APPLIED PSYCHOLOGY: HEALTH AND WELL-BEING, 2011, 3 (2), 172–192

doi:10.1111/j.1758-0854.2011.01050.x

Sources of Perceived Self-Efficacy as Predictors of

Physical Activity in Older Adults aphw_1050 172..192

Lisa M. Warner*
Freie Universität Berlin and German Centre of Gerontology, Berlin,
Germany

Benjamin Schüz
German Centre of Gerontology, Berlin, Germany

Keegan Knittle
Leiden University, The Netherlands

Jochen P. Ziegelmann
Freie Universität Berlin, Germany

Susanne Wurm
German Centre of Gerontology, Berlin, Germany

According to Bandura’s self-efficacy theory, there are four sources of self-

efficacy: past experience, vicarious experience, verbal persuasion, and percep-
tion of physical states. The aims of the study were twofold: To review previous
research on the sources of self-efficacy and to examine the sources in predicting
self-efficacy for exercise in older adults. A sample of 309 older adults was
assessed at two time points for exercise, exercise-specific self-efficacy, and four
sources of self-efficacy. Past experiences, vicarious experiences, and subjective
health had significant direct effects on self-efficacy and indirect effects on exer-
cise via self-efficacy. Persuasive arguments did not predict self-efficacy. This
suggests that future research should target past experience and vicarious expe-
rience as sources of self-efficacy.

Keywords: exercise, multimorbidity, objective health, sources of self-efficacy,

subjective health

* Address for correspondence: Lisa M. Warner, Health Psychology, Freie Universität Berlin,
Habelschwerdter Allee 45, 14195 Berlin, Germany. Email: [email protected]

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology. Published by Blackwell Publishing Ltd., 9600 Garsington
Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 173
INTRODUCTION
Perceived self-efficacy plays a crucial role in the initiation of new behaviors
and in changing old habits. According to Bandura (1997, p. 3), “perceived
self-efficacy refers to beliefs in one’s capabilities to organise and execute the
courses of action required to produce given attainments”. Self-efficacy
beliefs influence the activities which people choose to engage in, the level of
effort they expend in those activities, the extent to which they persevere
in the face of difficulties, and the cognitive evaluations and emotional
reactions brought about by successes and failures (Bandura, 1997). Besides
general self-efficacy beliefs, individuals hold efficacy beliefs of varying
strength across specific life domains. For example, a person might deem
him- or herself fully capable of successfully managing a company, but at
the same time feel incapable of quitting smoking or exercising regularly.
These task- or domain-specific self-efficacy beliefs predict related outcomes
better and are more amenable to change (Bandura, 1997; Leganger, Kraft,
& Roysamb, 2000).
Self-efficacy is one of the factors studied most often in health behavior
research. Domain-specific forms of self-efficacy have been shown to predict
smoking, alcohol consumption, physical activity, dietary behavior, condom
use, dental hygiene behavior, and treatment compliance among chronically ill
individuals (Luszczynska & Schwarzer, 2005). Thus, there is strong evidence
that health behavior-specific self-efficacy predicts health behaviors and health
behavior change. However, much less research exists about the origins and
sources of self-efficacy beliefs—information which could prove vital in the
development of health behavior change interventions.

Sources of Self-Efficacy for Health Behavior

Social Cognitive Theory (Bandura, 1997) proposes four main sources of
self-efficacy for any particular behavior: mastery experience, vicarious expe-
rience, verbal persuasion, and somatic and affective states. In the following,
we will discuss these sources, as well as studies that have examined the impact
of these sources upon self-efficacy.
Mastery experiences are postulated to be “the most effective source
of efficacy information because they provide the most authentic evidence of
whether one can master whatever it takes to succeed” (Bandura, 1997, p. 80).
Mastery is hypothesised to be effective because people judge their capabilities
according to their own direct experiences and observable successes or failures
during goal pursuit. Building on operant conditioning, Bandura (1997)
assumes that previous successes foster self-efficacy beliefs, and thereby
enhance behavior likelihood because they represent desired outcomes;
whereas failures undermine self-efficacy beliefs because they represent

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
174 WARNER ET AL.

undesired outcomes (Bandura, 1997). However, one’s subjective attributions

of what caused success or failure can alter this paradigm: Failure (when
attributed to low effort or insurmountable barriers to achievement) can
actually enhance self-efficacy, and, likewise, success (when attributed to
chance or help from others) can diminish self-efficacy (Leganger et al., 2000).
Self-efficacy for a particular behavior therefore not only predicts future
behavior; it is also influenced by one’s interpretations of past behavior
(Orsega-Smith, Payne, Mowen, Ho, & Godbey, 2007).
Several correlational studies suggest that mastery experiences are a source
of health behavior-specific self-efficacy beliefs, such as self-efficacy for exer-
cise, walking, or disease self-management (e.g. McAuley, Elavsky, Jerome,
Konopack, & Marquez, 2005; O’Brien Cousins & Tan, 2002; Ott, Greening,
Palardy, Holderby, & DeBell, 2000). Qualitative studies report mastery expe-
riences to be an important source of self-efficacy as well, for example for
rehabilitative exercise and muscular endurance (Feltz & Riessinger, 1990;
Resnick, 2002). Additionally, evidence from intervention studies indicates
that mastery experience prompts self-efficacy beliefs (Ashford, Edmunds, &
French, 2010).
Another suggested source of self-efficacy beliefs is vicarious experience, or
modeling (Bandura, 1997). Seeing others perform a behavior and observing
the consequences of their actions is assumed to increase people’s beliefs
in their own capability to master similar tasks, with comparable results
(Bandura, 1997). Vicarious experiences are a powerful source of self-efficacy
beliefs because they provide the observer with strategies and techniques
needed to attain desired goals (Wise & Trunnell, 2001). Few correlational
studies have measured vicarious experiences and investigated their associa-
tion with health behaviors. One such study found that modeling predicted
self-efficacy for condom use (Wulfert & Wan, 1993). Vicarious experiences,
when included as part of interventions, have been shown to effectively
enhance health behavior-specific self-efficacy (Ashford et al., 2010). In a
study that examined the effects of vicarious experience, mastery experience,
and verbal persuasion upon self-efficacy, vicarious experience was found
(behind mastery experience) to be the second strongest predictor of self-
efficacy (Wise & Trunnell, 2001). Vicarious experiences are also frequently
mentioned when participants are asked to report the basis of their efficacy
beliefs in qualitative studies (Feltz & Riessinger, 1990; Resnick, 2002).
Verbal persuasion, which according to Social Cognitive Theory is another
source of self-efficacy, is seen as an expression of faith in the capabilities of an
individual by someone else (Bandura, 1997). Although verbal persuasion is
the most widely used technique in interventions in the health domain, it is
also both theoretically and empirically the weakest source of self-efficacy
(Ashford et al., 2010). According to Bandura, verbal persuasion is limited in
its power to create enduring increases in self-efficacy, maybe because it is

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 175
easier to convince people that they are incapable of doing something and to
hinder them from trying than it is to convince them of the opposite (Bandura,
1997). Verbal persuasion should therefore have the largest effects upon self-
efficacy among people who, based on previous experience, already have a
moderate level of efficacy beliefs. In this vein, Wise and Trunnell (2001) found
that verbal persuasion only increased self-efficacy when combined with per-
formance accomplishments. Studies that investigate the effects of verbal per-
suasion on self-efficacy for health behaviors often find no relationship (Ott
et al., 2000), as people can respond to persuasion with disbelief or reactance
(Miller, Lane, Deatrick, Young, & Potts, 2007). Hence, it is not surprising
that Ashford et al. (2010) found a negative relation between efforts to per-
suade participants of their abilities to be physically active and their perceived
self-efficacy for physical activity. As verbal persuasion and vicarious experi-
ence are in fact social influences, it can become problematic when they are
conceptualised conjointly as social support, thereby making it impossible to
compare the unique contribution of each upon self-efficacy beliefs (e.g.
McAuley, Jerome, Marquez, Elavsky, & Blissmer, 2003b).
Somatic and affective states are deemed to be the fourth source of self-
efficacy perceptions. In general, people tend to read physiological signs, such
as arousal or tension, as signs of being unprepared for a task or of poor
performance (Wood & Bandura, 1989). Hence, people are more likely to feel
competent if they do not experience aversive arousal (Conger & Kanungo,
1988). Bandura postulates that “physiological indicators of efficacy play an
especially influential role in health functioning and in activities requiring
physical strength and stamina” (Bandura, 1997, p. 106). For example,
fatigue, aches, and pains can be attributed to physical incapability and reduce
self-efficacy (Wood & Bandura, 1989). Somatic states are therefore of par-
ticular importance in exercise behavior. Especially among older adults, objec-
tive health problems, pain, and fear of injury constitute important barriers to
engaging in regular exercise (e.g. Lim & Taylor, 2005). However, the lack of
physical activity among older adults cannot be fully explained by deteriorat-
ing objective health or injury (Rhodes et al., 1999). Rather, subjectively
feeling healthier is associated with exercise self-efficacy in older adults (e.g.
O’Brien Cousins & Tan, 2002). In line with that, among chronically ill and
older adults, subjective measurements of health (e.g. ratings of vitality or
pain) have been shown to predict exercise self-efficacy better than more
objective measurements (e.g. comorbid conditions; Perkins, Baum, Taylor, &
Basen-Engquist, 2009).
Although qualitative research indicates that physical sensations are often
detrimental to exercise-specific self-efficacy beliefs (Feltz & Riessinger, 1990;
Resnick, 2002), until now, intervention studies have largely neglected altering
perceptions or misinterpretations of somatic and affective states (Ashford
et al., 2010).

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
176 WARNER ET AL.

Aims of the Current Study

Even though it is of special importance to empirically investigate the sources
from which health behavior self-efficacy originates, as well as which of these
sources has the greatest impact, efficacy beliefs have rarely been investigated
as dependent variables in the health domain. Within other domains of self-
efficacy, such as social, academic, mathematics, educational, or managerial
self-efficacy, however, this question has been investigated (S.L. Anderson &
Betz, 2001; Britner, 2008; Lent, Lopez, & Bieschke, 1991; Schunk, 2003; Van
Vianen, 1999). Within these domains, comparisons of all four sources of
efficacy beliefs revealed mastery experience as the most consistent predictor
of self-efficacy beliefs. However, relative importance of the sources seems to
depend to a large extent upon which kind of behavior is investigated, as for
example for job and coming out self-efficacy (self-efficacy for disclosing one’s
homosexuality), mastery experiences were less important than other sources
(M.K. Anderson & Mavis, 1996; Chiles & Zorn, 1995).
Correlational studies in the domain of physical activity most often only
analyze the association of somatic and affective states with exercise self-
efficacy (e.g. Maly, Costigan, & Olney, 2006; Perkins et al., 2009). A recent
systematic review by Ashford et al. (2010) identified 27 intervention studies
that targeted self-efficacy for physical activity and showed that mastery expe-
riences, along with vicarious experiences, had the strongest effects on self-
efficacy. However, none of the studies included actually measured whether
participants had experienced mastery, nor did they systematically compare
prompts of mastery experience against prompts for other sources of self-
efficacy. Only one intervention study in the exercise domain (Wise & Trun-
nell, 2001) experimentally tested three of the assumed self-efficacy sources
against each other, showing that mastery experiences more effectively
improved self-efficacy than either vicarious experience or verbal persuasion.
However, to our knowledge, no study has yet compared all four theoretical
sources of exercise self-efficacy against each other. Therefore, our study was
undertaken to set a first example of how to measure the four sources of
self-efficacy and compare them to one another. This will be accomplished not
only by investigating the direct effects of the sources on self-efficacy beliefs,
but also by testing for indirect effects of these sources through exercise
self-efficacy, upon exercise behavior in a sample of older adults with multiple
illnesses. Multimorbidity (the co-occurrence of two or more chronic condi-
tions at the same time) is a serious problem for older adults, as those affected
report not only more frequent health care utilisation, treatment complica-
tions, and medication interactions, but also experience decline in quality of
life and autonomy (Fortin et al., 2004; Gijsen et al., 2001). By activating older
adults’ self-efficacy beliefs to perform regular exercise, it is possible to delay
need for care and enable older adults with chronic conditions to maintain

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 177
their autonomy and quality of life for a longer period of time (Bassey, 2002;
Chin A Paw, van Uffelen, Riphagen, & van Mechelen, 2008).
Based on theoretical assumptions of Social Cognitive Theory and on the
above-mentioned evidence from correlational, qualitative, experimental,
and meta-analytic studies on the sources of specific self-efficacy beliefs, we
propose the following hypotheses:

Hypothesis 1: The strongest source of exercise self-efficacy is mastery experi-

ence, followed by vicarious experience and subjective health perceptions.
Verbal persuasion has either no relation or a negative one to exercise self-
efficacy. Subjective perceptions of health have a stronger relation to exercise
self-efficacy than the objective health status.
Hypothesis 2: The size of the indirect effects through exercise self-efficacy on
exercise behavior is highest for mastery experience, followed by vicarious
experiences and subjective health.
Hypothesis 3: The higher the level of exercise self-efficacy, the higher the
frequency of exercise.

METHODS

Participants and Procedure

Participants in the PREFER (Personal Resources of Elderly People with
Multimorbidity: Fortification of Effective Health Behaviour) project were
recruited from the database of the third German Ageing Survey (DEAS;
Wurm, Tomasik, & Tesch-Römer, 2010), a population-representative survey
of community-dwelling adults aged 40 and over. The third wave of the DEAS
took place from July to September 2008. The DEAS served as recruitment
pool and baseline assessment (Time 0 [T0]) for the PREFER study. Partici-
pants of the DEAS were eligible for PREFER if they were (a) 65 years or
older, (b) suffered from at least two chronic physical conditions, mentioned in
either the Charlson Comorbidity Index (Charlson, Szatrowski, Peterson, &
Gold, 1994) or the Functional Comorbidity Index (Groll, To, Bombardier, &
Wright, 2005), and (c) had given consent to be contacted for further studies.
From the initial sample of 6,205 DEAS participants (aged 40–85), 443
(7.14%) fulfilled all inclusion criteria. Of these 443 participants, 309 (69.7%)
gave informed consent to take part in the PREFER project and made an
appointment for the baseline measurement (Time 1 [T1]; March 2009). Par-
ticipants were visited at home by trained interviewers, where they completed
a 30-minute personal interview with objective health tests. In addition, they
were given a questionnaire to be filled in and returned in a prepaid return
envelope. The second measurement point (Time 2 [T2]; September 2009), also
consisting of an interview and questionnaire data, was completed by 271
(87.7% of T1) persons.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
178 WARNER ET AL.

Participants came from all regions of Germany, with 108 (35%) living in
the eastern federal states (former German Democratic Republic). Partici-
pants reported on average 5.49 chronic conditions (SD = 2.86) at T1, with
hypertension (67.64%), osteoarthritis (63.11%), hyperlipidaemia (49.19%),
arthritis (31.07%), and peripheral vascular disease (30.74%) being the five
most prominent conditions. All participants were mentally and physically fit
enough to take part in a 30-minute standardised interview and to fill in the
questionnaire.

Measurements
Mastery experience was operationalised as physical exercise frequency at T1
and T2 by asking participants in the interviews, “On how many days of the
past week have you exercised, e.g. hiking, football, aerobics, swimming?”
Answers could range from 0 to 7 days (Craig et al., 2003). Assessing only the
past week, and assessing days instead of minutes per week, was done to avoid
memory biases that can occur in older adults (Rikli, 2000). Additionally, as
interview data tend to be more reliable than physical activity data from
questionnaires (Washburn, 2000), the exercise frequency item was assessed in
the personal standardised interview.
Two items from the Support for Exercise Habits Scale (Sallis, Grossman,
Pinski, Patterson, & Nader, 1987) were used as measurements of vicarious
experience. Participants rated the frequency that their friends and their family
served as exercise models in the last 3 months in the T1 questionnaire. The
two items read as follows: (a) “In the last 3 months, friends, acquaintances, or
neighbors have exercised with me”, and (b) “In the last 3 months, members of
my family (e.g. partner, children, siblings, or grandchildren) have exercised
with me”. The items were answered on a 5-point scale ranging from 1
([almost] never) to 5 ([almost] always), and were averaged to create the
variable vicarious experience. The Pearson’s correlation between the items
was .29 (p < .001).
Two different items from the Support for Exercise Habits Scale (Sallis
et al., 1987) were used to measure verbal persuasion. Participants rated the
frequency with which their friends and their family encouraged them to
exercise in the last 3 months. The two items read as follows: (a) “In the last 3
months, friends, acquaintances, or neighbors have encouraged me to stick
with my exercise program”, and (b) “In the last 3 months, members of my
family (e.g. partner, children, siblings, grandchildren) have encouraged me to
stick with my exercise program”. The items were answered on a 5-point scale
ranging from 1 ([almost] never) to 5 ([almost] always), and were averaged to
form the variable verbal persuasion. The Pearson’s correlation between the
items was .50 (p < .001).

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 179
Somatic states were assessed with two variables: Objective and subjective
health. Objective health was measured in the T1 interview with a peak expi-
ratory flow meter that assesses the maximum pulmonary expiratory flow. It
has been shown that peak pulmonary expiratory flow is a reliable and sensi-
tive indicator of fitness in older and frail adults (Cook et al., 1995). Partici-
pants were asked to maximally inhale and exhale as fast and forcefully as
possible into the instrument while standing. To be sure that inexperience or
problems with the apparatus did not bias this measurement, participants
completed the task twice. Scores could range from 60 to 800 liters per minute,
and the better result of the two trials was taken as measurement of objective
health (Cook et al., 1991). Subjective health was assessed in the interview at
T1 with the single item: “How would you describe your current state of
health?” Answers could be given on a 5-point scale ranging from 1 (very poor)
to 5 (excellent).
Self-efficacy for exercise was assessed in the T1 questionnaire with three
items based on a study by Scholz, Sniehotta, and Schwarzer (2005). The items
were (a) “I am confident that I can exercise on a regular basis”, (b) “I am
confident that I can exercise on a long-term basis”, and (c) “I am confident
that I can exercise on a regular basis, even if I have to do this on my own”.
Responses could range from 1 (not true at all) to 4 (exactly true), and the item
scores were summed to create the total self-efficacy score (Cronbach’s
alpha = .94).
Control variables were participants’ gender, age, and education level
(1 = low education, at most 9 years’ school education, 2 = medium education,
secondary school, 3 = higher education, qualifying for university admission)
taken from the International Standard Classification of Education (ISCED;
UNESCO, 1997). This specific set of control variables was chosen because
female gender, older age, and lower education have all been linked to lower
exercise frequency (e.g. DiPietro, 2001; Rhodes et al., 1999).

Analytic Procedure
Descriptive data analysis and Pearson correlations were carried out with
SPSS 18.0. Descriptive data analyses revealed significant skewness for exer-
cise frequency at T1 and T2, as most participants indicated that they had
exercised on none of the last seven days (60.5% at T1; 57.0% at T2). There-
fore, exercise frequency was reciprocally transformed (1/x + 1) and then
reversed, so that higher scores again indicated higher exercise frequency. This
procedure yielded a more satisfactory distribution. Means, standard devia-
tions, and ranges of exercise frequency in Table 1 are reported before trans-
formation; correlation and regression coefficients were calculated using the
transformed exercise frequency.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 180

TABLE 1
Descriptive Statistics and Correlations

M SD Range 1 2 3 4 5 6 7 8 9
WARNER ET AL.

1. Age T1 73.23 5.10 65–85

Association of Applied Psychology.

2. Gender 1 = men (58%), -.09
2 = women (42%)
3. Education level (1 = low, 2.23 0.66 1–3 -.07 -.41***
2 = medium, 3 = high)
4. Mastery Experience/Exercise 1.19 2.01 0–7 -.10 -.01 .15**
Frequency T1
5. Vicarious experience T1 2.21 1.06 1–5 -.13* -.17** .16** .23**
6. Verbal persuasion T1 2.31 1.15 1–5 .03 -.14* .10 .08 .45***
7. Subjective health T1 3.36 0.80 1–5 -.07 -.03 .11 .23*** .10 -.05
8. Objective health T1 345.51 134.48 70–780 -.24*** -.50*** .36*** .14* .21*** .13* .16**
9. Exercise self-efficacy T1 2.94 0.90 1–4 -.20** -.07 .16** .40*** .41*** .13* .30*** .23***
10. Exercise frequency T2 1.30 2.00 0–7 -.11 -.04 .23*** .50*** .37*** .24*** .14* .26*** .40***

Note: T1 = Time 1, T2 = Time 2.

* p < .05; ** p < .01; *** p < .001.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 181
To estimate direct and indirect effects of multiple independent variables via
self-efficacy upon exercise frequency at T2, we used the nonparametric boot-
strapping procedure implemented in MPlus 5.21 (Muthén & Muthén, 1998–
2009). The total effect (c) of an independent variable (IV; sources of self-
efficacy) upon the dependent variable (DV; exercise frequency) is the sum of
the direct effect (c′) of the IVs on the DV and the indirect effect (a*b) of the
IV on the DV through the proposed mediator self-efficacy (M); where a is the
effect of the IV upon the M, and b is the effect of the M upon the DV,
partialling out the effect of the IV. The total indirect effect is the sum of a*b
weights. The estimated standardised path coefficients are equivalent to stan-
dardised beta weights in multiple regression analyses. The 95 per cent confi-
dence intervals for the indirect effects are estimated using bootstrapping
procedures with 1,000 resamples. Significance at the a = .05 level is assumed
if the 95 per cent confidence interval of an indirect effect does not include
zero. Bootstrapping has the advantage of not imposing the assumptions of
normality of the sampling distribution (MacKinnon, Fairchild, & Fritz,
2007). This procedure does not, like former tests of mediation, impose the
condition of a significant c path (MacKinnon et al., 2007). Missing data were
imputed using the Full Information Maximum Likelihood method.

RESULTS

Sample Statistics
Descriptive statistics and intercorrelations can be found in Table 1. At T1,
participants were on average 73.23 years old (SD = 5.10), and 41.7 per cent
were women. Around 12.6 per cent indicated a low, 52.1 per cent a medium,
and 35.3 per cent a higher education level according to the International
Standard Classification of Education (ISCED; UNESCO, 1997).
A paired sample t-test with listwise deletion revealed no significant differ-
ence in exercise frequency from T1 to T2, MT1 = 1.19, SD = 2.01, MT2 = 1.30,
SD = 2.00; t(276) = -.31, p > .05. A majority of participants reported no exer-
cise sessions within the last 7 days at both T1 (60.5%) and T2 (57.0%).

Attrition Analyses
Those 35 participants who dropped out between T1 and T2 were examined
for significant differences in the study variables at T1. Participants who
dropped out indicated significantly lower subjective health and exercise fre-
quency at T1 than participants who completed the second measurement (for
subjective health: MDrop-outs = 2.89, SD = .93; MCompleters = 3.42, SD = .76,
p < .001; for exercise frequency: MDrop-outs = 0.57, SD = 1.00; MCompleters = 1.27,
SD = 2.09, p < .01). In longitudinal research on aging (including the present

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
182 WARNER ET AL.

study), this selective attrition is not problematic when examining associations

between variables and not descriptive outcomes (e.g. Kempen & van Son-
deren, 2002).

Prediction of Exercise Self-Efficacy

In order to investigate Hypothesis 1, the sources of self-efficacy (mastery
experience, vicarious experience, verbal persuasion, and subjective and objec-
tive health) were entered into the model predicting exercise self-efficacy, while
statistically controlling for gender, age, and education. As shown in Table 2,
self-efficacy was significantly and positively related to mastery experience,
vicarious experience, and subjective health (a paths). Contrary to Hypothesis
1, mastery experiences, vicarious experiences, and subjective health had
almost equally strong positive relations to self-efficacy (Paternoster, Brame,
Mazerolle, & Piquero, 1998): Comparisons of beta values revealed no signifi-
cant differences between the effects of mastery experiences and vicarious
experiences, z = 0.81, p = .21, or between the effects of mastery experiences
and subjective health, z = 0.87, p = .19. The control variables, verbal persua-
sion and objective health, had no relation with exercise self-efficacy. Overall,
the model explained 33 per cent of the variance in exercise self-efficacy.

Prediction of Exercise Frequency

According to Hypothesis 2, the sources of self-efficacy should exert indirect
effects on exercise frequency at T2 through self-efficacy. Therefore, the direct
relationships between the sources of self-efficacy at T1 and exercise frequency
at T2 were first examined without introducing self-efficacy into the model
(c path), while controlling for gender, age, and education. Gender, education,
mastery experience, vicarious experience, and objective health were shown to
have significant direct effects upon exercise frequency (see Table 2). In the
next step, exercise self-efficacy was introduced into the model. This model
revealed—in line with Hypothesis 3—a significant and positive relationship
between self-efficacy at T1 and exercise frequency at T2 (b path).
There were significant positive indirect effects for mastery experience,
vicarious experience, and subjective health on exercise frequency at T2
through self-efficacy (a*b paths). Vicarious experience had a slightly larger
indirect effect than mastery experience. Verbal persuasion and objective
health had no indirect associations with exercise frequency at T2 through
self-efficacy. No indirect effects were calculated for the control variables, as
they were assumed to have direct relations to exercise only. Overall, this
model explained 39 per cent of the variance in exercise frequency.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 TABLE 2
Direct and Indirect Effects of the Sources of Exercise Self-Efficacy on Exercise Frequency

Direct effects
Total effects IV,DV Indirect effects Lower 95% CI Higher 95% CI
IV , M M , DV IV , DV controlled for M IV, M , DV limit for limit for

Association of Applied Psychology.

Independent variables (path a) (path b) (path c) (path c⬘) (a*b) a*b a*b

Gender .03 .15** .17** .16**

Age T1 -.09 .15** <.01 .02
Education T0 .02 .15** .14** .14**
Mastery Experience/ .26*** .15** .39*** .35*** .04* 0.008 0.069
Exercise Frequency T1
Vicarious experience T1 .31*** .15** .22** .17** .05* 0.012 0.080
Persuasion T1 -.02 .15** .11 .11 <-.01 -0.020 0.013
Subjective Health T1 .19*** .15** <.01 -.02 .03* 0.004 0.053
Objective Health T1 .09 .15** .17** .16** .01 -0.010 0.035

Note: T1 = Time 1, T2 = Time 2; IV = Independent variables, M = Exercise self-efficacy T1, DV = Exercise frequency T2; reported are standardised path coefficients (b) and
limits of bootstrapped confidence intervals (CI) for a*b product terms. All analyses are controlled for gender, age, and education; N = 309.
* p < .05; ** p < .01; *** p < .001, R2 for self-efficacy = .33, R2 for exercise = .39.
PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
183
184 WARNER ET AL.

DISCUSSION
Self-efficacy has been established as one of the most influential constructs in
health psychology, as it affects both the initiation and maintenance of
health behaviors (Luszczynska & Schwarzer, 2005). In this study, we exam-
ined four theoretical sources of self-efficacy: mastery experiences, vicarious
experiences, verbal persuasion, and somatic states (Bandura, 1997). Our
paper aimed at both reviewing the evidence for the four sources of
self-efficacy in health psychology, and, for the first time, to measure and
compare the impact of these sources on exercise self-efficacy and exercise
behavior.

Prediction of Exercise Self-Efficacy

We found that mastery experience, vicarious experience, and subjective
health were positively associated with exercise self-efficacy, but no clear
hierarchy emerged among the sources. This was not in line with Bandura’s
assumption that mastery experiences are the single greatest predictor of
self-efficacy (Bandura, 1997). Apart from measurement issues elaborated in
the limitations section below, there are several possible explanations why
vicarious and mastery experiences might have equally strong effects. Exercise
undergoes a strong normative effect, as peers have an impact on exercise
behavior in all age groups (Fein, Plotnikoff, Wild, & Spence, 2004; King
et al., 2000). In addition, Bandura claims that modeling has stronger effects
in individuals with little experience with a behavior, which might have been
the case in our sample, as aging stereotypes prevail, according to which older
adults—especially older women—are not assumed to exercise, which makes it
less likely that mastery experiences exist (Bandura, 1997; O’Brien Cousins,
2000; Vertinsky, 1998). The relatively strong effect of subjective health may
be explained by the perceived fitness level being frequently reported as a
source of self-efficacy in older adults (Resnick, 2002).
Verbal persuasion and objective health status were not related to exercise
self-efficacy in our study. We expected this null-effect for verbal persuasion,
as in a number of studies it has been shown to be a poor predictor of
self-efficacy (Ashford et al., 2010; Ott et al., 2000; Wise & Trunnell, 2001).
We did not, however, expect the null-effect of objective health upon exercise
self-efficacy, as somatic states should have an impact upon self-efficacy beliefs
within domains that require strength (Bandura, 1997). We found that sub-
jective ratings of health, but not objective health, affect self-efficacy. This
seems to contradict findings showing that older adults perceive physical
barriers as large hindrances to exercise (Lim & Taylor, 2005; O’Brien
Cousins, 2000). However, these studies did not test the objective health status
of their participants, and their finding of poor health as the most important

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 185
barrier to exercise in older adults might in fact reveal that perceived poor
health is the most problematic barrier to building up self-efficacy beliefs (Lim
& Taylor, 2005; O’Brien Cousins, 2000).

Prediction of Exercise Frequency

We were also interested in the indirect association of the sources with exercise
behavior. We found indirect effects of mastery experience, vicarious experi-
ence, and subjective health upon exercise frequency via exercise self-efficacy.
Verbal persuasion and objective health status had no indirect association
with exercise frequency, in part because these two potential sources were not
significantly related to exercise self-efficacy.
Furthermore, we found direct positive effects of female gender, higher
education, and better objective health status upon exercise frequency. The
finding for education is in line with previous studies that have demonstrated
the effect of educational level on exercise (DiPietro, 2001), but the finding for
gender was unexpected, as older men typically report more frequent exercise
than older women (Lim & Taylor, 2005; Rhodes et al., 1999). However, the
relationship to gender in our study is difficult to interpret, as the bivariate
correlation between gender and exercise frequency was not significant. This
points to a suppressor effect, which we will not interpret further, in light of
the clear, pre-existing link between female gender and less exercise among
older adults (Lim & Taylor, 2005; Rhodes et al., 1999; Tu, Gunnell, &
Gilthorpe, 2008).
Although we found a direct relation between objective health status and
exercise, the lack of a relation between objective health and self-efficacy
beliefs is also challenging to interpret, as we expected that people judge their
ability to exercise, at least to some extent, on their current health status
(O’Brien Cousins, 2000). However, our measurement of objective health—a
lung capacity test—might have been too distal to participants’ considerations
of their exercise abilities, as individuals might consider their leg strength,
functional limitations, or pain as more influential than their level of lung
fitness.

Limitations of the Current Study and Suggestions for

Future Research
Our study is a post-hoc analysis of an existing dataset. Therefore, we have to
acknowledge limitations in the assessment of the sources of self-efficacy. At
the same time, future research on the sources of self-efficacy can benefit from
our findings:
We assessed mastery experiences as past performance in our study. Even
though past performance has been shown to be a useful proxy for mastery

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
186 WARNER ET AL.

experience in previous research (Duncan, Rodgers, Hall, & Wilson, 2011;

McAuley et al., 2003b), a better measurement of the concept would incorpo-
rate a subjective evaluation of successes and failures, as well as participants’
attributions of these successes and failures to internal or external factors
(Leganger et al., 2000).
Our measurement for vicarious experiences was derived from the Support
for Exercise Habits Scale (Sallis et al., 1987). As it was assessed whether
participants exercised with a companion, this operationalisation might have
incorporated not only vicarious experience, but past performance as well.
Although it is difficult to distinguish pure modeling from participatory mod-
eling, as older adults prefer to exercise with a partner (Clark, 1996; Wallace,
Raglin, & Jastremski, 1995), future research should make an effort to assess
vicarious experiences in a way that ensures discriminant validity. Even
though vicarious experiences are often assessed by asking participants
whether they had a behavioral model in their environment (S.L. Anderson &
Betz, 2001; Britner, 2008; Lent et al., 1991), future research might also profit
from assessing whether the observer knew of failures and successes that were
experienced by the model due to exercising, as successful models are more
likely to be imitated (Van Vianen, 1999).
Verbal persuasion from family and friends was also assessed using the
Support for Exercise Habits Scale (Sallis et al., 1987). The items utilised
represent a major component of verbal persuasion—namely encouragement
(Lent et al., 1991; Van Vianen, 1999). The measurement of verbal persuasion
might, however, be extended by discouragement items that are often assessed
as well in scales to measure verbal persuasion in other domains of research
(Lent et al., 1991). The assessment of verbal persuasion could also be further
improved by considering persuasive arguments from health professionals
such as nurses or general practitioners, as they constitute an important source
of health information and have been shown to predict exercise motivation
(e.g. Glasgow, Eakin, Fisher, Bacak, & Brownson, 2001).
The assumed fourth source of self-efficacy—somatic and affective states—
was only partly represented in our study. We considered subjective and
objective health, but did not have satisfactory measurement for affective
states. Previous research, for example, found fatigue and depression to be
associated with self-efficacy in breast cancer patients (Perkins et al., 2009),
which suggests a role of affective states in predicting self-efficacy. McAuley
and colleagues (McAuley, Jerome, Elavsky, Marquez, & Ramsey, 2003a) set
an example of how to measure affective states relevant for exercise self-
efficacy by using the Feeling Scale (Rejeski, Best, Griffith, & Kenney, 1987)
together with the question “How did exercise make you feel today?”
(McAuley et al., 2003a). Future research should investigate ways to assess
affective states both immediately before and after exercise, as the immediate
impact of these variables upon self-efficacy has not yet been examined.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 187
Apart from measurement issues, there are some further limitations of the
current study. Our sample of older multimorbid participants might be self-
selected for two reasons: First, participants who completed all assessments of
the study reported higher ratings of subjective health and more frequent
exercise. Second, as participants of our study are all community-dwelling
volunteers who participated in an unpaid survey, our study sample could be
biased towards healthier and more autonomous individuals with multiple
illnesses, as older adults in poor health are more likely to be hospitalised or
living in care facilities. However, our study was drawn from the population-
representative German Ageing Survey (Wurm et al., 2010), which suggests
that it is representative with regard to the inclusion criteria.
Furthermore, as causality cannot be established by means of our research
design, alternative interpretations are possible. For example, subjective and
objective health are typically investigated as outcomes of physical exercise,
rather than as predictors of exercise self-efficacy, and, as such, facilitators of
exercise behavior. Hence, we only investigated one direction of a circular
process, in which improved health evaluations lead to more self-efficacy,
which in turn leads to more frequent exercise, improved subjective health,
and more mastery experience. To investigate these interrelations, studies with
longer-term follow-ups with more frequent measurement points and inter-
vention studies are necessary. Although many intervention studies in the
health domain have prompted self-efficacy beliefs, systematic comparisons of
the sources of self-efficacy are sparse. We found only one experimental study
that systematically tested three of the sources against one another in the
health domain (Wise & Trunnell, 2001). More studies utilising such factorial
designs are warranted, to allow empirical evidence, and not just theoretical
assumptions, to guide the construction of future interventions.
Most research to this point has considered the sources of self-efficacy as
behavior change techniques rather than as psychological constructs to be
measured and tested against one another. Although it may not be prudent at
this time to develop a scale that measures all four sources of exercise self-
efficacy, we have demonstrated that measuring and analyzing the sources of
self-efficacy themselves can provide valuable information for interventionists
seeking to stimulate health behavior change in older adults with multiple
medical conditions. Such a scale could also serve as a manipulation check in
interventions and could provide evidence of how interventions work (Michie
& Abraham, 2004).

Practical Implications and Conclusions

Our study provides not only implications for future research, but also
suggestions for interventions. As older individuals, especially older women,
were found to lack experience with many exercise behaviors, providing safe

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
188 WARNER ET AL.

opportunities for experiencing mastery is an important way to build up

exercise self-efficacy (Lehr, 1992). This could also mean confronting preva-
lent stereotypes that indicate that exercise is inappropriate or dangerous for
older adults (Vertinsky, 1998). Caution should be taken with graded mastery,
however, as a meta-analysis found it to be detrimental to exercise self-efficacy
(Ashford et al., 2010).
According to our results and those from previous research, providing older
adults with vicarious experiences should enhance their self-efficacy for exer-
cise (e.g. O’Brien Cousins, 1996). However, considering that only 27 per cent
of Germans aged 60 or older report regular exercise (2 or more hours a week),
competent exercise role models are sparse in this population (Robert Koch-
Institut, 2005). A solution might be exercise groups that consist of peers or
peer educators to enhance role model exhibition (Beauchamp, Carron,
McCutcheon, & Harper, 2007).
In accordance with other studies, verbal persuasion did not have a positive
impact on self-efficacy for exercise in our study and can therefore not be
recommended as an intervention technique to enhance self-efficacy (e.g.
Ashford et al., 2010; Ott et al., 2000).
With regard to affective and somatic states, older adults might misinterpret
physical feelings, symptoms, or subjective health perceptions prior to and
during exercise as threatening, as signs of vulnerability to physical dysfunc-
tion, or as a sign of their incapacity to perform a new task (Lee, Arthur, &
Avis, 2008). It is thus important to prevent such misinterpretations by edu-
cating older adults about initial physical symptoms that can accompany
physical exercise (e.g. breathlessness, muscle aches, and feelings of unease)
and to provide them with more accurate interpretations, such as insufficient
aerobic fitness or muscle strength that will improve as exercise becomes more
routine (Lee et al., 2008).

ACKNOWLEDGEMENTS
The German Ageing Survey was funded under Grant 301-1720-2/2 by the
German Federal Ministry for Family, Senior Citizens, Women, and Youth.
The first and second authors are funded by the German Federal Ministry of
Education and Research (Grant No. 01ET0702); the fourth author is funded
by Grant No. 01ET0801 by the same funding body. The first and third
authors received a tandem grant from the European Health Psychology
Society for their cooperation. The content is the sole responsibility of the
authors.

REFERENCES
Anderson, M.K., & Mavis, B.E. (1996). Sources of coming out self-efficacy for lesbi-
ans. Journal of Homosexuality, 32, 37–52.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 189
Anderson, S.L., & Betz, N.E. (2001). Sources of social self-efficacy expectations: Their
measurement and relation to career development. Journal of Vocational Behavior,
58, 98–117.
Ashford, S., Edmunds, J., & French, D.P. (2010). What is the best way to change
self-efficacy to promote lifestyle and recreational physical activity? A systematic
review with meta-analysis. British Journal of Health Psychology, 15, 265–288.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
Bassey, E.J. (2002). Exercise for the elderly: An update. Age and Ageing, 31, 3–5.
Beauchamp, M.R., Carron, A.V., McCutcheon, S., & Harper, O. (2007). Older adults’
preferences for exercising alone versus in groups: Considering contextual congru-
ence. Annals of Behavioral Medicine, 33, 200–206.
Britner, S.L. (2008). Motivation in high school science students: A comparison of
gender differences in life, physical, and earth science classes. Journal of Research in
Science Teaching, 45, 955–970.
Charlson, M.E., Szatrowski, T.P., Peterson, J., & Gold, J. (1994). Validation of a
combined comorbidity index. Journal of Clinical Epidemiology, 47, 1245–1251.
Chiles, A.M., & Zorn, T.E. (1995). Empowerment in organizations: Employees’ per-
ceptions of the influences on empowerment. Journal of Applied Communication
Research, 23, 1–25.
Chin A Paw, M.J.M., van Uffelen, J.G.Z., Riphagen, I., & van Mechelen, W. (2008).
The functional effects of physical exercise training in frail older people: A system-
atic review. Sports Medicine, 38, 781–793.
Clark, D.O. (1996). Age, socioeconomic status, and exercise self-efficacy. The Geron-
tologist, 36, 157–164.
Conger, J.A., & Kanungo, R.N. (1988). The empowerment process: Integrating
theory and practice. Academy of Management Review, 13, 471–482.
Cook, N.R., Albert, M.S., Berkman, L.F., Blazer, D., Taylor, J.O., & Hennekens,
C.H. (1995). Interrelationships of peak expiratory flow rate with physical and
cognitive function in the elderly: MacArthur Foundation Studies of Aging. Jour-
nals of Gerontology Series A: Biological Sciences and Medical Sciences, 50A,
M317–M323.
Cook, N.R., Evans, D.A., Scherr, P.A., Speizer, F.E., Taylor, J.O., & Hennekens,
C.H. (1991). Peak expiratory flow rate and 5-year mortality in an elderly popula-
tion. American Journal of Epidemiology, 133, 784–794.
Craig, C.L., Marshall, A.L., Sjöström, M., Bauman, A.E., Booth, M.L., Ainsworth,
B.E. et al. (2003). International physical activity questionnaire: 12-country reliabil-
ity and validity. Medicine and Science in Sports and Exercise, 35, 1381–1395.
DiPietro, L. (2001). Physical activity in aging: Changes in patterns and their relation-
ship to health and function. Journals of Gerontology Series A: Biological Sciences
and Medical Sciences, 56A, 13–22.
Duncan, L.R., Rodgers, W.M., Hall, C.R., & Wilson, P.M. (2011). Using imagery to
enhance three types of exercise self-efficacy among sedentary women. Applied
Psychology: Health and Well-Being, 3, 107–126.
Fein, A.J., Plotnikoff, R.C., Wild, T.C., & Spence, J.C. (2004). Perceived environment
and physical activity in youth. International Journal of Behavioral Medicine, 11,
135–142.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
190 WARNER ET AL.

Feltz, D.L., & Riessinger, C.A. (1990). Effects of in vivo emotive imagery and per-
formance feedback on self-efficacy and muscular endurance. Journal of Sport &
Exercise Psychology, 12, 132–143.
Fortin, M., Lapointe, L., Hudon, C., Vanasse, A., Ntetu, A.L., & Maltais, D. (2004).
Multimorbidity and quality of life in primary care: A systematic review. Health and
Quality of Life Outcomes, 2, doi: 10.1186/1477-7525-1182-1151.
Gijsen, R., Hoeymans, N., Schellevis, F.G., Ruwaard, D., Satariano, W.A., & van
den Bos, G.A. (2001). Causes and consequences of comorbidity: A review. Journal
of Clinical Epidemiology, 54, 661–674.
Glasgow, R.E., Eakin, E.G., Fisher, E.B., Bacak, S.J., & Brownson, R.C. (2001).
Physician advice and support for physical activity: Results from a national survey.
American Journal of Preventive Medicine, 21, 189–196.
Groll, D.L., To, T., Bombardier, C., & Wright, J.G. (2005). The development of a
comorbidity index with physical function as the outcome. Journal of Clinical
Epidemiology, 58, 595–602.
Kempen, G.I.J.M., & van Sonderen, E. (2002). Psychological attributes and changes
in disability among low-functioning older persons: Does attrition affect the out-
comes? Journal of Clinical Epidemiology, 55, 224–229.
King, A.C., Castro, C., Wilcox, S., Eyler, A.A., Sallis, J.F., & Brownson, R.C. (2000).
Personal and environmental factors associated with physical inactivity among
different racial ethnic groups of US middle-aged and older-aged women. Health
Psychology, 19, 354–364.
Lee, L.-L., Arthur, A., & Avis, M. (2008). Using self-efficacy theory to develop
interventions that help older people overcome psychological barriers to physical
activity: A discussion paper. International Journal of Nursing Studies, 45, 1690–
1699.
Leganger, A., Kraft, P., & Roysamb, E. (2000). Perceived self-efficacy in health
behaviour research: Conceptualisation, measurement and correlates. Psychology &
Health, 15, 51–69.
Lehr, U.M. (1992). Physical activities in old age: Motivation and barriers. In S.
Harris, R. Harris, & W.S. Harris (Eds.), Physical activity, aging and sports (Vol. 2,
pp. 52–62). Albany, NY: Center for the Study of Aging.
Lent, R.W., Lopez, F.G., & Bieschke, K.J. (1991). Mathematics self-efficacy: Sources
and relation to science-based career choice. Journal of Counseling Psychology, 38,
424–430.
Lim, K., & Taylor, L. (2005). Factors associated with physical activity among older
people: A population-based study. Preventive Medicine, 40, 33–40.
Luszczynska, A., & Schwarzer, R. (2005). Social cognitive theory. In M. Conner & P.
Norman (Eds.), Predicting health behaviour (2nd edn. rev., pp. 127–169). Bucking-
ham: Open University Press.
McAuley, E., Elavsky, S., Jerome, G.J., Konopack, J.F., & Marquez, D.X. (2005).
Physical activity-related well-being in older adults: Social cognitive influences.
Psychology and Aging, 20, 295–302.
McAuley, E., Jerome, G.J., Elavsky, S., Marquez, D.X., & Ramsey, S.N. (2003a).
Predicting long-term maintenance of physical activity in older adults. Preventive
Medicine, 37, 110–118.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
 PERCEIVED SELF-EFFICACY PREDICTS PHYSICAL ACTIVITY 191
McAuley, E., Jerome, G.J., Marquez, D.X., Elavsky, S., & Blissmer, B. (2003b).
Exercise self-efficacy in older adults: Social, affective, and behavioral influences.
Annals of Behavioral Medicine, 25, 1–7.
MacKinnon, D.P., Fairchild, A.J., & Fritz, M.S. (2007). Mediation analysis. Annual
Review of Psychology, 58, 593–614.
Maly, M.R., Costigan, P.A., & Olney, S.J. (2006). Determinants of self efficacy for
physical tasks in people with knee osteoarthritis. Arthritis Care & Research, 55,
94–101.
Michie, S., & Abraham, C. (2004). Interventions to change health behaviours:
Evidence-based or evidence-inspired? Psychology & Health, 19, 29–49.
Miller, C.H., Lane, L.T., Deatrick, L.M., Young, A.M., & Potts, K.A. (2007). Psy-
chological reactance and promotional health messages: The effects of controlling
language, lexical concreteness, and the restoration of freedom. Human Communi-
cation Research, 33, 219–240.
Muthén, L.K., & Muthén, B.O. (1998–2009). MPlus user’s guide (5th edn.). Los
Angeles, CA: Muthén & Muthén.
O’Brien Cousins, S. (1996). Exercise cognition among elderly women. Journal of
Applied Sport Psychology, 8, 131–145.
O’Brien Cousins, S. (2000). “My heart couldn’t take it”: Older women’s beliefs about
exercise benefits and risks. Journals of Gerontology. Series B, Psychological Sci-
ences and Social Sciences, 55, 283–294.
O’Brien Cousins, S., & Tan, M. (2002). Sources of efficacy for walking and climbing
stairs among older adults. Physical & Occupational Therapy in Geriatrics, 20,
51–68.
Orsega-Smith, E.M., Payne, L.L., Mowen, A.J., Ho, C.-H., & Godbey, G.C. (2007).
The role of social support and self-efficacy in shaping the leisure time physical
activity of older adults. Journal of Leisure Research, 39, 705–727.
Ott, J., Greening, L., Palardy, N., Holderby, A., & DeBell, W.K. (2000). Self-efficacy
as a mediator variable for adolescents’ adherence to treatment for insulin-
dependent diabetes mellitus. Children’s Health Care, 29, 47–63.
Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using the correct
statistical test for equality of regression coefficients. Criminology, 36, 859–866.
Perkins, H.Y., Baum, G.P., Taylor, C.L.C., & Basen-Engquist, K.M. (2009).
Effects of treatment factors, comorbidities and health-related quality of life on
self-efficacy for physical activity in cancer survivors. Psycho-Oncology, 18, 405–
411.
Rejeski, W.J., Best, D., Griffith, P., & Kenney, E. (1987). Sex-role orientation and the
responses of men to exercise stress. Research Quarterly for Exercise and Sport, 58,
260–264.
Resnick, B. (2002). Geriatric rehabilitation: The influence of efficacy beliefs and
motivation. Rehabilitation Nursing, 27, 152–159.
Rhodes, R.E., Martin, A.D., Taunton, J.E., Rhodes, E.C., Donnelly, M., & Elliot, J.
(1999). Factors associated with exercise adherence among older adults. An indi-
vidual perspective. Sports Medicine, 28, 397–411.
Rikli, R.E. (2000). Reliability, validity, and methodological issues in assessing physi-
cal activity in older adults. Research Quarterly for Exercise and Sport, 71, 89–96.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
192 WARNER ET AL.

Robert Koch-Institut (2005). Ergänzende Wertetabellen zu den Abbildungen des

Themenhefts 26—Körperliche Aktivität. Gesundheitsberichterstattung des Bundes
[Additional data for graphics from booklet 26: Physical activity. German Health
Report]. Berlin: RKI.
Sallis, J.F., Grossman, R.M., Pinski, R.B., Patterson, T.L., & Nader, P.R. (1987). The
development of scales to measure social support for diet and exercise behaviors.
Preventive Medicine, 16, 825–836.
Scholz, U., Sniehotta, F.F., & Schwarzer, R. (2005). Predicting physical exercise in
cardiac rehabilitation: The role of phase-specific self-efficacy beliefs. Journal of
Sport & Exercise Psychology, 27, 135–151.
Schunk, D.H. (2003). Self-efficacy for reading and writing: Influence of modeling,
goal setting, and self-evaluation. Reading & Writing Quarterly: Overcoming Learn-
ing Difficulties, 19, 159–172.
Tu, Y.-K., Gunnell, D., & Gilthorpe, M. (2008). Simpson’s Paradox, Lord’s Paradox,
and suppression effects are the same phenomenon: The reversal paradox. Emerging
Themes in Epidemiology, 5, 2. doi:10.1186/1742-7622-5-2.
UNESCO (1997). ISCED 1997: International Standard Classification of Education.
Retrieved 16 May 2010, from http://www.uis.unesco.org/publications/ISCED97.
Van Vianen, A.E.M. (1999). Managerial self-efficacy, outcome expectancies, and
work-role salience as determinants of ambition for a managerial position. Journal
of Applied Social Psychology, 29, 639–665.
Vertinsky, P. (1998). “Run, Jane, run”: Central tensions in the current debate about
enhancing women’s health through exercise. Women & Health, 27, 81–111.
Wallace, J.P., Raglin, J.S., & Jastremski, C.A. (1995). Twelve month adherence of
adults who joined a fitness program with a spouse vs. without a spouse. Journal of
Sports Medicine and Physical Fitness, 35, 206–213.
Washburn, R.A. (2000). Assessment of physical activity in older adults. Research
Quarterly for Exercise and Sport, 71, S79–S88.
Wise, J.B., & Trunnell, E.P. (2001). The influence of sources of self-efficacy upon
efficacy strength. Journal of Sport & Exercise Psychology, 23, 268–280.
Wood, R., & Bandura, A. (1989). Social cognitive theory of organizational manage-
ment. Academy of Management Review, 14, 361–384.
Wulfert, E., & Wan, C.K. (1993). Condom use: A self-efficacy model. Health Psy-
chology, 12, 346–353.
Wurm, S., Tomasik, M.J., & Tesch-Römer, C. (2010). On the importance of a positive
view on ageing for physical exercise among middle-aged and older adults: Cross-
sectional and longitudinal findings. Psychology & Health, 25, 25–42.

© 2011 The Authors. Applied Psychology: Health and Well-Being © 2011 The International
Association of Applied Psychology.
Copyright of Applied Psychology: Health & Well-Being is the property of Wiley-Blackwell and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.