The findings of Study 1 were used to develop a proposed model of perceived motivationally-relevant social agent influence during athlete development (Figure 6). This proposed model included five categories of perceived motivationally-relevant social agent influence that were consistent across athlete developmental stage: relationship factors, interpersonal interactions, support for development, support for performance, and feedback/evaluation. Each social agent was found to play a role in each category of
motivationally-relevant influence at some point during athlete development, but the roles and types of influence fluctuated across each developmental stage. To allow further testing, refinement and application of the proposed model, the nature and structure of the model required confirmation. This chapter reports the findings of a study which aimed to determine and refine the structure of the proposed model of motivationally-relevant social agent
influence. The next chapter then reports the findings of the second part of this study – Study 2b – which aimed to apply the refined proposed model to explore similarities and differences in perceived motivationally-relevant social agent influence between athletes at different stages of development. Study 2a and 2b are linked in that they use the same participant data to refine and then apply the proposed model of motivationally-relevant social agent influence.
5.1.1. Categories of Perceived Motivationally-Relevant Social Agent Influence. The following section will summarise the extant literature regarding each of the five categories of perceived motivationally-relevant social agent influence. This discussion will set the context for the decisions made when conducting the confirmatory factor analysis including the structure of models to be compared.
5.1.1.1. Relationship Factors.
Positive relationships with social agents can lead to positive performance and wellbeing outcomes (Davis & Jowett, 2014). Study 1 identified that the quality of the relationship between athletes and social agents that were perceived to be motivationally- relevant. The motivationally-relevant relationship factors identified in Study 1 included
friendships with social agents, social agents acting as role models, belongingness to a team, closeness, similarity, choosing which social agents to form relationships with, and continuity of relationships.
5.1.1.2. Interpersonal Interactions.
The specific motivationally-relevant interpersonal interactions between social agents and athletes which were identified in Study 1 included supportive behaviours, behaviours which promoted positive athlete emotions, behaviours which support athlete intrinsic
motivation, and behaviours perceived as being a negative influence. The interactions between participants in Study 1 and the relevant social agents were considered as being positive and negative. This dichotomy resonates with the findings of Weiss and Theeboom (1996) who identified conflict and conflict resolution as dimensions of peer friendships in sport, which were common across all social agents.
5.1.1.3. Support for Development.
Many social agent behaviours were identified in Study 1 which were perceived to be supportive of athlete development, including technical and psychological components as well as preparing athletes to contextualise their learning and prepare for competition. In line with previous research each social agent played a role in providing athletes with this type of support for their development as players, although coaches were the dominant social agent during each developmental stage in line with previous research (e.g., Keegan et al., 2014b).
5.1.1.4. Support for Performance.
Social agent behaviours, which were perceived to facilitate and support athlete performance in Study 1 participants, included practical support, social support, effort support and providing access to participation and competition opportunities. These types of support broadly align to models of social support offered by social agents in sport. For instance, Rees and Hardy (2000) identified four dimensions of social support amongst higher level
performers: emotional, esteem, informational and tangible, which were further supported by Sheridan et al.’s (2014) systematic review of social support in youth sport.
5.1.1.5. Feedback and Evaluation.
Study 1 identified many social agent behaviours which support athlete motivation by providing feedback and evaluative information. These behaviours included assisting athletes in reflection, providing praise for positive performances and helping athletes prepare for the future by placing experiences within long term development contexts. Quality of feedback from social agents has been suggested to predict athlete motivation, performance and enhance coach-athlete relationships when provided by the coach (Carpentier & Megeau, 2013).
Feedback and evaluative information have also been proposed to be a dimension within models of athlete social support (Sheridan et al., 2014).
5.1.2. Understanding the True Nature of Motivationally-Relevant Perceived Social Agent Influence.
5.1.2.1. Unclear Model Structure.
Study 1 demonstrated that at some point during athlete development, coaches, parents and peers all played a role in each category of motivational influence, yet their roles
fluctuated between the stages of athlete development. Table 8 summarises the overall influence of each social agent and how these changed across developmental stages as identified in Study 1. Similar findings have been identified in previous literature (e.g.,
Keegan et al., 2009, 2010a, 2014a) although there is a lack of clear understanding of how the profile of perceived social agent influence might change during development. For instance, the findings from Study 1 were not clear as to whether coaches, parents and peers each have unique motivational influences as perceived by athletes, or whether athletes have
motivational needs that can be fulfilled by any social agent. In order to better understand these changes in perceived social agent influence across time as well as differences in influence between stages of athlete development (i.e., the structure of the proposed model), the proposed model for perceived motivationally-relevant social agent influence needed to be tested and, if required, refined to better reflect the nature of perceived social agent influence on athlete motivation.
Table 8: Summary of Perceived Social Agent Influence Across Each Player Developmental Stage
Social Agent Sampling Stage Specialising Stage Investment Stage
Coach
Dominant social agent within sport context, although has mainly similar
influences as coaches and/or peers. Provides all aspects of motivational- influence, apart from providing access to
sport. Perceived as only social agent supporting players’ psychological development and engendering positive
feelings.
Very important influence, mainly in combination with parents and peers. Quality of relationship with players less important. Supports players with praise.
Dominant perceived social agent influence as performance level increases. Quality of relationship important especially when there is
continuity in this relationship. Perceived as sole provider of praise and
many developmental influences.
Parent
Comparatively little perceived influence at this stage. Quality of relationship not
perceived as being motivationally relevant within sport. Solely responsible
for providing access to the sport.
No unique influence – each influence shared with coaches and peers. Closeness only relationship factor of
influence. Main influences in development, performance and interpersonal interaction categories.
Sole provider of practical support. Players choosing which types of support to gain from parent. Few additional relationship factors perceived
as influential. Most influences shared with coaches and peers.
Peer
Provides wide range of influence across each category. No influence for which they are solely responsible. Relationship factors important and shared with coach.
Dominant social agent of influence involved in almost every type apart from praise. Quality of relationship
with player of particular sole importance at this stage.
Influential across many categories, but friendships with player only unique
influence. Quality of relationship continues to be important, yet players choose which peers to influence their
5.1.2.2. Possible Model Structures.
The changing nature of the profiles of perceived motivationally-relevant social agent influence make the proposed model of perceived motivationally-relevant social agent influence unclear between three possible options. Literature to date has identified individual social agent roles and roles that social agents share (e.g., Keegan et al., 2009, 2010a, 2014a, 2014b) and, therefore, multiple structures for the proposed model may exist. First, it may be possible that athletes require different types of support from each social agent at different developmental stages. This structure would be similar to findings by Vazou et al. (2006) who identified similar coach- and peer-created climates being important but no interaction
between them. Second, athletes may require a specific type and amount of support at each developmental stage, but the support is not related to social agents and therefore coaches, parents and/or peers can provide the support needed to meet their motivational needs. This additive model of perceived social agent influence would resonate with findings from Ullrich-French and Smith (2006) who identified independent and additive parent and peer influences on athlete motivation. Third, a combination of these two approaches may exist whereby some roles or types of support are specific to particular social agents, whereas others can be fulfilled by more than one social agent. Testing different models of perceived
motivationally-relevant social agent influence will allow for a greater understanding of which of these profiles best suits participant data and will then facilitate a greater understanding of whether social agents differ in the types and amount of influence they provide at each developmental stage.
5.1.2.3. Challenges in Research to Date.
The extant literature relating to perceived motivationally-relevant social agent influences suggest that there is an overreliance on studying young athletes and team sports (e.g., Harwood et al., 2015) which has limited the generalisability of research findings to other population groups. Another factor that has limited the ability of researchers to date in exploring the separate and combined influence of parents, coaches and peers on athlete motivation during development is the use of different psychometric measures for each respective social agent (e.g., Atkins et al., 2014; Davies et al., 2014). For instance, some studies have measured coach and athlete motivational influences using different
psychometric measures, making it challenging to compare the perceived influences of these respective social agents. Researchers have called for more appropriate approaches to exploring concurrent perceived social agent influence (e.g., measuring the same constructs
and ideally with the same measure for each social agent; Harwood et al., 2015). Some researchers have adapted psychometric tests and subscales that were designed for one social agent to then use to measure the construct with multiple social agents. For example, in a study exploring the independent and combined influence of peers and parents on athlete motivational outcomes, Ullrich-French and Smith (2006) used the peer-specific Sport Friendship Quality Scale (Weiss & Smith, 1999) to assess athlete-peer relationship quality, and an adapted version to assess athlete-parent relationship quality.
5.1.2.4. The Need for the Present Study
In summary, Study 1 identified five categories of perceived motivationally-relevant social agent influence that were consistent across athlete developmental stages. Previous research has identified similar categories and dimensions of perceived motivationally- relevant social agent influence. Study 1 did not, however, identify the structure of the proposed model of motivationally-relevant perceived social agent influence. For example, it was not clear whether the perceived influence of social agents was independent or additive, or whether it was a combination of these two, and research to date has also not provided clarity regarding any theoretical structure. Part of the reason that previous research has been unable to determine this has been due to studies using different psychometric measures to explore the independent or combined motivational influence of social agents, making it challenging to accurately compare different theoretical models.
To better understand the complexities of perceived motivationally-relevant social agent influence, a study was required to determine the structure of the proposed model which allowed for direct comparison between each social agent. This comparison was achieved in the present study by adapting psychometric measures developed for one social agent to be used to measure the perceived influence of each of the three social agents. Furthermore, due to complex nature of the proposed model of perceived motivationally-relevant social agent influence a cross-sectional study was identified as the most appropriate approach prior to any future research testing and applying the model beyond this thesis. Despite the cross-sectional nature of the study limiting the opportunity to determine causality (e.g., the relationship between different levels of perceive social agent influence and athlete motivation and/or performance), it was deemed an appropriate starting point for model development and refinement.
5.2. Aims and Objectives