The Role of Self-Efficacy and Friend Support on Adolescent Vigorous Physical Activity

Self-Efficacy for Physical Activity

Self-efficacy is an individual’s belief in his or her ability to perform a specific action required to attain a desired outcome. People with strong self-efficacy beliefs set higher goals, invest more effort into the pursuit of their goals, and are more likely to try harder if barriers and setbacks emerge (Bandura, 1997). Translated to physical activity this means that self-efficacy relates to individual perceptions about being able to perform physical activity regularly, to adopt and maintain it even if the conditions are difficult, and to reinitiate it when lapses occur (Bandura, 1997). Accordingly, studies using motion detectors monitoring physical activity have shown that self-efficacy is related to a high level of physical activity among 10- to 16-year-old adolescents (Strauss, Rodzilsky, Burack, & Colin, 2001). In general, adolescents with higher levels of self-efficacy for physical activity were found to report higher physical activity levels and to maintain the practice for an extended period of time (Dishman et al., 2004; Rovniak et al., 2002; Sterdt et al., 2014). As expected by social cognitive theory, self-efficacy was found to be a major instigating force in forming an intention to be active (Bandura, 1997; Luszczynska et al., 2010). Studies comparing constructs from different health behavior change theories show that the effects of self-efficacy on physical activity are stronger than those of other social cognitive determinants in adolescents (Rovniak et al., 2002).

Received Friend Support

Many of those who are regularly physically active do so in groups rather than alone—reflecting the fact that physical activity is a behavior performed mainly in social contexts among adolescents (Smith, 2003). Accordingly, social support has been identified as an important predictor of physical activity in this age group (Sterdt et al., 2014). Social support refers to the function and quality of social relationships, such as either perceived availability of help or support actually received (Schwarzer & Knoll, 2007). Studies examining physical activity outcomes have found that social support adds predictive value to both intention and behavior and has a stronger influence in predicting physical activity intention than subjective norms (Rhodes, Plotnikoff, & Courneya, 2008). In addition, the sources of support (e.g., parents, teachers, friends) can make a difference. Support provided by friends can be considered a key variable because many physical activities are performed in a social peer group context among adolescents. Furthermore, peer support was found to become more important than parental support at that age group (adolescents aged 12-16 years; Edwardson, Gorely, Pearson, & Atkin, 2013). Underlining these assumptions, studies in the physical activity domain have reported on the importance of friends as sources of social support for adolescents (Duncan, Seeley, Gau, Strycker, & Farmer, 2012; Sallis, Prochaska, & Taylor, 2000; Sterdt et al., 2014, Voorhees et al., 2005).

The Interplay of Self-Efficacy and Social Support

Social cognitive theory (Bandura, 1997) suggests various possible interactions of self-efficacy and social support. On the one hand, they could strengthen one another. This synergistic interaction was found for older adults who, when reporting to have high levels of both resources, were most active (Warner, Ziegelmann, Schüz, Wurm, & Schwarzer, 2011a). On the other hand, they could interfere with each other if very self-efficacious adolescents also receive large amounts of support, which makes them feel overprotected or controlled. This possible interaction has, to date, only been tested in the elderly, who showed preference for lower support levels if self-efficacy was high (Newsom & Schulz, 1998; Warner, Ziegelmann, Schüz, Wurm, Tesch-Römer, et al., 2011b). The third possible interaction—the compensation hypothesis—posits that both resources could compensate for that lacking in the respective other (Bandura, 1997; Dishman et al., 2009). A study that tested this compensation hypothesis in young girls, however, found that receiving lower social support for physical activity could not be mitigated by self-efficacy beliefs—only if both resources were high, girls between 13 and 18 years of age kept their activity levels—favoring the synergistic view in which both resources combined have a stronger effect than one alone (Dishman et al., 2009).

The Current Study

We aimed to examine the role of personal (self-efficacy) and social (friend support) resources and how they relate to vigorous physical activity and the intention to practice it. The first question, then, was whether both self-efficacy and friend support combined need to be present for vigorous physical activity (synergistic effect; Dishman et al., 2009; Warner, Ziegelmann, Schüz, Wurm, & Schwarzer, 2011a). The second question was whether one single resource is sufficient, which would reflect a compensatory effect of one for the other (compensatory effect; Warner et al., 2015), or whether they might interfere among adolescents (interference effect; Warner, Ziegelmann, Schüz, Wurm, Tesch-Römer, et al., 2011b). The first hypothesis pertains to a simple mediation effect. It was expected that intention to be physically active would serve as a mediator between self-efficacy and the behavioral outcome, which is in line with social cognitive theory (Bandura, 1997). The second hypothesis addressed the conditional effect that was expected when friend support comes into the equation as a putative moderator of the self-efficacy–intention relationship, as the effect of self-efficacy on intention may depend on levels of received friend support.

Method

Results

Means, standard deviations, and intercorrelations are displayed in Table 1.

OPEN IN VIEWER

Table 1.

Means (M), Standard Deviations (SD), and Correlations.

Variable	M	SD	1	2	3	4	5	6
1. Sex (137 female, 89 male; coded 1 and 2, respectively)			—	−.08	−.10	.15*	−.04	−.03
2. Age (in years)	13.50	0.59		—	−.01	−.05	−.05	.00
3. Self-efficacy (scale 1-7)	6.13	0.91			—	.27***	.81***	.12
4. Friend support (scale 1-4)	2.39	0.94				—	.34***	.28**
5. Intention (scale 1-7)	6.03	1.10					—	.20**
6. Vigorous physical activity (time in minutes)	82.94	63.74						—

*

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

Testing the simple mediation hypothesis for the relationship from self-efficacy to vigorous physical activity via intention yielded an indirect effect of b = 12.83, 95% confidence interval (CI) [2.24, 24.40]. This confirmed the assumptions and allowed to proceed to the conditional process analysis, examining the second hypothesis. Figure 1 displays the conditional process model with standardized parameter estimates, in which a moderated mediation takes place, as reflected by the interaction term of self-efficacy and friend support on intention.

OPEN IN VIEWER

Figure 1.

Interaction between self-efficacy and received friend support on vigorous physical activity.

In detail, the following unstandardized parameters were estimated. The effect of self-efficacy on intention was b = 1.17, 95% CI [0.97, 1.36]; the effect of friend support on intention was b = 0.93, 95% CI [0.40, 1.46]; and their interaction was b = −0.13, 95% CI [−0.21, −0.04]. Of the intention variance, 72% were accounted for by this set of three predictors, which, however, was mainly due to self-efficacy (while also controlling for age and sex). On the right side of the model, the effect of intention on vigorous physical activity was b = 13.37, 95% CI [0.32, 26.42]; the direct effect of self-efficacy on vigorous physical activity was not significant, b = −4.08, 95% CI [−19.62, 11.45]; and sex and age were nonsignificant.

To probe the interaction, the regression lines were plotted using three specific values of the moderator friend support, namely, those at the mean of the moderator, and at ±1 standard deviation from the mean (−1 SD, M, +1 SD). Figure 2 illustrates that at high levels of self-efficacy, there was also a high level of intention, independent of support. At this level of self-efficacy, it did not matter how much friend support was received, self-efficacy predicted intention. On the other hand, at low levels of self-efficacy, support made a difference: If students lacked self-efficacy as well as support, they were less likely to form an intention for vigorous physical activity. In adolescents with low self-efficacy, who received above average friend support, intention to engage in physical activity was facilitated. This interaction confirms the second hypothesis that the resources do not only add up but also interact in a way that friend support partly compensates for low levels of self-efficacy in adolescents. However, the high support slope was significant as well, indicating that support could not fully buffer lack of self-efficacy. The conditional indirect effect of self-efficacy on vigorous physical activity was highest in the subgroup of students receiving lower levels of support, pointing to the fact that highly supported individuals’ self-efficacy was mediated less strongly via intention on vigorous physical activity.

OPEN IN VIEWER

Figure 2.

Interaction between physical activity self-efficacy and received friend support on intention.

Discussion

The current study examined the joint effects of friend support and self-efficacy as predictors of intention and vigorous physical activity in adolescence—a phase in life that is known to be characterized by decreasing physical activity levels (Dumith et al., 2011) and where additional vigorous activity can have added benefits (Department of Health, 2014; Gutin et al., 2005). The first hypothesis examining the simple mediation between self-efficacy and vigorous physical activity by intention was confirmed, which is in line with most health behavior theories such as social cognitive theory (Bandura, 1997). Moreover, there was a direct effect of self-efficacy and social support on intention and also an interaction, demonstrating that the strength of relationship between self-efficacy and intention depended on how much support was received from friends. The conditional process analysis was performed to examine the second hypothesis pointing to a moderated mediation. If adolescents have low levels of self-efficacy, they need friend support to build their intention, and indirectly, vigorous physical activity at a later point in time. Lack of friend support, however, is no disadvantage when adolescents harbor a high level of self-efficacy. This means that high self-efficacy compensates for low support, but high friend support can only partly buffer low self-efficacy. Overall, the results of the current study provide initial support for the suggestion of a synergistic effect (Dishman et al., 2009; Warner, Ziegelmann, Schüz, Wurm, & Schwarzer, 2011a) as well as compensatory effect (Dishman et al., 2009; Warner et al., 2015) between self-efficacy and social support for adolescent vigorous physical activity. High levels of both resources were found to be beneficial in forming an intention to be active among adolescents and did not lead to adverse side effects due to possible reactance of high self-efficacious and highly supported individuals. The current study did not find evidence for the interference hypothesis between self-efficacy and social support, which was found in older adults, who showed lower levels of autonomy if they reported to have high levels of both resources (Warner, Ziegelmann, Schüz, Wurm, Tesch-Römer, et al., 2011b).

A meta-analysis on the effects of physical activity interventions on body mass index of children found that these interventions often fail to increase the amount of physical activity and, therefore, did not result in the expected health benefits (Metcalf, Henley, & Wilkin, 2012). A recent review and meta-analysis on the effects of interventions that aimed to increase moderate and vigorous physical activity in school lessons, however, indicated that interventions during physical education lessons can increase the proportion of time that students spend exercising (Lonsdale et al., 2013). Along these lines, integrating physical activity in a peer context might be especially important in adolescence. To inform future interventions for vigorous physical activity in adolescents, the results of the current study add some practical suggestions: interventionists who target either self-efficacy, peer support, or both to increase physical activity among adolescents do not have to worry about “overdosing” friend support, since even students who believed in their ability to perform physical activity regularly showed no adverse effects of additional friend support. This beneficial effect of support should, however, not be taken for granted from all sources of support (e.g., not be transferred to parents and teachers) as Bandura (1997) warns that verbal persuasion might cause opposite effects in recipients if it is perceived as controlling (Verloigne et al., 2014).

The results of the current study also suggest that it is most important in adolescents to target self-efficacy, because self-efficacy can make up for low support from friends (compensation effect). For adolescents with low self-efficacy, increasing friend support will, however, only partly increase their intention—a combined support and self-efficacy intervention might show better effects for adolescents with low self-efficacy. Furthermore, the indirect effect of self-efficacy on physical activity was highest in the subgroup of students with low support, which means that they needed a strong intention to be physically active, whereas those with much support did not necessarily need a strong intention to translate their self-efficacy into physical activity. This finding can be interpreted in the way that those adolescents, who reported a great deal of support, were the ones who were socially motivated for action. This motivation is possibly due to participation in group programs or memberships in sports teams where commitment to the group activity plays a major role and time and place are predefined—potentially making individual self-regulatory processes less important. Taken together, these results strengthen previous research that found peer support to be a key resource for physical activity in adolescents and that suggested to make parents aware of this strong effect of peers, support exercising in groups and clubs, and trying to increase adolescents’ ability to provide adequate support for one another (Edwardson et al., 2013).

Some conceptual and methodological limitations of the current study need to be mentioned to evaluate these results and their implications for future research and practice. Given the main focus of this article was to investigate the interplay of self-efficacy and friend support on intention and vigorous physical activity, the behavior measure only assessed the vigorous physical activity performed “yesterday.” To investigate changes in naturally occurring vigorous physical activity over time, baseline measures of behavior as well as longer follow-ups would be advisable. In addition, behavior was assessed using a self-report 1-day physical activity recall. This decision was based on research suggesting that young people are better able to report their physical activity more accurately if asked for a 24-hour recall rather than longer periods (Van Hoye et al., 2014). Although such self-reports are a frequently utilized practice in research on physical activity, and the validity of single-item self-reports has been shown to be satisfactory (Hamilton, White, & Cuddihy, 2012; Prince et al., 2008), measuring physical activity with daily diaries of several 24-hour recalls or objective methods such as accelerometry may be less prone to memory bias or social desirability bias (Van Hoye et al., 2014; Welk et al., 2004). A further limitation was the study population only consisting of one grade level (i.e., Grade 9) and those who had recent experience of engaging in vigorous physical activity. Given that previous research has found school grade–level differences among social cognitions (Mummery, Spence, & Hudec, 2000) as well as potential differences among active and nonactive individuals (Bauman et al., 2012; Hamilton & White, 2010), future research should investigate if these effects hold across a broad range of school grade groups and different physical activity intensity levels. Finally, the prospective design of the current study does not eliminate mechanisms of reversed causation, for example, that more vigorous physical activity leads to building more sustained self-efficacy and positive responses from friends and also intention to maintain that level of physical activity.

In conclusion, despite the added benefits of vigorous physical activity, adolescents do not engage in such exercise enough. Self-efficacy and social support are identified as key determinants of physical activity; however, the mechanism that reflects the interplay of these two factors is not well understood. The current study sheds some light on this relationship. In sum, this study confirms the assumption of an interactive resource mechanism that has been found in some of the literature on physical activity in older adults suggesting a synergistic effect between both resources as well as a compensation of low support by high self-efficacy and partly also the other way round. Further research needs to manipulate either self-efficacy or support or both factors to unveil possible causal pathways that may be responsible for such mechanisms.