Content uploaded by Jana Nikitin
Author content
All content in this area was uploaded by Jana Nikitin on Nov 28, 2019
Content may be subject to copyright.
Running head: AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 1
Age Differences in Implicit Theories About Willpower: Why Older People Endorse a
Nonlimited Theory
Article Accepted for publication in Psychology and Aging
This manuscript might not exactly replicate the final version published in the journal. It is not the copy
of record.
Please cite as: Job, V., Sieber, V., Rothermund, K., & Nikitin, J. (2018). Age differences in implicit
theories about willpower: Why older people endorse a nonlimited theory. Psychology and Aging, 33,
940-952. doi:0.1037/pag0000285
Veronika Job1, Vanda Sieber2, Klaus Rothermund3, Jana Nikitin4
1Technische Universität Dresden
2University of Zurich
3Friedrich-Schiller University Jena
4University of Basel
Author Note
Veronika Job, Faculty of Psychology, Technische Universität Dresden; Vanda Sieber,
Department of Psychology, University of Zurich; Klaus Rothermund, Department of
Psychology, Friedrich-Schiller University Jena; Jana Nikitin, Faculty of Psychology,
University of Basel
This research was supported by a grant from the Swiss National Science Foundation to
Veronika Job (100014_159395) and to Jana Nikitin (100019_159399).
Correspondence should be addressed to Veronika Job, Faculty of Psychology, Technische
Universität Dresden, Zellescher Weg 17, 01069 Dresden Zurich, Germany. Email:
veronik.job@tu-dresden.de
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 2
Abstract
What people believe about their capacity to exert self-control (willpower), whether it
is a limited or a nonlimited resource, affects their self-regulation and well-being. The present
research investigated age-related differences in people’s beliefs—called implicit theories—
about willpower. Study 1 (n = 802, age range 18–83 years) showed that with higher age
people are more likely to believe that willpower is a nonlimited resource. Study 2 (n = 423)
with younger (age 18–35 years) and older adults (age 60–98 years) replicated this finding and
showed that age and a nonlimited willpower theory are related to perceived autonomy on
demanding tasks (i.e., sense of self-determination), which might explain the age-related
differences in willpower theories. Finally, experimental Studies 3a (n = 302) and 3b (n = 497)
manipulated an autonomous mindset in younger (age 18–35 years) and older adults (age 60–
87 years) and provided evidence for a causal effect of perceived autonomy on self-control-
beliefs, supporting the proposed developmental mechanism.
Keywords: willpower theories, lay beliefs, implicit theories, autonomy, ageing
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 3
Age Differences in Implicit Theories About Willpower: Why Older People Endorse a
Nonlimited Theory
Young people in the western cultural context tend to believe that their willpower (their
capacity to exert self-control) is a limited resource (Job, Dweck, & Walton, 2010; Savani &
Job, in press). People who hold such a limited-resource theory about willpower believe that
exerting self-control depletes their energy and makes them more vulnerable to fail in the face
of accumulating self-control demands such as working on strenuous mental tasks for a long
time or wanting to resist multiple temptations. Research showed that this limited-resource
theory explains reduced self-control capacity on consecutive tasks in the laboratory (ego-
depletion effect; Baumeister, Vohs, & Tice, 2007) and stress-related lapses of self-control in
everyday life (Bernecker & Job, 2015a; Job et al., 2010; Job, Walton, Bernecker, & Dweck,
2015), which were observed in student samples in the USA and Western Europe.
Less is known, however, about the development of implicit theories beyond young
adulthood. In the present research, we asked the question whether older adults differ from
young adults in their implicit theories about willpower. Does the experience of declining
physical and cognitive capacity associated with ageing strengthen and increase the belief that
self-control capacity, too, is a precious and limited resource? Or, in contrast, do positive
developments in autonomy (the possibility to freely decide how one wants to spend one’s
time due to reduced external obligations and constraints) across adulthood promote the
development of a more nonlimited theory about willpower? We started out testing these two
competing assumptions.
Implicit Theories about Willpower
Implicit theories are “people’s basic assumptions about themselves and their world”
(Dweck, 1996, p. 69). People develop implicit theories about the nature of various human
characteristics, such as intelligence or morality to explain phenomena in the world and to
make choices concerning their own behavior (Dweck, 1999). For example, believing that
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 4
intelligence is malleable makes people strive for learning goals (Dweck & Leggett, 1988).
Typically, laypersons are not consciously reflecting on the theories they have in their
everyday life. They are also not aware of how these theories affect their behavior. This is the
reason why lay theories are called “implicit”. However, people are typically capable of
recognizing their own theory when different theories are stated in a questionnaire.
Recently, researchers started to investigate people’s implicit theories about self-
control capacity, colloquially called willpower (Job et al., 2010; Martijn, Tenbult,
Merckelbach, Dreezens, & De Vries, 2002). It was shown that some people hold a so-called
limited theory, which is assessed with items such as “After a strenuous mental activity, your
energy is depleted and you must rest to get it refueled again”. Others, however, do not share
this belief. In contrast, they think that engaging in demanding tasks may further activate their
self-control capacity and increase their ability to master new demands. This so-called
nonlimited theory is assessed with items like “After a strenuous mental activity, you feel
energized for further challenging activities”. This latter belief was not labeled “unlimited” but
“nonlimited” to specify that people with a nonlimited theory of willpower reject the view that
willpower is easily depleted by acts of self-control. They do not have to believe that
willpower is limitless and that they can exert self-control endlessly without rest and recovery
(Job et al., 2015).
In a series of laboratory experiments, Job and colleagues (2010) showed that only
people who believe that willpower is a limited resource show the ego-depletion effect, that is,
they perform worse on an unrelated task after they exerted self-control. Conversely, people
who believe that willpower is not limited show no impairment over a series of demanding
self-control tasks (Chow, Hui, & Lau, 2014; Salmon, Adriaanse, De Vet, Fennis, & De
Ridder, 2014). These results were found both when measuring implicit theories about
willpower as an individual difference variable and when manipulating them experimentally,
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 5
supporting the assumption of their causal role (Clarkson, Otto, Hirt, & Egan, 2016; Job,
Walton, Bernecker, & Dweck, 2013; Miller et al., 2012).
Research indicates that a limited-resource theory is dysfunctional in university
students’ everyday lives. When they face high demands, the self-regulation of students
endorsing a limited-resource theory is impaired, and they receive lower grades than students
with a nonlimited theory about willpower (Job et al., 2015). Moreover, implicit theories about
willpower affect well-being. For example, a longitudinal study demonstrated that the more
students endorsed a limited theory, the greater was their drop in well-being as demands
increased across an academic term (Bernecker, Hermann, Brandstätter, & Job, 2017).1
Although implicit theories about willpower do have a conceptual overlap with other
constructs that are related to successful self-regulation and agency (e.g., trait self-control, self-
efficacy, internal control beliefs), the core feature of these willpower beliefs is unique and is
not contained in the other constructs. While all of the other constructs focus on how
successful people believe that they typically are in reaching their goals or in their ability to
successfully execute the actions or behaviors that are necessary to reach a goal or to overcome
obstacles and difficulties in achieving a certain outcome, the main focus of willpower theories
are peoples’ beliefs about the nature of self-control capacity over time, that is, when self-
regulatory demands accumulate. Someone thus could feel highly efficacious in exerting self-
control on a specific demanding occasion (indicating high self-efficacy or internal control),
while at the same time believing that his/her capacity to exert control is limited or will
decrease over time. Accordingly, previous research showed that willpower theories are
predictive of outcomes over and above what is predicted by self-efficacy or trait self-control
in situations with high demands that are chronic and accumulate across longer time intervals
(e.g., Bernecker & Job, 2015a; Job, Walton, Bernecker, & Dweck, 2015).
Possible Age Differences in Willpower Theories
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 6
Whenever theories about willpower were measured in student populations (average
age ranging between 21 and 24 years), the mean for willpower beliefs was significantly
different from the scale midpoint suggesting that there was a general tendency to endorse a
limited-resource theory. So far, however, nothing is known about the development of implicit
theories about willpower across adulthood. Knowledge about age differences and about the
mediating mechanism of possible developmental change with regard to willpower theories
could advance our understanding not just of willpower theories and their development but it
could also provide important insights into age differences relating to self-regulation and
coping with demands (e.g., regarding developmental changes in processes like optimization,
compensation, tenacity, and assimilation; Brandtstädter, Wentura, & Greve, 1993; Freund &
Baltes, 1998; Heckhausen, 1997; Rothermund & Brandtstädter, 2003). The present research
was conducted to test possible competing assumptions on age differences and the
development of willpower theories across adulthood. Based on what we know so far, age
could be either associated with a more limited or a more nonlimited resource theory.
Changes in physical capacity. As they age, people experience physiological changes
that involve decrements in their physical capacity (e.g., muscular strength and endurance;
Freund, Nikitin, & Riediger, 2010; Kaiser, 2009). With increasing age, people lose the ability
to deal with stress and changes in their environment due to age-related reductions in
flexibility of the physiological system and the corresponding difficulty in down-regulating
arousal (for summary, see Charles, 2010). Older adults require prolonged phases to physically
recuperate from strenuous effort and they are more vulnerable to acute and chronic diseases
(Rockwood & Mitnitski, 2007; Spirduso, Francis, & McRae, 2005). Further, aging is
characterized by changes in cognitive functioning, specifically in biologically based cognitive
mechanics, such as processing speed and working memory (Lindenberger & Baltes, 1997;
Salthouse, 2004). After having grown rapidly during infancy and childhood, these functions
successively decline during adulthood and show an accelerated decline in very old age (Craik
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 7
& Bialystok, 2006; Taylor, 2005). In particular, executive control functions, which develop
relatively late, are assumed to decline early, according to a “last in, first out” principle (Raz,
2000).
In sum, with advancing age people are increasingly likely to be confronted with the
limited nature of their physiological and cognitive capacity and they experience these limits
specifically when they are confronted with high demands. Experiencing that important aspects
of their physiologically based functioning are declining might foster the perception of the
physiological bases of their mental capacities and contribute to the idea that their cognitive
capacity is, analogously to muscle strength, a limited resource that gets depleted whenever
they use it. Therefore, it is plausible to assume that with increasing age people are more likely
to endorse more limited theories about willpower. The present research tests whether there is
a positive relationship between age and a limited theory about willpower. This hypothesis
assumes that age-related decline in physical and cognitive functioning (i.e., health, Spiro &
Brady, 2011) is responsible for the development of a limited-resource theory.
Changes in Autonomy. Ageing is not only characterized by constant decline. If one
considers life before very old age (fourth age, 80+) there are aspects of life that may
considerably improve with increasing age. One such aspect is autonomy (Erikson, 1963; Ryff,
1995; Sheldon & Kasser, 2001), which is defined as sense of self-determination (Ryff &
Keyes, 1995). Typically, people gain decisional independence and autonomy in their careers
as they get older (Kanfer, Beier, & Ackerman, 2013; Kooij, De Lange, Jansen, Kanfer, &
Dikkers, 2011; Zacher, Rosing, Henning, & Frese, 2011). Once child-care duties ceased and
even more so after people passed the retirement line, people become free from external
obligations and constraints, and they gain free time and an amount of autonomy with regard
to how they want to spend their days that is unparalleled by any time of their adult life before
(the post-[re]productive freedom; Baars, 2012; Bildtgård & Öberg, 2017). Moreover, old age
is less regulated by age-related norms and expectations, providing more freedom to select the
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 8
goals one wants to set and pursue (Baltes, Reese, & Lipsitt, 1980). Accordingly, research
documented a positive relationship between age and experiencing autonomy with regard to
everyday tasks as well as personal goals (Sheldon, Houser-Marko, & Kasser, 2006; Sheldon,
Kasser, Houser-Marko, Jones, & Turban, 2005). We suggest that this gain in autonomy
promotes the development of a rather nonlimited theory about willpower as people age.
Why should autonomy be related to a nonlimited theory about willpower? As
autonomous behavior is consistent with one’s preferences, values, and interests, behavioral
regulation is more harmonious and less prone to conflict. In contrast, when people pursue
goals in an externally controlled way because they are pressured to do so (low autonomy)
they are more likely to experience behavioral conflict and tension. As a result, people feel
more drained and fatigued from pursuing goals non-autonomously (Ryan & Deci, 2008;
Sheldon & Kasser, 1995; Thayer & Moore, 1972). High autonomy should, therefore, promote
the experience that, even though some tasks might be difficult and challenging, one has the
energy to stay focused for as long as necessary. This experience, in turn, should promote a
more nonlimited view on willpower. Recent research provides preliminary support for this
assumption (Sieber, Flückiger, Mata, Bernecker, & Job, 2017). In two longitudinal studies,
autonomous goal striving during a period of four months predicted a change in willpower
theories towards a more nonlimited-resource theory. Moreover, an experimental study showed
that when people are reflecting on a task they were working on autonomously, they were
more likely to endorse a nonlimited theory as compared to people who reflected on working
on a task for non-autonomous reasons.
Given that autonomous goal striving promotes a nonlimited theory about willpower
and given that people in general tend to gain autonomy as they get older, age might be
positively associated with a nonlimited theory about willpower.
Present Research
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 9
A first, high-powered correlational study was conducted to test the relationship
between age and implicit theories about willpower. Theories about willpower were measured
in a community sample covering the age range from 18 to 83 years. In addition, this study
tested whether age and theories about willpower are related to overall health status. Since
Study 1 showed that age was positively related to a nonlimited theory about willpower,
Studies 2, 3a, and 3b were conducted with the aims of replicating the relationship found in
Study 1, and exploring autonomy as a possible developmental mechanism underlying this
relationship. Study 2 tested whether age-differences in willpower theories between a group of
younger vs. older adults would be related to differences in reported autonomous motivation
with regard to a recently completed strenuous mental task. Finally, experimental Studies 3a
and 3b were designed to test the proposed process explaining developmental change by
manipulating an autonomous vs. externally controlled mindset in samples of younger and
older adults. We expected that thinking of a situation where they did work on a strenuous task
non-autonomously would reduce participants’ endorsement of a nonlimited theory about
willpower. In contrast, reflecting on a task they worked on autonomously would enhance
participants’ nonlimited willpower theory.
Study 1
Study 1 tested whether there is an association between implicit theories about
willpower and age and whether such an association is related to age-related differences in
overall health status.
Method
This research was conducted consistent with the ethical guidelines of the University of
Zurich, Department of Psychology; all studies were considered exempt from formal ethical
review.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 10
Participants and procedure. A total of n = 880 participants (Mage = 49.61 years,
SDage = 16.65 years, range 18–85, 49% females) was recruited in Germany via Respondi AG
(www.respondi.com/en/). This market-research institution has a member community from
which target groups can be recruited for online studies according to personal panel data. Since
we aimed for an equal distribution across the full age range, participants were recruited in 20-
years steps, that means, n = 285 adults from 18 to 39 years, n = 293 from 40 to 59 years, and
n = 302 adults older 60 years. Seventy-six participants had to be excluded from the study
because they did not answer at least one of two control questions correctly (“For technical
purposes, please click on the ‘5’). The control questions, designed to catch participants who
simply click through the responses without actually reading the items, were placed randomly
throughout the questionnaire. In addition, two participants in the group between age 40 and
59 were excluded because they indicated an age outside of the specified age range. Thus, the
following analyses are based on a sample of n = 802 adults (48.3% females). Male and female
participants did not differ in the their mean age (males: Mage = 49.67, SDage = 16.20, females:
Mage = 48.87, SDage = 16.86), t(800) < 1, p = .494. Age was marginally associated with
highest education completed, F(2, 799) = 2.60, p = .075. Participants who reported
completion of a high school (n = 59) were significantly younger (Mage = 45.32, SDage = 17.43)
than those who completed vocational training (n = 511, Mage = 50.10, SDage = 16.45), Mdiff = -
4.78, SE = 2.27, p = .035. Participants with a university degree (n = 232, Mage = 48.48, SDage
= 16.29) did not differ with regard to their age from the other two groups (ps ≥ .189). Not
surprisingly, participants who indicated to have paid work (n = 433) were younger (Mage =
43.12, SDage = 11.53) than those who did not (n = 369, Mage = 54.53, SDage = 18.23), t(800) =
10.38, p < .001. In contrast, participants who were retired (n = 243) were on average older
(Mage = 67.13, SDage = 6.73) than those who were not (n = 559, Mage = 41.52, SDage = 13.14),
t(800) = 28.78, p < .001. There was no age difference between people who were unemployed
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 11
(n = 44, Mage = 47.04, SDage = 13.09) and the rest of the group (n = 758, Mage = 49.41, SDage =
16.69), t < 1, p = .361.
Participants answered demographic questions and completed a scale assessing beliefs
about willpower. Since this study was conducted as a small part of a larger research project
investigating age-related motivational underpinnings of daily social lives (see Nikitin &
Freund, 2018), participants completed various other questionnaires for a total of
approximately 30 minutes. Each participant was paid €3.
Measures. Implicit theories about willpower were measured using an 8-item
shortened version of the original 12-item scale (Job et al., 2010; see supplement for full list of
questions used in this research). Participants answered items such as “After a strenuous
mental activity my energy is depleted and I must rest to get it refueled again” (limited-
resource theory) and “My mental stamina fuels itself; even after strenuous mental exertion I
can continue doing more of it” (nonlimited-resource theory) on a scale from 1 (strongly
disagree) to 7 (strongly agree). Items concerning the limited-resource theory were reverse-
scored, therefore a high value represents a higher agreement with a nonlimited resource
theory (
a
= .72, M = 4.27, SD = 0.98). Further, participants rated their overall physical health
on a 7-point scale ranging from 1 (very bad) to 7 (very good) (M = 4.73, SD = 1.43).
Results and Discussion
Table 1 depicts correlations between the main study variables. As expected, age was
negatively correlated with health (r = -.210, p < .001). Health, on the other hand, was
positively correlated with a nonlimited willpower theory (r = .252, p < .001). This correlative
pattern seems to be consistent with the assumption that an age-related deterioration in health
might lead to a more limited theory about willpower. However, if this is the case, it is
overturned by other, counteracting processes. The correlation between age and theory about
willpower was positive, r = .237, p < 0.001 (see Figure 1). The older participants were, the
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 12
more likely they were to endorse a nonlimited willpower theory. A regression analysis
controlling for health confirmed the positive relationship between age and a nonlimited
willpower theory, b = .02, SEb = .002, t = 9.08, p < .001. Further, the test for a quadratic
relationship between age and willpower theory (controlling for the linear trend) was non-
significant, t = -1.55, p < .121.
The results suggest that, even though people experience physical decline as they get
older and despite the fact that poor health is related to a more limited theory about willpower,
older participants might be more likely to believe that their willpower is not limited. We
suggest that the positive relationship between age and a nonlimited theory about willpower
can be explained by increasing autonomy people experience in their everyday life as they get
older. Studies 2 and 3a/b were conducted to test this assumption.
Study 2
Study 2 tested whether older adults would be more likely to be autonomously
motivated on strenuous mental tasks they work on in their everyday life as compared to
younger adults. To test this assumption, we asked participants to recall a recent situation
where they worked on a mentally demanding task. We decided to use this idiosyncratic
approach to maximize equivalence in perceived demands across all participants. Confronting
young and old adults with the same situation (e.g., a task in the laboratory) often results in
age-dependent differences in how this situation is perceived (e.g., typical paradigms assessing
emotion regulation and recognition poorly match the way these processes occur in older
adults; Kunzmann & Richter, 2009; Rohr, Wieck, & Kunzmann, 2017; Streubel &
Kunzmann, 2011). If we had confronted all our participants with one standardized situation, it
would have been likely that older and younger participants would have perceived it as
differently demanding. Therefore, we decided against using a standardized task for all
participants. In our study, although participants in the different age groups potentially
remembered different activities from different domains, the critical dimension––perceived
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 13
mental demands––was held constant. We hypothesized that older adults would indicate more
autonomy concerning the recalled mentally demanding task and that the perceived autonomy
would be associated with stronger endorsement of a nonlimited theory about willpower.
Method
Participants. Participants (U.S. residents) were recruited via Amazon’s Mechanical
Turk (MTurk) to participate in a study on how people work on different tasks in exchange for
$0.50. Two panels were created (18–35 years and 60+ years)2 with the assistance of
TurkPrime (Litman, Robinson, & Abberbock, 2016). Out of 546 persons who agreed to
participate, n = 423 completed the study. Out of those, n = 21 participants had to be excluded,
as they were not part of the required age-groups, which led to a sample size of n = 402.
Additionally, n = 7 participants had to be excluded, as they did not indicate a concrete task,
which led to a total sample size of n = 395 (56% females, Mage = 43.37, SDage = 18.92, range:
18–98 years). Based on the correlation between age and willpower theory in Study 1 (r =
.237) we aimed for a sample with high power (1 − β = 0.95) to detect a medium to small
effect size.
The subsample of young adults (n = 233) consisted of 52% females (Mage = 28.09,
SDage = 4.03, range: 18–35 years). The sample was diverse concerning working status (61.4%
were working fulltime, 19.3% worked part-time, 9.4% were looking for work, 7.3% were full
time students, 2.1% were part time students and one person indicated to be retired). The
subsample of older adults (n = 162) consisted of 62% females (Mage = 65.35, SDage = 5.36,
range: 60–98 years). Thereby, 26.5% were working fulltime, 16% worked part-time, 4.9%
were looking for work, 51.2% were retired, and 1.2% were part time students.
The age groups did not significantly differ in the income distribution, c2(5) = 4.71, p =
.453. Moreover, there was no difference between the age groups in the distribution of gender,
c2(2) = 4.25, p = .119, whereas the distribution of working status was statistically significant,
c2(5) = 155.99, p < .001.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 14
Procedure. After giving informed consent and reporting demographic information,
participants were asked to think about a mentally demanding task they had worked on lately.
Once they had remembered a task, they were asked to indicate how mentally demanding the
task had been on a scale ranging from 1 = not at all to 10 = very much (M = 7.94, SD = 1.37).
On a next page, participants were instructed to quickly describe the situation. Finally,
participants answered questions concerning autonomous goal striving and their implicit
theories about willpower.
Autonomous goal striving. Following the procedure proposed by Sheldon and Elliot
(1999) to measure autonomy of self-set goals, participants were asked to indicate the reasons
for working on the mentally demanding task they had recalled. Autonomous motivation was
reflected by two statements (“I worked on this task because it gives me pleasure and
enjoyment” and “I worked on this task because I believe it is important”) as well as the
externally controlled motivation (“I worked on this task because I would feel bad [guilty,
ashamed or anxious] otherwise” and “I worked on this task because it is requested or the
situation requires it”). Participants provided their agreement on a scale ranging from 0
(strongly disagree) to 5 (strongly agree). Following previous research (Sheldon et al., 2004),
a relative autonomy score was computed for each participant by subtracting the scores for
externally controlled goal striving from the scores for autonomous goal striving (M = 0.62,
SD = 4.05).
Theories about willpower. Implicit theories about willpower were measured with the
full twelve-item scale developed by Job and colleagues (2010). The questions were answered
on a 6-point scale (1= strongly disagree; 6 = strongly agree). Items concerning the limited-
resource theory were reverse-scored, therefore a high value represents a higher agreement
with a nonlimited resource theory (
a
= .82, M = 3.53, SD = 0.77).
Results and Discussion
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 15
The two age groups did not differ with regard to perceived difficulty of the strenuous
mental task they had remembered, t(393) = -1.075, p = .283. Replicating the results from
Study 1, younger participants were less likely to endorse a nonlimited theory about willpower
(M = 3.45, SD = 0.78) as compared to older participants (M = 3.64, SD = 0.75), t(393) = -
2.49, p = .013, d = -0.25, 95%CI [-0.351, -0.041]. Further, the two groups differed in the
strength of experienced autonomy. Younger participants reported less autonomy (M = 0.22,
SD = 3.66) as compared to older participants (M = 1.18, SD = 4.51), t(299) = -2.23, p = .026,
d = -0.23, 95%CI [-1.798, -0.114]. Finally, there was a significant correlation between
autonomy and implicit theories about willpower, r = .286, p < .001. The more autonomy
participants experienced with regard to the task they had recalled, the more likely they were to
endorse a nonlimited theory about willpower.
Study 2 replicates the results of Study 1 by documenting a positive relationship
between age and a nonlimited theory about willpower. Participants in the older age group
were more likely to believe that willpower was nonlimited as compared to participants in the
younger age group. Further, the results of Study 2 indicate that older adults engage in
strenuous mental activities in their everyday life with a more autonomous motivation as
compared to younger adults. Perceived autonomy, in turn, was positively related to a
nonlimited theory about willpower. This finding is consistent with our hypothesis that people
develop a nonlimited theory about willpower as they gain autonomy across adulthood.
However, since Study 2 is merely correlational, it cannot provide any definite
conclusions about causal relationships and developmental processes (Lindenberger, von
Oertzen, Ghisletta, & Herzog, 2011). Therefore, we took an experimental approach in Studies
3a and 3b and manipulated the postulated age-related process (as suggested by Freund &
Isaacowitz, 2013).
Studies 3a and 3b
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 16
Studies 3a and 3b were conducted to test the causal effect of autonomy on implicit
theories about willpower and their age-related differences. Based on previous research (Sieber
et al., 2017), we expected that high (as compared to low) autonomy makes people believe that
willpower is a nonlimited resource. If autonomy is the explaining factor of age-related
differences in beliefs about willpower, then manipulating autonomy experimentally should
also affect people’s willpower theories. Accordingly, assigning older and younger people to
autonomy conditions that run counter to prevalent age differences in autonomy should reduce
or eliminate age differences in willpower and should make old and young people more similar
in their respective beliefs: Specifically, inducing low autonomy in the group of older adults
should lead them to report levels of beliefs about willpower that are more similar to what
younger adults report. Similarly, letting young adults reflect on a demanding activity they
completed autonomously should enhance their feelings of autonomy and thereby also increase
the nonlimited theory in young adults, again making them more similar to older adults with
regard to their willpower beliefs. These hypotheses were tested in a 2 (age: young, old) ´ 2
(autonomy: low, high) between-subjects design. We expected to find two main effects of age
and autonomy.
Both studies followed the exactly same procedure making Study 3b basically a
replication of Study 3a. The only difference was that data were sampled with two different
online recruiting platforms. While we used a sample from Mechanical Turk in Study 3a, in
Study 3b we used a sample from Prime Panels provided by TurkPrime (Litman et al., 2016), a
very large and representative online panel allowing us to target participants aged above 60
years.
Participants Study 3a
Data were collected online via Mechanical Turk using the TurkPrime tool (Litman et
al., 2016) to recruit participants of pre-specified age ranges (≤ 35 years, ≥ 60 years). Again,
we selected a sample with high power (1 − β = 0.95) to detect an assumingly medium to small
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 17
effect size. Participants were n = 226 young (age M = 28.15 years, SD = 3.83 years, range 19–
35, 49% females) and n = 155 older adults (age M = 64.72, SD = 4.92, range 60–87, 67%
females). The difference in the gender distribution between the age groups was statistically
significant, c2(1) = 11.22, p = .001. Within the younger group, 65.5% participants worked
full-time, 16.8% worked part-time, 8.4% looked for work, 8.9% were full- or part-time
students, and one was retired. Within the older group, 26.5% participants worked full-time,
24.5% worked part-time, 4.5% looked for work, none was student, and 44.5% were retired.
The age differences in the distribution of working status were statistically significant, c2(5) =
143.94, p < .001. The age groups did not significantly differ in the income distribution, c2(5)
= 5.90, p = .316.
Participants Study 3b
Data were collected online via TurkPrime (Litman et al., 2016). We selected a sample
with high power (1 − β > 0.95) to detect an assumingly medium to small effect size.
Participants were n = 242 young (age M = 27.60 years, SD = 5.11 years, range 18–35, 62%
females) and n = 255 older adults (age M = 67.17, SD = 5.38, range 60–86, 64% females).
The difference in the gender distribution between the age groups was statistically non-
significant, c2(1) = 1.05, p = .591. Within the younger group, 47.9% participants worked full-
time, 15.3% worked part-time, 21.1% looked for work, 13.2% were full- or part-time
students, and 2.5% were retired. Within the older group, 13.4% participants worked full-time,
6.3% worked part-time, 3.5% looked for work, one older participant was a part-time student,
and 76.4% were retired. The age differences in the distribution of working status were
statistically significant, c2(5) = 288.5, p < .001. Unlike Study 3a, the age groups did
significantly differ in the income distribution, c2(4) = 10.42, p = .034. Among young
participants 19.2% earned less than $20.000 (compared to 10% older participants).
Procedure Studies 3a and 3b
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 18
Participants were randomly assigned to one of two experimental conditions defined by
autonomy. In both conditions, participants were instructed to think about a mentally
demanding task they worked on lately. In the high-autonomy condition (3a: n = 188; 3b: n =
264), participants should think about a task they worked on because of the mere enjoyment it
provided to them. In the low-autonomy condition (3a: n = 193; 3b: n = 233), they should think
about a task they worked on because someone else wanted them to or because it was required
by the situation. All participants were then asked to briefly describe the task. Subsequently,
participants provided information on how mentally demanding the task was (“How mentally
demanding would you rate the task on a scale from 1 to 10?” 1 = not at all, 10 = very much;
Study 3a: M = 7.82, SD = 1.48; Study 3b: M = 7.24; SD = 2.10). The question on perceived
mental demands of the task was included to ensure that the results were not driven by reports
of differently demanding tasks in the groups. In addition, the autonomy subscale from the
balanced measure of psychological needs was administered to measure whether the
manipulation was successful (BMPN, Sheldon & Hilpert, 2012). Participants were asked to
report their autonomy feelings while working on the task, serving as a manipulation check
(six items, e.g., “While working on that task, I was free to do things the way I want to”, 1 =
strongly disagree, 6 = strongly agree; Study 3a: a = .82, M = 4.40, SD = 1.16; Study 3b: a =
.77, M = 4.43; SD = 1.08). Finally, participants reported their willpower theories. These were
assessed with the same 12 items used in Study 2 (Study 3a: a = .86, M = 3.61, SD = 0.86;
Study 3b: a = .80, M = 3.62, SD = 0.80). Higher values represent a more nonlimited theory
about willpower.
Results Study 3a
Perceived autonomy. First, we tested if the two autonomy conditions affected
participants’ reports of experienced autonomy (manipulation check). This was the case.
Participants in the high-autonomy condition reported higher levels of autonomy (M = 4.92, SE
= 0.06) than participants in the low-autonomy condition (M = 3.90, SE = 0.09), t(379) = 9.46,
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 19
p < .001, d = 0.66. There was no interaction with age-group, F < 1, p = .780, suggesting that
the manipulation was equally effective for young and for older adults. Importantly, supporting
our assumption about age-related differences in autonomy, older participants reported more
autonomy (M = 4.69, SE = 0.09) as compared to younger participants (M = 4.21, SE = 0.08),
t(379) = 4.06, p < .001, d = 0.66. Figure 2 shows the means in perceived autonomy by
experimental conditions and age groups.
Willpower theories. To test the hypothesis that beliefs about willpower differ as a
function of autonomy manipulation and age (two main effects), we ran a 2 (autonomy: low,
high) ´ 2 (age group: young, old) between-subjects ANOVA. As hypothesized, the autonomy
manipulation had a main effect on willpower theories, F(1, 377) = 12.73, p < .001, d = 0.37.
Participants in the high-autonomy condition reported a more nonlimited theory (M = 3.79, SE
= 0.06) than participants in the low-autonomy condition (M = 3.47, SE = 0.06). Age was also
related to theories about willpower, F(1, 377) = 8.54, p = .004, d = 0.30. As expected, older
adults reported a more nonlimited theory (M = 3.76, SE = 0.07) than younger adults (M =
3.50, SE = 0.06). Unexpectedly, there was a marginal Age Group ´ Autonomy Condition
interaction, F(1, 377) = 3.81, p = .052, d = 0.29. As shown in Figure 2, the difference in
willpower theories between the age groups was significant in the high-autonomy condition,
t(146) = -3.43, p = .001, d = 0.50, but not in the low-autonomy condition, t(191) < 1, p = .491.
In addition, older adults endorsed a nonlimited theory in the high-autonomy more than in the
low-autonomy condition, t(153) = 3.66, p < .001, d = 0.59, whereas there was no significant
difference for the younger participants between the two conditions, t(224) = 1.25, p = .213.3
Task demands. In order to rule out that the results are driven by differently
demanding tasks, we tested whether the groups differed in their reports on how demanding
the reported task was. A 2 (autonomy: low, high) ´ 2 (age group: young, old) between-
subjects ANOVA revealed a marginal effect of the autonomy manipulation (low autonomy: M
= 7.98, SE = 0.11, high autonomy: M = 7.69, SE = 0.11), F(1, 377) = 8.02, p = .054, d = 0.20.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 20
However, there was no significant age effect (younger group: M = 7.71, SE = 0.10, older
group: M = 7.95, SE = 0.12), F(1, 377) = 2.11, p = .147, nor a significant Age Group ´
Autonomy Condition effect, F(1, 377) = 1.07, p = .301. Based on the difference in the
perceived task demands between the two experimental groups, we included task demands as a
covariate and ran the main analysis once more. Perceived demands had no effect on theories
about willpower, F(1, 376) < 1, p = .992, and the pattern of results of the main analysis (i.e.,
age and autonomy as predictors of beliefs about willpower) did not change after including
perceived demands as a covariate (autonomy condition: F[1, 376] = 12.58, p < .001, d = 0.36;
age group: F[1, 376] = 8.46, p = .004, d = 0.30; Age Group ´ Autonomy Condition: F[1, 376]
= 3.79, p = .052, d = 0.20).
Discussion. Study 3a provides evidence for a causal effect of autonomy on the
adoption of a nonlimited theory about willpower. As a main effect, participants who were
instructed to think of a task they had worked on autonomously, reported more of a nonlimited
theory about willpower. Further, Study 3a replicated the core finding of this research: As a
main effect, older adults were more likely to endorse a nonlimited theory about willpower as
compared to younger adults. This difference, however, disappeared, when older (and
younger) participants had been brought to think about a strenuous mental task they had
worked on under conditions of low autonomy. Under these conditions, their endorsement of a
nonlimited theory decreased and was comparable to the willpower theory of younger adults.
Unexpectedly, thinking of an autonomous task did not lead to a substantial increase in non-
limited willpower beliefs in younger adults. One explanation for this lack of an effect for the
younger participants could be that their low levels of autonomy with regard to strenuous
mental tasks in everyday life might undermine the impact of such a memory based
manipulation. However, since we did not predict this result and since the interaction was only
a statistical trend, it was important to test whether this result would replicate in Study 3b
before further interpreting this finding.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 21
Results Study 3b
Perceived autonomy. Participants in the high-autonomy condition reported higher
levels of autonomy (M = 4.67, SE = 0.06) than participants in the low-autonomy condition (M
= 4.13, SE = 0.07), t(493) = 5.75, p < .001, d = 0.52. Further, older participants reported
significantly higher autonomy (M = 4.69, SE = 0.06) as compared to younger participants (M
= 4.15, SE = 0.07), t(493) = 5.70, p < .001, d = 0.51. There was no interaction with age-
group, F < 1, p = .574. These results suggest that the manipulation was effective and they
support the assumption on age-related differences in autonomy (see Figure 3).
Willpower theories. To test the hypothesis that beliefs about willpower differ as a
function of autonomy manipulation and age (two main effects), we ran a 2 (autonomy: low,
high) ´ 2 (age group: young, old) between-subjects ANOVA. As hypothesized, the autonomy
manipulation had a main effect on willpower theories, F(1, 493) = 6.76, p = .010, d = 0.24
(see Figure 3). Participants in the high-autonomy condition reported a more nonlimited theory
(M = 3.71, SE = 0.05) than participants in the low-autonomy condition (M = 3.53, SE = 0.05).
Age was also related to theories about willpower, F(1, 493) = 14.87, p < .001, d = 0.35. As
expected, older adults reported a more nonlimited theory (M = 3.76, SE = 0.05) than younger
adults (M = 3.48, SE = 0.05). As opposed to Study 3a, there was no Age Group ´ Autonomy
Condition interaction, F(1, 439) < 0.5 p = .716.4
Task demands. Neither the autonomy conditions nor the age groups differed in how
demanding the task was they had thought about (F < 0.5 for the main effects). There was only
a marginal interaction between autonomy condition and age group, F(1, 493) = 2.75, p = .098,
d = 0.16. Within the high autonomy condition, there was a trend for older participants to think
of a more demanding task than younger participants, F(1, 493) = 3.13, p = .077, d = 0.16.
Including task demands as a covariate in the main analysis did not change the pattern of
results on theories about willpower (autonomy condition: F[1, 492] = 6.86, p = .009, d = 0.24;
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 22
age group: F[1, 492] = 15.27, p < .001, d = 0.35; Age Group ´ Autonomy Condition: F[1,
492] < 0.5, p = .798).
Discussion. Study 3b replicated the main findings of Study 3a by showing that a
manipulation of autonomy changes people’s theories about willpower. Participants who were
led to think of a demanding task they had worked on for autonomous reasons were more
likely to endorse a nonlimited theory about willpower than participants who were led to think
of a demanding task they had worked on for controlled reasons. This result provides evidence
for a causal effect of autonomy on the adoption of a nonlimited theory about willpower.
Further, Study 3b––again––confirmed the results from the previous studies on age differences
in theories about willpower and perceived autonomy. Older adults across both autonomy
conditions reported to have felt more autonomous when working on the demanding tasks as
compared to younger participants and they were more likely to endorse a nonlimited theory
about willpower than younger participants. Reducing older participants’ autonomy as well as
enhancing younger participants’ autonomy made the two age groups more similar in their
nonlimited willpower theory endorsement. This pattern of results supports our assumption
that one factor underlying the age differences in willpower theories are age-related differences
in perceived autonomy.
Study 3b did not replicate the Autonomy Condition ´ Age Group interaction
documented in Study 3a. Given that this interaction was (a) not hypothesized, (b) only
marginally significant, and (c) that it did not replicate in Study 3b, we assume that it was
caused by chance and/or by characteristics of the one specific sample. Moreover, a joint
analysis with a data-set combing both studies confirmed the two main effects (autonomy
condition: F[1, 873] = 17.07, p < .001, d = 0.28; age group: F[1, 873] = 23.15, p < .001, d =
0.33), whereas the Age Group ´ Autonomy Condition interaction was nonsignificant, F(1,
873) = 1.11, p = .293.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 23
General Discussion
This research started out testing two competing assumptions: with higher age, people
could be either less or more likely to believe that willpower is a nonlimited resource. The first
hypothesis postulated that due to a decline in physical functioning, older people might be
more likely to adopt a limited theory about willpower. Although there was a negative
relationship between age and health and also between health and a nonlimited theory about
willpower, older adults were still, overall, more likely to endorse a nonlimited theory about
willpower. This pattern of results suggests that even though there might be an effect of
decreasing functioning on willpower theories, it is counteracted by a second ageing-related
process that promotes more of a nonlimited-resource theory about willpower with increasing
age.
Studies 2 and 3a/b tested the process accounting for this relationship. On a
correlational basis, Study 2 showed that age and willpower theories are related to perceived
autonomy in a way that is consistent with the assumption that enhanced autonomy in older
adults promotes a nonlimited willpower theory. When thinking about a recent strenuous
mental task, older participants reported more autonomous motivation as compared to younger
participants. They indicated that they had worked on the mentally demanding task because it
was fun or because it was important to them more than younger participants did. In turn, the
more autonomous motivation participants reported, the stronger was their endorsement of a
nonlimited theory about willpower.
Experimental Studies 3a and 3b provided causal evidence for the suggested process by
manipulating the mechanism. On the one hand, when older participants were instructed to
reflect on a task they had been forced to work on (low autonomy), the mediating mechanism
that promotes a nonlimited theory was blocked, and they became more similar to younger
participants in their willpower theories. On the other hand, when younger participants were
made to reflect on a demanding task they worked on because it was fun (high autonomy), they
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 24
showed enhanced endorsement of a nonlimited willpower theory and thus became more
similar to older participants. This effect, however, emerged in Study 3b only. In Study 3a,
young participants were less responsive to the high autonomy manipulation. These results
suggest that it might not always be possible to induce feelings of autonomy in people with
generally low autonomy experiences. If it is difficult for them to remember a high autonomy
situation such a memory-based manipulation might run on empty. In contrast, reducing older
participants’ autonomy by evoking memories of controlled motivation reliably resulted in
reduced endorsement of a nonlimited theory. Taken together, these results suggest that
increasing age is associated with greater autonomy on strenuous tasks in everyday life and
that this autonomy, in turn, promotes a nonlimited theory about willpower.
This research is consistent with previous theorizing and findings on changes in
autonomy across the lifespan. One positive aspect of getting older is that, up to a certain
point, one gains the freedom to decide on how one wants to spend time and what tasks and
goals one wants to pursue (Erikson, 1963; Ryff, 1995; Sheldon & Kasser, 2001). With
increasing years in a company, one typically gains seniority and status, which can result in
more autonomy in structuring one’s own and others’ work (Kanfer et al., 2013; Kooij et al.,
2011; Zacher et al., 2011). Even more impactful is retirement, which typically leaves people
with much more free time on their disposal (Costa, 1998; Gauthier & Smeeding, 2003; Weiss,
2005). Based on Sieber and colleagues work (2017), we have argued that experiencing a
greater amount of autonomy when pursuing demanding tasks makes people endorse a
nonlimited theory. Working on strenuous tasks in a self-chosen way because it is personally
important or fun is more likely to be experienced as energizing than working on such a task
for external reasons (e.g., out of obligation or because a boss or the studies demand it). This
experience of having the energy available to stay focused on a task, as suggested by the
present research, makes people think of their willpower as a nonlimited resource.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 25
The present findings are compatible with the observation that, although aging people
are confronted with declining physical and cognitive functioning, they are for a relatively long
time (typically until they reach their eighties) able to handle these challenges very well
(Brandtstädter, Rothermund, & Schmitz, 1998; Ebner & Freund, 2007; Heckhausen, 2005).
For example, they are still successful in pursuing their personal goals (Ebner, Freund, &
Baltes, 2006; Freund et al, 2010; Ryff, 1991; Sheldon & Kasser, 2001). Changes in autonomy
and theories about willpower can be a key to an understanding of successful self-regulation
and coping in old age. We propose that they mutually influence and are influenced by these
processes.
On the one hand, increased freedom and autonomy can be seen as an important
prerequisite for a flexible selection of goals and evaluative standards. Enhanced autonomy
allows individuals to focus on those areas in which they maintained high levels of functioning
and to disengage from others in which success has become improbable. This explains why
older adults can maintain a sense of personal control over important outcomes despite losses
and declines, for example, by processes of secondary control, re-orientation, or elective
selectivity (Brandtstädter & Rothermund, 1994; Freund & Baltes, 1998; Heckhausen, 1997).
These accommodative coping processes probably mediate the relation between autonomy and
nonlimited theories. Focusing on what is feasible and important may prevent older people
from getting stuck in barren commitments and wasting their energy in futile endeavors.
On the other hand, developing and endorsing a nonlimited implicit theory of
willpower with advancing age also fosters attempts to tenaciously adhere to important
personal goals and projects, especially when goal pursuit becomes more challenging.
Nonlimited theories then form the basis for enduring effort investment, compensation, and
assimilative tenacity (Baltes & Baltes, 1990; Heckhausen & Schulz, 1995; Rothermund &
Brandtstädter, 2003). All these strategies enable older adults to maintain meaningful and
stable life goals, control, and purpose that, in turn, promote older adults’ cognitive, mental,
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 26
and physical functioning (e.g., Gall, Evans, & Howard, 1997; Nimrod; 2007; Lövdén,
Ghisletta, & Lindenberger, 2005; Robbins, Lee, & Wan, 1994). Future research may
investigate whether the endorsement of a nonlimited willpower theory positively predicts self-
regulation as people age and how far it enables them to compensate for age-related decline in
physical and cognitive functioning.
In the present research, older participants mostly belonged to the category of “young
old” also called the “third age” (Laslett, 1989). For this phase, experiences of increased
autonomy and freedom are typical for most old people. In modern societies, however, aging
does not end with the third age. Instead, many people face and expect a phase of increasing
functional decline that is called “the fourth age”. During this period, active forms of goal
pursuit and coping reach their limits, and it becomes more and more difficult to maintain
autonomy and to ward off dependency (Baltes, 1997). The changes that occur during the
fourth age probably lead to a completely new outlook on what is important in life, and what
people care for when they have to cope with life’s finitude (Brandtstädter et al., 2010;
Tornstam, 1997). Although speculative and not based on evidence, we would assume that
nonlimited theories of willpower decline again during this last phase of life in face of those
challenges, when declining health impairs autonomy.
Relatedly, the results of Study 1 might further suggest that the positive development
with regard to a nonlimited theory about willpower might reach its limits once health-related
issues become heavy. Bad health was associated with the endorsement of a more limited
theory about willpower. However, the causal relationship between health and willpower
theory remains unclear. A limited theory about willpower might cause an earlier decline in
health due to worse health behavior shown by people who believe that willpower is a limited
resource (Job et al., 2010; 2015). For example, a limited theory about willpower was shown to
be related to poor therapy adherence in patients with Type 2 Diabetes (Bernecker & Job,
2015b). Still, it is also plausible––as we suggested––that very poor health can promote the
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 27
belief that willpower is limited, possibly causing a vicious cycle of declining health and
enhanced limited willpower theory.
Limitations and Future Directions
Correlational findings on age differences in willpower theories cannot rule out the
possibility of cohort effects. It is possible that within the last forty years there was a change in
societal values emphasizing recreation and wellness as an important part of life. This might
have shaped the belief – mostly in younger generations – that a person’s mental capacity is a
limited resource that requires adequate rest and recovery after stressful episodes. Such an
alternative explanation of the age group differences in willpower beliefs can be tested by
longitudinal approaches only. However, Studies 3a and 3b of the present research provided
positive causal evidence for the effect of autonomy on theories about willpower. In
combination with previous research and theorizing on age-related increases in autonomy
(Baltes et al., 1980; Bildtgård & Öberg, 2017; Kanfer et al., 2013), we can conclude that age-
related increases in autonomy are one mechanism that explains age-related differences in
willpower beliefs.
Secondly, we would like to emphasize that although the age differences in perceived
autonomy were robust across three studies, the effects were overall of small to medium effect
size (r = .12 in Study 2; d = 0.66 in Study 3a; d = 0.51 in Study 3b). It is therefore possible
that, in addition to autonomy, other so far unknown processes are involved in producing age
differences in willpower theories. One possibility could be that with higher age, people
acquire self-control strategies or skills that enable them to exert self-control for longer periods
of time and, accordingly, develop a more nonlimited theory about willpower. For example,
research on emotion regulation suggests that older adults use better emotion-regulation
strategies that enable them to regulate their emotions more successfully than younger adults
(e.g., Charles, 2010). Similarly, older adults typically report higher levels of satisfaction with
their social relationships than younger adults because they more successfully engage in
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 28
strategies that optimize positive social experiences and minimize negative ones (e.g., Luong,
Charles, Fingerman, 2011). Alternatively, due to an age-related increase in accommodative
flexibility (e.g., Brandtstädter, Wentura, & Greve, 1993) older adults may be better in
downgrading the importance of goals over which they have only limited control and to focus
on life domains that they can influence (Brandtstädter & Rothermund, 1994). Such an
accommodative reorientation to domains where personal control is high and avoiding
domains where personal control over outcomes is low might also foster the experience that
willpower and control resources are non-limited.
Finally, the present research relied solely on online sampling procedures. This allowed
the efficient recruitment of large samples providing all reported studies with high power.
However, we cannot be sure about the representativeness of our samples for the old and
young population. It is possible that specifically older participants who enjoy mental work
(high autonomy for demanding tasks) and who have a more nonlimited theory about
willpower are more likely to enroll at the sampling platforms (Respondi, Mechanical Turk,
TurkPrime) that we used for data collection. Future research should aim for representative
samples to replicate the relationship between age and implicit theories about willpower.
Conclusion
Ageing is often seen as dominated by a gradual decline in physical and mental
capacity. It is plausible that losses in health and reduced resilience to demands make people
perceive their willpower as a limited resource, which could, in turn, further undermine their
self-regulation and overall functioning. The present research, indeed, provides some evidence
for a positive relationship between health and a nonlimited willpower theory. However, we
find that despite declines in health, higher age is associated with higher endorsement of a
nonlimited willpower theory related to increased autonomy in older adults. This result sheds
light on positive developments people experience as they age. Most people will be
increasingly free in choosing activities and goals they want to pursue as they get older. Such
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 29
enhanced autonomy promotes feelings of nonlimited willpower capacity, which, in turn, may
further contribute to the maintenance of overall functioning and well-being in old age.
Importantly, people do not have to wait necessarily until they get old to have more autonomy.
The present research suggests that creating conditions that promote autonomous motivation
(e.g., at the workplace) can positively affect theories about willpower in all ages.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 30
References
Baars, J. (2012). Critical turns of aging, narrative and time. International Journal of Ageing
and Later Life, 7, 143–165. doi:10.3384/ijal.1652-8670.11171
Baltes, P. B. (1997). On the incomplete architecture of human ontogeny: Selection,
optimization, and compensation as foundation of developmental theory. American
Psychologist, 52, 366-80. doi:10.1037/0003-066X.52.4.366
Baltes, P. B., Reese, H. W., & Lipsitt, L. P. (1980). Life-span developmental psychology.
Annual Review of Psychology, 31, 65-110.
Baltes, P. B., & Baltes, M. M. (1990). Psychological perspectives on successful aging: The
model of selective optimization with compensation. In P. B. Baltes & M. M. Baltes
(Eds.), Successful aging: Perspectives from the behavioral sciences (pp. 1–34). New
York, NY: Cambridge University Press.
Baumeister, R. F., Vohs, K. D., & Tice, D. M. (2007). The strength model of self-control.
Current Directions in Psychological Science, 16, 351–355. doi: 10.1111/j.1467-
8721.2007.00534.x
Bernecker, K., Herrmann, M., Brandstätter, V., & Job, V. (2017). Implicit theories about
willpower predict subjective well-being. Journal of Personality, 85, 136–150.
doi:10.1111/jopy.12225
Bernecker, K., & Job, V. (2015a). Beliefs about willpower moderate the effect of previous
day demands on next day’s expectations and effective goal striving. Frontiers in
Psychology, 6:1496. doi:10.3389/fpsyg.2015.01496
Bernecker, K., & Job, V. (2015b). Beliefs about willpower are related to therapy adherence
and psychological adjustment in patients with type 2 diabetes. Basic and Applied Social
Psychology, 37, 188–195. doi:10.1080/01973533.2015.1049348
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 31
Bildtgård, T., & Öberg, P. (2017). Time as a structuring condition behind new intimate
relationships in later life. Ageing and Society, 35, 1505–1528.
doi:10.1017/S0144686X14000452
Brandtstädter, J., & Rothermund, K. (1994). Self-percepts of control in middle and later
adulthood: Buffering losses by rescaling goals. Psychology and Aging, 9, 265–273.
doi:10.1037/0882-7974.9.2.265
Brandtstädter, J., Rothermund, K., Kranz, D., & Kühn, W. (2010). Final decentrations:
Personal goals, rationality perspectives, and the awareness of life’s finitude. European
Psychologist, 15, 152–63. doi:10.1027/1016-9040/a000019
Brandtstädter, J., Rothermund, K., & Schmitz, U. (1998). Maintaining self-integrity and self-
efficacy through adulthood and later life: The adaptive functions of assimilative
persistence and accommodative flexibility. In J. Heckhausen & C. S. Dweck (Eds.),
Motivation and self-regulation across the life span (pp. 365–388). New York, NY:
Cambridge Uninversity Press.
Brandtstädter, J., Wentura, D., & Greve, W. (1993). Adaptive resources of the aging self:
Outlines of an emergent perspective. International Journal of Behavioral Development,
16, 323–49.
Charles, S. T. (2010). Strength and vulnerability integration: A model of emotional well-
being across adulthood. Psychological Bulletin, 136, 1068–1091. doi:10.1037/a0021232
Chow, J. T., Hui, C. M., & Lau, S. (2015). A depleted mind feels inefficacious: Ego-depletion
reduces self-efficacy to exert further self-control. European Journal of Social
Psychology, 45, 754–768. doi:10.1002/ejsp.2120
Clarkson, J. J., Otto, A. S., Hirt, E. R., & Egan, P. M. (2016). The malleable efficacy of
willpower theories. Personality and Social Psychology Bulletin, 42, 1490–1504.
doi:10.1177/0146167216664059
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 32
Costa, D. L. (1998). The evolution of retirement: An American economic history, 1880–1990.
University of Chicago Press.
Craik, F., & Bialystok, E. (2006). Cognition through the lifespan: Mechanisms of change.
Trends in Cognitive Sciences, 10, 131–138. doi:10.1016/j.tics.2006.01.007
Deci, E. L., & Ryan, R. M. (2008). Self-determination theory: A macrotheory of human
motivation, development, and health. Canadian Psychology/Psychologie Canadienne,
49, 182–185. doi :10.1037/a0012801 !
Dweck, C. S. (1996). Implicit theories as organizers of goals and behavior. In Gollwitzer, P.
M., & Bargh, J. A.: The Psychology of action: Linking cognition and motivation to
behavior. New York, NY: The Guilford Press.
Dweck, C. S. (1999). Self-Theories. Lillington, NC: Taylor & Francis.
Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and
personality. Psychological Review, 95, 256–273. doi:10.1037//0033-295X.95.2.256
Ebner, N. C., & Freund, A. M. (2007). Personality theories of successful aging. In J. A.
Blackburn & C. N. Dulmus (Eds.), Handbook of gerontology: Evidence-based
approaches to theory, practice, and policy (pp. 87–116). Hoboken, NJ: John Wiley.
Ebner, N. C., Freund, A. M., & Baltes, P. B. (2006). Developmental changes in personal goal
orientation from young to late adulthood: From striving for gains to maintenance and
prevention of losses. Psychology and Aging, 21, 664–678. doi:10.1037/0882-
7974.21.4.664
Erikson, E. (1963). Childhood and society. New York, NY: Norton.
Francis, Z., & Job, V. (in press). Lay theories of willpower. Social and Personality
Psychology Compass.
Freund, A. M., & Baltes, P. B. (1998). Selection, optimization, and compensation as strategies
of life management: Correlations with subjective indicators of successful aging.
Psychology and Aging, 13, 531–43. doi:10.1037/0882-7974.13.4.531
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 33
Freund, A. M., & Isaacowitz, M. (2014). Beyond age comparisons: A plea for the use of a
modified Brunswikian approach to experimental designs in the study of adult
development and aging. Human Development, 351–371. doi:10.1159/000357177
Freund, A. M., Nikitin, J., & Riediger, M. (2012). Successful aging. In R. M. Lerner, A.
Easterbrooks, & J. Mistry (Eds.), Comprehensive handbook of psychology: Volume 6:
Developmental psychology (pp. 615–638). New York, NY: Wiley.
Gall, T. L., Evans, D. R., & Howard, J. (1997). The retirement adjustment process: Changes
in the well-being of male retirees across time. The Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 52, 110–117.
doi:10.1093/geronb/52B.3.P110
Gauthier, A. H., & Smeeding, T. M. (2003). Time use at older ages: Cross-national
differences. Research on Aging, 25, 247–274. doi:10.1177/0164027503025003003
Heckhausen, J. (1997). Developmental regulation across adulthood: Primary and secondary
control of age-related challenges. Developmental Psychology, 33, 176–87.
doi:10.1037/0012-1649.33.1.176
Heckhausen, J. (2005). Competence and motivation in adulthood and old age: Making the
most of changing capacities and resources. In A. J. Elliot & C. S. Dweck (Eds.),
Handbook of competence and motivation (pp. 240–256). New York, NY: Guilford
Publications.
Heckhausen, J., & Schulz, R. (1995). A life-span theory of control. Psychological Review,
102, 284–304. doi:10.1037/0033-295X.102.2.284
Job, V., Dweck, C. S., & Walton, G. M. (2010). Ego depletion - is it all in your head? Implicit
theories about willpower affect self-regulation. Psychological Science, 21, 1686–1693.
doi:10.1177/0956797610384745
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 34
Job, V. & Walton, G. M. (2017). Lay theories of self-control. In C. Zedelius, B. C. N. Muller,
& J. Schooler (Eds.) The science of lay theories: How beliefs shape our cognition,
behavior, and health (pp. 47-69). Cham, Germany: Springer.
Job, V., Walton, G. M., Bernecker, K., & Dweck, C. S. (2015). Implicit theories about
willpower predict self-regulation and grades in everyday life. Journal of Personality and
Social Psychology, 108, 637-647. doi:10.1037/pspp0000014
Kanfer, R., Beier, M. E., & Ackerman, P. L. (2013). Goals and motivation related to work in
later adulthood: An organizing framework. European Journal of Work and
Organizational Psychology, 22, 253–264. doi:10.1080/1359432X.2012.734298
Kaiser, R. M. (2009). Physiological and clinical considerations of geriatric patient care. In D.
G. Blazer & D. C. Steffens (Eds.), The American Psychiatric Publishing textbook of
geriatric psychiatry (4th ed., pp. 45–61). Arlington, VA: American Psychiatric
Publishing.
Kooij, D. T. A. M., Lange, A. H., Jansen, P. G. W., Kanfer, R., & Dikkers, J. S. E. (2011).
Does age make a difference? Predicting physical activity of South Koreans. Journal of
Organizational Behavior, 32, 197–225. doi:10.1002/job
Kunzmann, U., & Richter, D. (2009). Emotional reactivity across the adult life span: The
cognitive pragmatics make a difference. Psychology and Aging, 24, 879-889.
doi:10.1037/a0017347
Laslett, P. (1989). A fresh map of life: The emergence of the third age. London, UK:
Weidenfeld and Nicholson.
Lindenberger, U., & Baltes, P. B. (1997). Intellectual functioning in old and very old age:
Cross-sectional results from the Berlin Aging Study. Psychology and Aging, 12, 410–32.
doi:10.1037/0882-7974.12.3.410
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 35
Lindenberger, U., von Oertzen, T., Ghisletta, P., & Hertzog, C. (2011). Cross-sectional age
variance extraction: What’s change got to do with it? Psychology and Aging, 26, 34–47.
doi:10.1037/a0020525
Litman, L., Robinson, J., & Abberbock, T. (2017). TurkPrime.com: A versatile
crowdsourcing data acquisition platform for the behavioral sciences. Behavior Research
Methods, 49, 433–442. doi:10.3758/s13428-016-0727-z
Lövdén, M., Ghisletta, P., & Lindenberger, U. (2005). Social participation attenuates decline
in perceptual speed in old and very old age. Psychology and Aging, 20, 423–434.
doi:10.1037/0882-7974.20.3.423
Luong, G., Charles, S. T., & Fingerman, K. L. (2011). Better with age: Social relationships
across adulthood. Journal of Social and Personal Relationships, 28, 9–23.
doi:10.1177/0265407510391362
Martijn, C., Tenbult, P., Merckelbach, H., Dreezens, E., & De Vries, N. K. (2002). Getting a
grip on ourselves: Challenging expectancies about loss of energy after self-control. Social
Cognition, 20, 441–460. doi:10.1521/soco.20.6.441.22978
McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interactions and
moderator effects. Psychological Bulletin, 114, 376-390. doi:10.1037/0033-
2909.114.2.376
Nikitin, J. & Freund, A. M. (2018). Feeling loved and integrated or lonely and rejected in
everyday life: The role of age and social motivation. Developmental Psychology,
54,1186-1198. doi:10.1037/dev0000502
Nimrod, G. (2007). Retirees’ leisure: Activities, benefits, and their contribution to life
satisfaction. Leisure Studies, 26, 65–80. doi:10.1080/02614360500333937
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect
effects in simple mediation models. Behavior Research Methods, Instruments, &
Computers, 36, 717–731. doi:10.3758/BF03206553 !
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 36
Raz, N. (2000). Aging of the brain and its impact on cognitive performance: Integration of
structural and functional findings. In F. I. M. Craik & T. A. Salthouse (Eds.), The
handbook of aging and cognition (2nd ed., pp. 1-90). Mahwah, NJ: Lawrence Erlbaum.
Robbins, S. B., Lee, R. M., & Wan, T. T. H. (1994). Goal continuity as a mediator of early
retirement adjustment: Testing a multidimensional model. Journal of Counseling
Psychology, 41, 18. doi:10.1037/0022-0167.41.1.18
Rockwood, K., & Mitnitski, A. (2007). Frailty in relation to the accumulation of deficits.
Journal of Gerontology: Medical Sciences, 62, 722–727. doi:10.1093/gerona/62.7.722
Rohr, M. K., Wieck, C., & Kunzmann, U. (2017). Age differences in positive feelings and
their expression. Psychology and Aging, 32, 608-620. doi:10.1037/pag0000200
Rothermund, K., & Brandstädter, J. (2003). Coping with deficits and losses in later life: From
compensatory action to accommodation. Psychology and Aging, 18, 896–905.
doi:10.1037/0882-7974.18.4.896
Ryff, C. D. (1991). Possible selves in adulthood and old age: A tale of shifting horizons.
Psychology and Aging, 6, 286–295. doi:10.1037/0882-7974.6.2.286
Ryff, C. D. (1995). Psychological well-being in adult life. Current Directions in
Psychological Science, 4, 99–104. doi:10.1111/1467-8721.ep10772395
Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological well-being revisited.
Journal of Personality and Social Psychology, 69, 719–727. doi:10.1037/0022-
3514.69.4.719
Salmon, S. J., Adriaanse, M. A., Vet, E. De, Fennis, B. M., & Ridder, D. D. De. (2014).
“When the going gets tough, who keeps going?” Depletion sensitivity moderates the
ego-depletion effect. Frontiers in Psychology, 5:647. doi:10.3389/fpsyg.2014.00647
Salthouse, T. A. (2004). What and when of cognitive aging. Current Directions in
Psychological Science, 13, 140–144. doi:10.1111/j.0963-7214.2004.00293.x
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 37
Savani, K., & Job, V. (in press). Reverse ego-depletion: Acts of self-control can improve
subsequent performance in Indian cultural contexts. Journal of Personality and Social
Psychology.
Sieber, V., Flückiger, L., Mata, J., Bernecker, K., & Job, V. (2017). Autonomous goal striving
promotes a nonlimited theory about willpower. Manuscript in preparation.
Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfaction, and longitudinal well-
being: The self-concordance model. Journal of Personality and Social Psychology, 76,
482–497. doi:10.1037/0022-3514.76.3.482!
Sheldon, K. M., & Hilpert, J. C. (2012). The balanced measure of psychological needs
(BMPN) scale: An alternative domain general measure of need satisfaction. Motivation
and Emotion, 36, 439–451. doi:10.1007/s11031-012-9279-4!
Sheldon, K. M., Houser-Marko, L., & Kasser, T. (2006). Does autonomy increase with age?
Comparing the goal motivations of college students and their parents. Journal of
Research in Personality, 40, 168–178. doi:10.1016/j.jrp.2004.10.004
Sheldon, K. M., & Kasser, T. (2001). Getting older, getting better? Personal strivings and
psychological maturity across the life span. Developmental Psychology, 37, 491–501.
doi:10.1037/0012-1649.37.4.491!
Spriduso, W. W., Francis, K., & MacRae, P. G. (2005). Physical dimensions of aging (2nd
ed.). Champaign, IL: Human Kinetics.
Spiro, A., & Brady, C. B. (2011). Integrating health into cognitive aging: Toward a preventive
cognitive neuroscience of aging. The Journals of Gerontology Series B: Psychological
Sciences and Social Sciences, 66B, 17–25. doi:10.1093/geronb/gbr018
Streubel, B., & Kunzmann, U. (2011). Age differences in emotional reactions: Arousal and
age-relevance count. Psychology and Aging, 26, 966-978. doi:10.1037/a0023424
Taylor, L. M. (2005). Introducing cognitive development. Hove, UK: Psychology Press.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 38
Thayer, R. E., & Moore, L. E. (1972). Reported activation and verbal learning as a function of
group size (social facilitation) and anxiety-inducing instructions. The Journal of Social
Psychology, 88, 277–287. doi:10.1080/00224545.1972.9918685 !
Tornstam, L. (1997). Gerotranscendence: The contemplative dimension of aging. Journal of
Aging Studies, 11, 143-54. doi:10.1016/S0890-4065(97)90018-9
Weiss, R. S. (2005). The experience of retirement. Ithaca, New York: Cornell University
Press
Zacher, H., Rosing, K., Henning, T., & Frese, M. (2011). Establishing the next generation at
work: Leader generativity as a moderator of the relationships between leader age, leader-
member exchange, and leadership success. Psychology and Aging, 26, 241–252.
doi:10.1037/a0021429
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 39
Footnotes
1 So far, research has accumulated evidence suggesting that a nonlimited theory has
overall beneficial effects (for overviews see Job & Walton, 2017, or Francis & Job, in press).
However, there might be boundary conditions or detrimental effects of a nonlimited
willpower theory (e.g., in extreme physical or psychological circumstances, like torture or
famine). Further, research on willpower theories does not imply that people may exert
themselves endlessly without needing food or sleep. Therefore, it is important to stress that
this research should not be misused to press people to continue working if they feel the need
for rest and recovery.
2 We decided to collect data from two extreme groups with regard to age, following a
recommendation by McClelland and Judd (1993) that for a given sample size, extreme group
sampling substantially increases the power to detect an effect, compared to continuous or
random sampling.
3 Since gender was not equally distributed across age groups, we ran the same analysis
including it as a control variable. The results remained unchanged.
4 Since income had shown significant age differences, we ran the same analysis
including it as a control variable. The results remained unchanged.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 40
Table 1
Descriptive Statistics and Zero-Order Correlations for the Main Variables of Study 1
(N=802)
Variable
M
(SD)
Range
1
2
1.
Willpower Theory
4.29
(0.99)
1-7
2.
Age
49.28
(16.51)
18-83
.24**
3.
Health
4.78
(1.40)
1-7
.25**
-.21**
Note. Higher scores indicate higher agreement with the nonlimited willpower theory, higher
age and better health. **p < .01.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 41
Figure 1. Scatterplot depicting the relationship between age and a nonlimited
willpower theory in Study 1. R2 linear = 0.056.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 42
Figure 2. The effect of age and autonomy condition on perceived autonomy and
implicit theory about willpower Study 3a. Error bars represent standard error of the mean.
AGE DIFFERENCES IN THEORIES ABOUT WILLPOWER 43
Figure 3. The effect of age and autonomy condition on perceived autonomy and
implicit theory about willpower Study 3b. Error bars represent standard error of the mean.
