maintaining cardiorespiratory, musculoskeletal, and neu-
romotor tness in apparently healthy adults: Guidance
for prescribing exercise. Medicine and Science in Sports
and Exercise, 43, 1334–1359. PubMed doi:10.1249/
Gaskill, S.E., Ruby, B.C., Walker, A.J., Sanchez, O.A., Serfass,
R.C., & Leon, A.S. (2001). Validity and reliability of com-
bining three methods to determine ventilatory threshold.
Medicine and Science in Sports and Exercise, 33, 1841–
1848. PubMed doi:10.1097/00005768-200111000-00007
Hardcastle, S.J., Ray, H., Beale, L., & Hagger, M.S. (2014).
Why sprint interval training is inappropriate for a largely
sedentary population. Frontiers in Psychology, 5, 1505.
PubMed doi:10.3389/fpsyg.2014.01505
Hardy, C.J., & Rejeski, W.J. (1989). Not what, but how one
feels: The measurement of affect during exercise. Journal
of Sport & Exercise Psychology, 11, 304–317.
Haskell, W.L., Lee, I.M., Pate, R.R., Powell, K.E., Blair, S.N.,
Franklin, B.A., . . . Bauman, A. (2007). Physical activity
and public health: Updated recommendation for adults
from the American College of Sports Medicine and the
American Heart Association. Medicine and Science in
Sports and Exercise, 39, 1423–1434. PubMed doi:10.1249/
Helfer, S.G., Elhai, J.D., & Geers, A.L. (2015). Affect and
exercise: Positive affective expectations can increase
post-exercise mood and exercise intentions. Annals of
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
Exercise as Affective Experience 159
JSEP Vol. 38, No. 2, 2016
Behavioral Medicine, 49, 269–279. PubMed doi:10.1007/
Jung, M.E., Bourne, J.E., & Little, J.P. (2014). Where does
HIT t? An examination of the affective response to
high-intensity intervals in comparison to continuous
moderate- and continuous vigorous-intensity exercise
in the exercise intensity-affect continuum. PLoS One,
9, e114541. PubMed doi:10.1371/journal.pone.0114541
Kahneman, D. (2003). A perspective on judgment and choice:
Mapping bounded rationality. The American Psychologist,
58, 697–720. PubMed doi:10.1037/0003-066X.58.9.697
Kahneman, D., Wakker, P.P., & Sarin, R. (1997). Back
to Bentham? Explorations of experienced utility.
The Quarterly Journal of Economics, 112, 375–406.
Kendzierski, D., & DeCarlo, K.J. (1991). Physical Activity
Enjoyment Scale: Two validation studies. Journal of Sport
& Exercise Psychology, 13, 50–64.
Kiviniemi, M.T., Voss-Humke, A.M., & Seifert, A.L. (2007).
How do I feel about the behavior? The interplay of affective
associations with behaviors and cognitive beliefs as inu-
ences on physical activity behavior. Health Psychology,
26, 152–158. PubMed doi:10.1037/0278-6133.26.2.152
Lind, E., Joens-Matre, R.R., & Ekkekakis, P. (2005). What
intensity of physical activity do formerly sedentary
middle-aged women select? Evidence of a coherent pat-
tern from physiological, perceptual, and affective markers.
Preventive Medicine, 40, 407–419. PubMed doi:10.1016/j.
Lishner, D.A., Cooter, A.B., & Zald, D.H. (2008). Address-
ing measurement limitations in affective rating scales:
Development of an empirical valence scale. Cognition and
Emotion, 22, 180–192. doi:10.1080/02699930701319139
Loehr, V.G., & Baldwin, A.S. (2014). Affective forecasting
error in exercise: Differences between physically active
and inactive individuals. Sport, Exercise, and Performance
Psychology, 3, 177–183. doi:10.1037/spy0000006
Lunt, H., Draper, N., Marshall, H.C., Logan, F.J., Hamlin, M.J.,
Shearman, J.P., . . . Frampton, C.M.A. (2014). High inten-
sity interval training in a real world setting: A randomized
controlled feasibility study in overweight inactive adults,
measuring change in maximal oxygen uptake. PLoS One,
9, e83256. PubMed doi:10.1371/journal.pone.0083256
Marcus, B.H., Williams, D.M., Dubbert, P.M., Sallis, J.F., King,
A.C., Yancey, A.K., . . . Claytor, R.P. (2006). Physical
activity intervention studies: What we know and what
we need to know. Circulation, 114, 2739–2752. PubMed
Martin, S.B., Morrow, J.R., Jr., Jackson, A.W., & Dunn,
A.L. (2000). Variables related to meeting the CDC/
ACSM physical activity guidelines. Medicine and Sci-
ence in Sports and Exercise, 32, 2087–2092. PubMed
Marwick, T.H., Hordern, M.D., Miller, T., Chyun, D.A., Bertoni,
A.G., Blumenthal, R.S., ... Rocchini, A. (2009). Exercise
training for type 2 diabetes mellitus: Impact on cardiovas-
cular risk. Circulation, 119, 3244 3262.
O’Donovan, G., & Shave, R. (2007). British adults’ views on
the health benets of moderate and vigorous activity.
Preventive Medicine, 45, 432–435. PubMed doi:10.1016/j.
Rhodes, R.E., Fiala, B., & Conner, M. (2009). A review and
meta-analysis of affective judgments and physical activ-
ity in adult populations. Annals of Behavioral Medicine,
38, 180–204. PubMed doi:10.1007/s12160-009-9147-y
Rhodes, R.E., & Kates, A. (2015). Can the affective response
to exercise predict future motives and physical activity
behavior? A systematic review of published evidence.
Annals of Behavioral Medicine, 49, 715–731. PubMed
Russell, J.A. (1980). A circumplex model of affect. Journal
of Personality and Social Psychology, 39, 1161–1178.
Sheppard, K.E., & Partt, G. (2008). Acute affective responses
to prescribed and self-selected exercise intensities in young
adolescent boys and girls. Pediatric Exercise Science, 20,
129–141. PubMed
Simon, H.A. (1983). Reason in human affairs. Stanford, CA:
Stanford University Press.
Solomon, R.L. (1980). The opponent-process theory of acquired
motivation: The costs of pleasure and the benets of
pain. The American Psychologist, 35, 691–712. PubMed
Stanovich, K.E., & West, R.F. (2000). Individual differences
in reasoning: Implications for the rationality debate?
Behavioral and Brain Sciences, 23, 645–665. PubMed
Stevens, M., Lemmink, K.A., van Heuvelen, M.J., de Jong, J.,
& Rispens, P. (2003). Groningen Active Living Model
(GALM): Stimulating physical activity in sedentary
older adults; validation of the behavioral change model.
Preventive Medicine, 37, 561–570. PubMed doi:10.1016/j.
Swain, D., & Franklin, B.A. (2006). Comparison of car-
dioprotective benefits of vigorous versus moderate
intensity aerobic exercise. The American Journal of
Cardiology, 97, 141–147. PubMed doi:10.1016/j.amj-
Tudor-Locke, C., Brashear, M.M., Johnson, W.D., & Katzmar-
zyk, P.T. (2010). Accelerometer profiles of physical
activity and inactivity in normal weight, overweight, and
obese U.S. men and women. The International Journal of
Behavioral Nutrition and Physical Activity, 7, 60. PubMed
Welch, A.S., Hulley, A., Ferguson, C., & Beauchamp, M.R.
(2007). Affective responses of inactive women to a
maximal incremental exercise test: A test of the dual-mode
model. Psychology of Sport and Exercise, 8, 401–423.
Wilson, P.M., Rodgers, W.M., Blanchard, C.M., & Gessell, J.
(2003). The relationship between psychological needs,
self-determined motivation, exercise attitudes, and
physical tness. Journal of Applied Social Psychology,
33, 2373–2392. doi:10.1111/j.1559-1816.2003.tb01890.x
Zauberman, G., Diehl, K., & Ariely, D. (2006). Hedonic versus
informational evaluations: Task dependent preferences for
sequences of outcomes. Journal of Behavioral Decision
Making, 19, 191–211. doi:10.1002/bdm.516
Downloaded by Duke University on 09/07/16, Volume 38, Article Number 2
... Moreover, forecasted pleasure/utility concerns how pleasant or unpleasant future experiences are predicted to be. Researchers have theorized that both remembered and forecasted pleasure can help predict whether behavior will be repeated (Karl et al., 2021;Zenko et al., 2016). Hence, such constructs could yield considerable value when evaluating the extent to which AR experiences can help retain visitors. ...
... The slider was initially positioned at the origin (0). The descriptors and slider were visible to participants but the numbers were not (Zenko et al., 2016). ...
... Presence was assessed by the Augmented Reality Immersion questionnaire (Georgiou & Kyza, 2017). Remembered pleasure and forecasted pleasure were measured using visual analogue scales (Lishner et al., 2008;Zenko et al., 2016). User engagement was assessed using items derived from tom Dieck et al. (2018). ...
Full-text available
Augmented reality (AR) is an emergent technology in tourism. However, research concerning the AR user experience is relatively scarce and seldom addresses the intentions of designers. Accordingly, we sought to: (a) explore the design intentions underlying a multi-user, purpose-built AR experience; (b) assess the extent to which users’ realized experience aligned with the designers’ intended experience; and (c) examine the relationships between users’ internal states and their associated behavior, in alignment with a Stimulus-Organism-Response framework. In Study 1, designers ( n = 5) took part in a focus group and completed a design intentions survey. In Study 2, users ( n = 48) tested the AR experience, and a range of subjective (e.g., affective responses) and objective (i.e., visual attention) data were recorded. Findings indicated designer–user disparities primarily at the organism and response levels. Additionally, users’ affective responses to the AR experience were strongly associated with visitor engagement.
... Remembered and forecasted pleasure are typically linked; how one recalls an exercise session is presumed to influence anticipated affective responses to subsequent exercise sessions (e.g., Davis & Stenling, 2020). Zenko et al. (2016) observed strong positive associations of remembered pleasure and subsequent forecasted pleasure assessed at 15 min (r = .84), 24 hours (r = .86), ...
... This model combines physiological considerations (i.e., inclusion of high-intensity work that enhances physiological adaptations to exercise) and psychological considerations (i.e., promoting more positive affective responses). The opposingslopes model was developed based on evidence from behavioral economics and Solomon's (1980) "opponent process" theory of acquired motivation (see Hutchinson et al., 2020;Zenko et al., 2016). ...
... The opposing-slopes approach was first empirically tested in the context of exercise by Zenko et al. (2016), who randomly assigned participants to a 15-min bout of recumbent cycling of either increasing (UP) intensity (0-120% of watts corresponding to each participant's ventilatory threshold) or decreasing (DOWN) intensity (i.e., 120-0%). The DOWN condition elicited a positive slope of pleasure during exercise, meaning that participants felt increasingly more pleasure as the exercise task progressed. ...
... Increased concentrations of blood lactate combined with increased oxygen consumption are metabolically unsustainable and cause unpleasant feelings in the participants, urging them to stop the exercise [15]. The importance of this theory is that adherence to exercise may be influenced by affective responses, thus rendering them important components of exercise prescription [16,17]. ...
Full-text available
The current study examines the effects of a Tabata high-intensity interval training (HIIT) session on affective, cognitive and physiological indicators in women of different fitness levels. A total of 28 adult women (aged 24.2 ± 1.5 years) completed a 20 m shuttle run test and were then assigned to higher fitness and lower fitness groups (HF and LF, n = 14 each) according to their predicted aerobic power. On a separate occasion, participants completed a 30 min Tabata workout (six 4 min rounds separated by 1 min passive rest). Each round included eight exercises (20 s exercise and 10 s rest). Affective, physiological and cognitive responses were assessed prior to, during and after the protocol. Heart rate and blood lactate concentration increased similarly in both groups over time throughout the workout (p < 0.001). Total Mood Disturbance was higher for LF (111.4 ± 15.7) vs. HF (102.9 ± 11.7) (p = 0.48), vigor showed a level by time interaction of p = 0.006 and Activation–Deactivation Adjective Check List factors deteriorated over time (p < 0.001). The Concentration Grid Test was better overall for HF (10.5 ± 3.6) vs. LF (8.6 ± 3.6) (p = 0.05). The Feeling Scale and Rating of Perceived Exertion worsened similarly in both groups over time (p = 0.002 and p < 0.001, respectively). Positive and negative affect and arousal did not differ between groups or change over time (p > 0.05). These results show that, despite the different levels of aerobic fitness, physiological, metabolic, perceptual and affective responses were similar in the two groups of women during a 30 min Tabata session. This may imply that affective responses during this type of HIIT are independent of aerobic fitness.
... Regarding internal parameters, opting for self-selected intensity 46 or manipulating the structure of the session (e.g., ending the session with a lower intensity) 47 can promote positive affective experiences toward PA. In this perspective, practitioners (e.g., physical education teachers, health professionals) are uniquely placed to nurture environments that effectively promote positive affective experiences toward PA. ...
Full-text available
Most individuals are now aware of health benefits of physical activity (PA) but remain physically inactive. Mobilizing a multidisciplinary approach at the crossroads between decision-making sciences, we investigate why highlighting the health benefits of PA is unlikely to promote a sustained engagement in PA. Essential features of decision making – effort-discounting, delay-discounting and beliefs distortion – may weaken the subjective value attributed to health benefits, making the latter insufficient to trigger PA behaviors. We develop a decision model demonstrating that health benefits hold a weak subjective value, in comparison with the cost of engaging in PA (e.g., effort) and of our innate attraction toward sedentary alternatives. Instead, focusing on positive affective experiences could counteract the impact of aforementioned features and ultimately favor a regular engagement in PA. Tackling the current pandemic of physical inactivity would therefore require an urgent change in the promotion of PA, so as to make affective experiences central.
Full-text available
Introduction From a public health perspective, it is important to gain more insight into how people can be motivated to maintain effective exercise routines. It is a common belief that moderate-intensity exercise is more pleasant and enjoyable than high-intensity training. This study aims to provide insight into (1) participants' expectations and preferences for training intensity prior to training, (2) how longer-term participation affect participants' experience of endurance training with continuous moderate-intensity training and high-intensity interval training. Materials and methods A total of 22 participants (14 women and eight men) between the ages of 21–30 volunteered for participation. Participants were randomized and divided into two equal groups. A total of 17 participants, nine women and eight men, completed the study. One group did moderate-intensity longer-lasting training and the other did high-intensity interval training. All participants completed three training sessions per week for 8 weeks. Semi-structured interviews were conducted with each participant before and after completing the training intervention. Data was analyzed using thematic analysis. This study is a part of a larger study evaluating and comparing the effects on endurance capacity of high-intensity interval training and moderate-intensity training. Physiological data are previously published. Results The results describe participants expectations prior to training, and how they experienced the actual training. The overall experience of training comprises several factors that work together. Both expectations and actual experiences (e.g., of physical pleasantness or unpleasantness, of positive or negative emotions, and of actual results from the training) contribute to the participants' overall experience of exercise. Conclusion The major finding is that improved physical fitness was a stronger motivator than feelings of pleasantness. Experiencing good results seemed to downplay feelings of unpleasantness and reinforce positive feelings toward exercise. Lack of results reinforce negative feelings toward exercise. Participants reported high-intensity exercise as more unpleasant and exhaustive, but the interval training group were more satisfied and experienced the training as more motivating.
This study aimed to understand determinants of recalled in-task affective valence experienced during a regularly performed aerobic bout in adult exercisers aged 55+. Qualitative data were collected (January to March, 2021, during the COVID-19 pandemic) using interviews wherein individuals (N = 16, 69% women, 61 ± 5 years) recalled deviations in affective valence in response to a regularly completed bout. Using thematic analyses, two themes emerged regarding how COVID-19 impacted regular exercise behaviors: (a) "loss" and (b) "adaptation." Two themes encompassed the determinants of recalled in-task affective valence: (a) "person-specific conditions" and (b) "external conditions." Finally, an increase in duration/intensity during a pleasant session was indicated by 44% of the participants, while 75% indicated a decrease in duration/intensity during an unpleasant session. The participants indicated that affective valence was determined by previously cited and novel factors that relate to exercise performed in naturalistic environments. Volitional modifications to planned exercise volume appear more responsive to feelings of displeasure.
Background Women with polycystic ovary syndrome (PCOS) experience general and PCOS-specific barriers that limit their engagement with exercise and contributes to high attrition from exercise programs, hindering the potential benefits of exercise to address their increased cardio-metabolic risk. A positive remembered affective response can predict future intentions and adherence to exercise prescription. Objectives To compare the longitudinal changes in remembered affect to high-intensity interval training (HIIT) and moderate-intensity continuous training (MICT) in women with PCOS and to determine whether longitudinal changes in remembered affect are correlated with changes in fitness, body mass index, adherence and exercise enjoyment. Methods Physically inactive, overweight women with PCOS were randomly assigned to 12 weeks of either HIIT (n = 15) or MICT (n = 14) (3 sessions per week). Remembered affective valence (Feeling Scale) was collected after each exercise session. Cardiorespiratory fitness (VO2peak) was assessed at baseline and post-intervention. Exercise enjoyment was assessed post-intervention. Results The longitudinal changes in the remembered affect were more positive in the HIIT group compared to MICT (β = 0.017, p = 0.047). HIIT was also considered more enjoyable than MICT (p = 0.002). Adherence was high in both groups (>90%). We found a moderate correlation with longitudinal changes between the remembered affect and change in fitness (rs = 0.398) and exercise enjoyment (rs = 0.376) using the combined group, however, these were not statistically significant (p = 0.054 and p = 0.064, respectively). Conclusions HIIT demonstrated a more positive longitudinal remembered affective response and greater exercise enjoyment compared to MICT in overweight women with PCOS.
This study expanded the State Mindfulness Scale for Physical Activity (SMS-PA) to include acceptance items to better represent core elements of mindfulness. Young adults who just participated in physical activity (N = 394) completed a survey to assess state mindfulness and theoretically relevant constructs about affect, motivation, and body image. An exploratory factor analysis was used to reduce the item pool on half of the sample. A 19-item and 15-item version of the SMS-PA2 were further tested through confirmatory factor analysis on the second half of the sample demonstrating a theoretically based factor structure representing either a total score or four separate factors – monitoring of the mind and body and accepting of the mind and body. The SMS-PA2 scores demonstrated evidence supporting construct and incremental validity through associations with theoretically relevant variables. Initial evidence shows expanded predictive utility of the SMS-PA2.
This meta‐analytic study aimed to examine the effects of audiovisual stimuli on affective responses during and after exercise and their moderators. A total of 296 effect sizes (Hedge's g) were extracted from 46 independent studies covering 1292 participants. Meta‐analysis was performed using Comprehensive Meta‐Analysis Version 3.3, and potential moderating variables were analysed using univariate meta‐regression models. Audiovisual stimuli increased affective valence during (g = 0.793, 95% CI [0.666, 0.920]) and after exercise (g = 0.792, 95% CI [0.567, 1.016]), and arousal during (g = 0.920, 95% CI [0.742 1.097]) and after exercise (g = 0.666, 95% CI [0.390, 0.962]). There may be publication bias in the meta‐analysis, but the main findings are still valid. The type of audiovisual stimuli (audio‐video > audio or video), exercise habits (active > not reported), and exercise intensities (self‐selected > imposed) moderated the effects. In conclusion, the application of audiovisual stimuli during exercise can elicit positive affective responses. These results provide a viable intervention strategy for exercise and health practitioners to reduce the number of physically inactive individuals and improve exercise compliance and adherence.
Full-text available
The tendency to underestimate how enjoyable exercise will be—an affective forecasting error—is thought to undermine motivation for regular physical activity. We sought to clarify limitations of previous work by examining whether (a) physically inactive individuals show the same forecasting error as physically active individuals, and (b) experienced enjoyment mediates the relation between expected enjoyment and intentions, and whether physical activity levels moderate this relation. Prior to a 30-min workout, physically inactive (60 min of physical activity/week; N 18) and active (150 min of physical activity/week; N 24) individuals reported their expected enjoyment. Afterward, they reported experienced enjoyment and exercise intentions. We found a marginally significant interaction (p .07, partial 2 .08) between group (active, inactive) and time (expected, experienced enjoyment), suggesting the forecasting error differed for active and inactive individuals. Specifically, inactive individuals reported significantly lower expected enjoyment than active individuals (p .02, d .73), but reported similar levels of experienced enjoyment (p .27). We also found that experienced enjoyment mediated the relation between expected enjoyment and exercise intentions for inactive (ab .367, 95% confidence interval [CI] .075, .742) but not active individuals (ab .079, 95% CI .269, .089). The findings suggest that lower expectations for exercise enjoyment characterize physically inactive individuals and provide support for the conclusion that the affective forecasting error undermines motivation for regular physical activity. However, among inactive individuals, experienced enjoyment had a stronger relation with intentions to exercise regularly than expected enjoyment.
Full-text available
Contemporary theories of exercise behavior have been the products of the so-called cognitive revolution, which has shaped the dominant paradigm in psychology over the past several decades. Cognitive theories rely on the assumption that, in making behavioral decisions, humans collect relevant information and make their selections on the basis of a more-or-less rational analysis of this information. Although the dominance of cognitive theories in the field of exercise psychology is unquestionable, evidence suggests that they leave most of the variance in exercise behavior unaccounted and interventions based on them are of limited effectiveness in changing exercise behavior. This chapter reviews the history and evaluates the potential of an alternative approach, namely the hedonic theory of motivation. This idea, long neglected due the fascination of psychologists with informationprocessing models of the mind, attributes a substantial portion of the variance in decision-making to affective processes. Modern iterations of the idea emerging from the fields of neurology and behavioral economics reaffirm the ancient thesis that, in the long run, humans tend to repeat what makes them feel better and tend to avoid what makes them feel worse. Evidence from studies in the context of exercise suggests that affective responses to exercise vary greatly between individuals. Furthermore, despite a still-evolving methodological platform, preliminary studies show that affective responses to exercise predict subsequent exercise behavior. This line of research and theorizing offers a novel and intriguing perspective on the mechanisms underlying behavioral decision-making in the context of exercise. The literature reviewed in this chapter highlights the need for further research on the motivational implications of affective processes and lays the foundation for the development of a hedonic theory of exercise behavior.
Full-text available
Background: The efficacy of high-intensity interval training for a broad spectrum of cardio-metabolic health outcomes is not in question. Rather, the effectiveness of this form of exercise is at stake. In this paper we debate the issues concerning the likely success or failure of high-intensity interval training interventions for population-level health promotion. Discussion: Biddle maintains that high-intensity interval training cannot be a viable public health strategy as it will not be adopted or maintained by many people. This conclusion is based on an analysis of perceptions of competence, the psychologically aversive nature of high-intensity exercise, the affective component of attitudes, the less conscious elements of motivated behaviour that reflect our likes and dislikes, and analysis using the RE-AIM framework. Batterham argues that this appraisal is based on a constrained and outmoded definition of high-intensity interval training and that truly practical and scalable protocols have been - and continue to be - developed. He contends that the purported displeasure associated with this type of exercise has been overstated. Biddle suggests that the way forward is to help the least active become more active rather than the already active to do more. Batterham claims that traditional physical activity promotion has been a spectacular failure. He proposes that, within an evolutionary health promotion framework, high-intensity interval training could be a successful population strategy for producing rapid physiological adaptations benefiting public health, independent of changes in total physical activity energy expenditure. Summary: Biddle recommends that we focus our attention elsewhere if we want population-level gains in physical activity impacting public health. His conclusion is based on his belief that high-intensity interval training interventions will have limited reach, effectiveness, and adoption, and poor implementation and maintenance. In contrast, Batterham maintains that there is genuine potential for scalable, enjoyable high-intensity interval exercise interventions to contribute substantially to addressing areas of public health priority, including prevention and treatment of Type 2 diabetes and cardiovascular disease.