Health Education & Behavior

Health Education & Behavior
2014, Vol 41(1) 7–11
© 2013 Society for Public
Health Education
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1090198112473109
heb.sagepub.com
Brief Report
The interest in workplace health promotion activities targeting
physical activity and nutrition is high and is motivated by
public health and productivity concerns (Black, 2008;
Pencak, 1991; Wolfe, Ulrich, & Parker, 1987). However,
many health promotion initiatives struggle with low participation and have problems achieving optimal effect. Low participation may be because of a multitude of factors related to
individual, institutional, community, and governmental actions
or policies (Linnan, Sorensen, Colditz, Klar, & Emmons,
2001). Yet, since many worksite health promotion programs
are, for ethical reasons, made voluntary, participation is, in the
end, contingent on the individual employee’s motivation and
willingness to participate in a given program.
Human behavior can be described as a compromise
between processes residing within the person and external
and situational circumstances (Bandura, 1978; Blumberg &
Pringle, 1982; Buss, 1991). Accordingly, decisions to
change health behaviors may be viewed as the result of an
interaction between behavioral, cognitive, and environmental influences (Bandura, 1978). Notwithstanding the
plentiful range of external incentives that may motivate
individuals into action (Cialdini, 2001), a person’s selfefficacy (i.e., the perceived belief about his or her ability to
achieve a specific behavior in a particular situation) is often
considered important for understanding changes in health
behaviors (Bandura, 2004). Indeed, in social cognitive theory, self-efficacy is viewed as a core psychological belief
that affects basic processes of change, whether this is concerns, contemplations, mobilization of efforts, or dealing
with relapses (Bandura, 2004).
In a recent review, it was concluded that self-efficacy was
important for adherence to physical activity and exercise
therapy (Rhodes & Fiala, 2009). Although it makes sense
that a certain amount of self-efficacy is needed to engage, or
to change behavior, it also seems reasonable to assume that
exaggerated belief in one’s own efficacy could lead to the
rejection or evasion of offers of help or assistance. This
aspect of self-efficacy seems, however, to have been poorly
elucidated in research.
For this reason, as part of a process evaluation of a multipurpose in-house health promotion service in the Danish
police (Persson et al., 2013), we decided to address this issue
in relation to the use of an in-house wellness service. The
wellness service consisted of six full-time consultants (five
women) who served all employees in the Danish National
Police and the 12 police districts. Hence, the consultants
travelled a lot with the goal to visit the main stations in the
various districts at least once every 4 to 6 weeks. The work
of the wellness service focused on improving physical wellbeing related to the four major lifestyle factors (i.e., exercise,
eating, drinking, and smoking) and was available, free of
charge, to all employees. The service could be used during
473109 HEBXXX10.1177/109019811247310
9Health Education & BehaviorPersson et al.
2013
1National Research Centre for the Working Environment, Copenhagen,
Denmark
2Steno Health Promotion Center, Steno Diabetes Center, Gentofte,
Denmark
Corresponding Author:
Roger Persson, Lersø Parkallé 105, 2100 Copenhagen, Denmark.
Email: rpe@nrcwe.dk
The Relationship Between
Self-Efficacy and Help Evasion
Roger Persson, PhD1, Bryan Cleal, PhD2, Mette Øllgaard Jakobsen, MSc1,
Ebbe Villadsen1, and Lars L. Andersen, PhD1
Abstract
Objective. To examine the relationship between self-efficacy and not wanting help to change health behaviors. Method.
All employees in the Danish police department were invited to respond to an electronic questionnaire. All respondents
expressing a desire to change health behaviors in relation to smoking (n = 845), alcohol (n = 684), eating (n = 4431), and
physical activity (n = 5179) and who subsequently responded to questions on self-efficacy were included. Results. Both
the bivariate and multiple regression analyses showed that all four specific self-efficacy scores were positively related to
reporting that one did not want help. Conclusion. A high belief in one’s own ability to change lifestyle behaviors in relation
to smoking, alcohol, eating, and physical activity may lead to avoidance of help offers in a workplace setting.
Keywords
barriers, health promotion, motivation, police, self-effcacy, work
8 Health Education & Behavior 41(1)
working hours so long as the appointments did not conflict
with daily work tasks. At times the consultation served to
increase the employee’s awareness; on other occasions the
consultation served to support more radical lifestyle changes.
Furthermore, the counseling contained elements of primary,
secondary, and tertiary prevention. Although most activities
were directed toward individual employees, other activities,
such as group activities, health campaigns, and education,
were also occasionally endorsed. In view of the wellness service’s main focus, we decided to explore how well, statistically speaking, self-efficacy could predict the preference for
not receiving help, among participants who had acknowledged that they wanted to change health behaviors in relation
to smoking, alcohol, eating, and physical activity.
Materials and Method
Participants
The research was conducted in accordance with Danish law
and institutional guidelines on ethics in research. All potential users of the wellness service, that is, all administrative
workers (e.g., lawyers, clerks, and mechanics) and police
officers, were invited to respond to an electronic questionnaire (n = 15,284). To guarantee anonymity and alleviate
possible concerns about the access that employers would
have to the results, participants were required to log on to an
external server in order to complete the questionnaire. There
were in total 6,373 respondents (41.8%). Only those
employed in the Danish National Police and in the 12 police
districts and had responded to at least two third of the items
in the questionnaire were included (N = 6,062; mean age =
44 years, SD = 11 years). In the present study, we focus on
participants who had responded that they perceived a need
for lifestyle changes in relation to the four major life style
factors. Thus, of the 6,062 eligible participants, 845 wished
to cut down or quit smoking (14 %), 684 wished to reduce
their alcohol consumption (11 %), 4,431 wished to adopt
healthier eating habits (73 %), and 5,179 wanted to be more
physically active (85%). Demographic characteristics and
lifestyle factors for each of the subgroupings are presented
in Table 1.
Measures
Predictor variables. General self-efficacy was assessed with
three items that read, “I am confident that I can deal efficiently with unexpected events,” “I can solve most problems
if I invest the necessary effort,” and “I can usually handle
whatever comes my way.” All items had five response categories: always, often, sometimes, seldom, and never/hardly
ever. The mean score (range 1-5) was used as a continuous
predictor. Higher scores indicated greater self-efficacy.
Cronbach’s alpha varied between .71 and .72 when calculated in each of the four study samples.
Specific self-efficacy was assessed with one item for each
lifestyle factor: “If you decide to (smoke less/consume less
alcohol/eat healthier/be more physically active), do you
believe, that you can do it?” All specific self-efficacy items
were responded to on a scale from 1 to 10, where the respondents were asked to evaluate how easy it would be: 1 = do
not believe it is possible at all and 10 = it should be very easy
to do that.
Table 1. Demographic and Lifestyle Characteristics for Each of
the Four Subgroupings.
Smokers
Age in years, mean (SD) 45 (11)
Gender, % women 33
Proportion wanting to change
behavior, % of the study sample
14
Number of cigarettes/day, median
(25th-75th percentile)
12 (7-35)
Number of cigars/day, median
(25th-75th percentile)
0 (0-0)
Number of pipe chambers/day,
median (25th-75th percentile)
1 (0-4)
Alcohol consumers
Age in years, mean (SD) 47 (10)
Gender, % women 20
Proportion wanting to change
behavior, % of the study sample
11
Number of consumed units/week,
median (25th-75th percentile)
5 (3-7)
Proportion consuming alcohol up
to 3-4 times/week, %
76
Proportion consuming alcohol up
to 5-7 times/week, %
24
Healthier diet
Age in years, mean (SD) 43 (11)
Gender, % women 29
Proportion wanting to change
behavior, % of the study sample
73
Proportion consuming at least 1
serving of fruit/day, %
43
Proportion consuming at least 1
serving of vegetables/day, %
34
Proportion consuming fast food
at least 1 time/week, %
16
Physical activitya
Age in years, mean (SD) 44 (11)
Gender, % women 29
Proportion wanting to change
behavior, % of the study sample
85
Proportion physically active at
least 30 minutes 7 days/week, %
7
Proportion physically active at
least 30 minutes 0-3 day/week, %
57
aDefined as any activity that increases the respiration rate (e.g., heavy
garden work, walking at a fast pace, competitive sports, etc.), at least 30
minutes per day (%).
Persson et al. 9
Outcome. All four outcome variables were binary, and
reflected whether the participants had marked “I do not want
help” on a list that provided several options for help in relation to
smoking, alcohol consumption, eating, and physical activity.
Statistical Analyses
The statistical computations were made with IBM SPSS
Version 20 for Windows. P values less than .05 were considered statistically significant. Binary logistic regression
analyses were used to explore bivariate relationships.
Multiple binary logistic regression analyses were used to
estimate age-adjusted (continuous) and gender-adjusted
(categorical) relationships.
Results
The mean specific self-efficacy scores were as follows: for
smoking 5.9 (SD = 2.4; range = 0-10), alcohol 7.6 (SD = 2.0;
range = 1-10), eating 6.9 (SD = 1.8; range = 1-10), and
physical activity 7.0 (SD = 1.8; range 0-10). The mean general
self-efficacy scores in the four subsamples were as follows:
smoking 4.2 (SD = 0.5; range = 1.3-5.0), alcohol 4.1 (SD =
0.5; range = 1.7-5.0), eating 4.1 (SD = 0.5; range = 1.3-5.0),
and physical activity 4.1 (SD = 0.5; range = 1.3-5.0).
Bivariate Logistic Regression Analyses
The results for the unadjusted bivariate analyses are presented in Table 2. All four specific self-efficacy scores
were positively related to reporting that one did not want
help. The general self-efficacy score was related, with statistical significance, to not wanting help in relation to
changing physical activity patterns. Increasing age was
positively related to not wanting help as regards changing
eating and physical activity. Women were less likely to
report that they did not want help in relation to eating and
physical activity.
Multiple Logistic Regression Analyses
The results from the age- and gender-adjusted multiple
logistic regression analyses are presented in Table 3. The
four specific self-efficacy scores were positively related to
reporting that one did not want help with changing health
habits. The general self-efficacy score was positively related
to reporting that one did not want help with changing health
habits in relation to physical activity.
Discussion
This study explored how well self-efficacy could statistically predict the preference of not receiving any help among
of participants who had acknowledged that they wanted to
change health behaviors in relation to smoking, alcohol, eating,
and physical activity.
The results from both the bivariate and multiple regression
analyses showed that all four specific self-efficacy scores
were positively related to reporting that one did not want help.
The general self-efficacy score was less clearly associated
with the reports of not wanting help. General self-efficacy was
only statistically predictive for reporting that one did not need
help in relation to physical activity. Irrespective of statistical
significance, the results suggest that higher self-efficacy seems
to foster feelings of self-reliance. Whether self-reliance in the
present context represents a wise and accurate decision or a
potentially detrimental one is another question, which can not
be reliably answered with the present data. Yet the results
show that specific self-efficacy seems to be related to a selfreliant attitude that may affect attitudes toward help initiatives.
Hence, our observations underscore the possibility that high
self-efficacy might, in certain situations, act as an individual
barrier and hindrance to receiving help.
At first glance, our observations seem to contradict both
empirical evidence (Andersen, 2011; Rhodes & Fiala, 2009)
and social cognitive theory, where self-efficacy is viewed as
a core psychological belief that affects basic processes of
Table 2. Results From Bivariate Unadjusted Logistic Regression Analyses.
N %
Gender,
OR [95% CI]
Age,
OR [95% CI]
General Self-Efficacy,
OR [95% CI]
Specific Self-Efficacy,
OR [95% CI]
Smoking
I do not want help 243 29 0.96 [0.70, 1.32] 1.00 [0.98, 1.01] 1.05 [0.78, 1.40] 1.40 [1.30, 1.52]
Alcohol
I do not want help 520 76 0.98 [0.63, 1.52] 1.01 [0.99, 1.02] 1.14 [0.80, 1.62] 1.12 [1.03, 1.22]
Eating
I do not want help 483 11 0.64 [0.51, 0.80] 1.05 [1.04, 1.06] 1.19 [0.98, 1.44] 1.18 [1.12, 1.25]
Physical activity
I do not want help 893 17 0.77 [0.64, 0.89] 1.04 [1.03, 1.05] 1.33 [1.15, 1.54] 1.14 [1.09, 1.19]
Note. OR = odds ratio; CI = confidence interval. Gender: 0 = male, 1 = female; age (range = 18-65 years; effect per year increase); general self-efficacy was
measured as a mean of three items (1-5; effect per unit increase in mean score). Specific self-efficacy was measured with one item (1-10; effect per unit
increase).
10 Health Education & Behavior 41(1)
change, whether this concerns contemplation to change,
mobilization of efforts, or dealing with relapses (Bandura,
2004). It is, however, important to state that our observations
do not necessarily refute previous empirical findings and that
they are quite compatible with theories of social cognition.
Given that every person may, in theory, be ranged somewhere on the self-efficacy continuum, an individual’s degree
of self-efficacy will always be of relevance when attempting
to understand or describe human behavior. As such, the
results serve as a reminder that the impact of self-efficacy is
the outcome of an interaction with environmental influences
(Bandura, 1978). In addition, our observations also suggest
that self-efficacy may affect help-seeking behavior and, by
extension, that people with high self-efficacy may prefer not
to be helped. Hence, it appears important to distinguish
between what type of behavior self-efficacy facilitates and in
which situations.
Methodological Considerations
The external validity of our findings is strengthened by the
fact that the electronic survey was sent to all occupational
positions and groups in the Danish police and had a nationwide reach. Obviously, the overall response rate of 41 % is
a weakness. The response rate inevitably raises questions
about how well the participants represent all employees.
Even so, an analysis of e-mails from participants who
actively declined to participate in our survey indicated that
the reasons for not partaking were multifold, including both
positive and negative attribution of causes. Since all information was derived from self-reports, a number of potential
sources of common method bias also need to be considered
(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). The most
critical source seems to be that all scores have been delivered by the same person (i.e., common rater effects).
Alternatively, the trustworthiness of the results is increased
by the fact that the participants had to actively select the nohelp option among the extensive list of alternatives provided
for each lifestyle factor. Another issue that warrants attention is that the participants reported a generally high degree
of self-efficacy. The general self-efficacy score was, in particular, skewed to the right, almost to the point where there
was a risk of a ceiling effect. One potential explanation for
this high degree of self-efficacy could relate to the fact that
a large proportion of our participants were police officers
that conceived themselves as fairly resourceful persons.
However, since similarly high levels of self-efficacy have
been found in the Danish Work Environment Cohort Study
(Det Nationale Forskningscenter for Arbejdsmiljø, 2005),
one cannot exclude the possibility that questionnaire respondents are, in general, more prone to possess higher selfefficacy than nonrespondents. In any event, it is plausible
that the fairly compressed general self-efficacy scores make
it more difficult to find effects. Finally, it may be noted that
most smokers wanting to quit indicated that they needed
support to do so, whereas they selected the no-help option
most frequently in relation to alcohol.
Conclusion
A high belief in one’s own ability to change lifestyle behaviors
in relation to smoking, alcohol, eating, and physical activity
may lead to a decision to avoid help offers. Although the results
presented here cannot be used to provide a reliable answer with
regard to whether such a decision is detrimental or beneficial
for the health of an individual, the results suggest that a high
self-efficacy may, in certain situations, lead to decisions that
affect help-seeking behavior in a workplace setting.
Table 3. Multiple Logistic Regression Analyses: Age- and GenderAdjusted Analyses for Specific Self-Efficacy and General SelfEfficacy.
Variable OR 95% CI p
Smoking
Specific self-efficacy 1.41 [1.30, 1.52] <.001
Age 1.01 [0.99, 1.02] .538
Gender 0.92 [0.66, 1.30] .645
Alcohol
Specific self-efficacy 1.12 [1.03, 1.22] .008
Age 1.01 [0.99, 1.02] .988
Gender 1.03 [0.66, 1.60] .973
Eating
Specific self-efficacy 1.15 [1.08, 1.21] <.001
Age 1.04 [1.03, 1.05] <.001
Gender 1.31 [1.04, 1.66] .022
Physical activity
Specific self-efficacy 1.14 [1.09, 1.19] <.001
Age 1.04 [1.03, 1.05] <.001
Gender 1.05 [0.88, 1.25] .594
Smoking
General self-efficacy 1.04 [0.77, 1.40] <.790
Age 1.00 [0.98, 1.01] .542
Gender 1.05 [0.76, 1.45] .195
Alcohol
General self-efficacy 1.15 [0.80, 1.64] .440
Age 1.01 [0.99, 1.02] .958
Gender 0.99 [0.63, 1.54] .706
Eating
General self-efficacy 1.18 [0.98, 1.43] <.087
Age 1.04 [1.03, 1.05] <.001
Gender 1.31 [1.04, 1.66] .022
Physical activity
General self-efficacy 1.36 [1.17, 1.57] <.001
Age 1.04 [1.03, 1.05] <.001
Gender 1.10 [0.93, 1.31] .256
Note. OR = odds ratio; CI = confidence interval. Gender: 0 = male,
1 = female; specific self-efficacy was measured with one item (1-10; effect
per unit increase); general self-efficacy was measured as a mean of three
items (1-5; effect per unit increase in mean score). Variables in italics are
outcome variables and those in regular font are independent variables.
Persson et al. 11
Acknowledgments
We are grateful for the cooperation of participants, the wellness
service, the employer representatives from the Danish National
Police and the Danish police, and the representatives from The
Police Union and “HK-politiet” who partook and contributed to the
project in various ways.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for
the research, authorship, and/or publication of this article:
The study received financial support from the Danish National
Police.
References
Andersen, L. L. (2011). Influence of psychosocial work environment on adherence to workplace exercise. Journal of Occupational and Environmental Medicine, 53, 182-184.
Bandura, A. (1978). Self system in reciprocal determinism. American Psychologist, 33, 344-358.
Bandura, A. (2004). Health promotion by social cognitive means.
Health Education & Behavior, 31, 143-164.
Black, D. (2008). Working for a healthier tomorrow. London,
England: Department of Health.
Blumberg, M., & Pringle, C. (1982). The missing opportunity in
organizational research: Some implications for a theory of work
performance. Academy of Management Review, 7, 560-569.
Buss, D. M. (1991). Evolutionary personality psychology. Annual
Review of Psychology, 42, 459-491.
Cialdini, R. B. (2001). Influence: Science and practice (4 ed.). Boston, MA: Allyn & Bacon.
Det Nationale Forskningscenter for Arbejdsmiljø [National
Research Centre for the Working Environment]. (2005).
Den Nationale Arbejdsmiljøkohorte-NAK [The Danish
National Work Environment Cohort Study]. Retrieved from
http://www.arbejdsmiljoforskning.dk/da/arbejdsmiljoedata/
nak2005
Linnan, L. A., Sorensen, G., Colditz, G., Klar, N., & Emmons, K. M.
(2001). Using theory to understand the multiple determinants
of low participation in worksite health promotion programs.
Health Education & Behavior, 28, 591-607.
Pencak, M. (1991). Workplace health promotion programs: An
overview. Nursing Clinics of North America, 26, 233-240.
Persson, R., Cleal, B., Bihal, T., Hansen, S. M., Jakobsen, M. O.,
Villadsen, E., & Andersen, L. L. (2013). Why do people with
suboptimal health avoid health promotion at work? American
Journal of Health Behavior, 37, 43-55.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P.
(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.
Rhodes, R. E., & Fiala, B. (2009). Building motivation and sustainability into the prescription and recommendations for physical
activity and exercise therapy: The evidence. Physiotherapy
Theory and Practice, 25, 424-441.
Wolfe, R. A., Ulrich, D. O., & Parker, D. F. (1987). Employee
health management programs: Review, critique, and research
agenda. Journal of Management, 13, 603-615.