Introduction
Coaching is a social process, comprising a series of negotiated outcomes between structurally influenced agents within an ever-changing environment (Saury and Durand, 1998, Poczwardowski et al., 2002, Cushion et al., 2003 and Cushion and Jones, 2006,). In this respect, the coaching process can be considered the result of the dynamic interaction between coaches, athletes and the socio-cultural context (Smith and Smoll, 1993, Côté et al., 1995a, Côté et al., 1995b, Langley, 1997, Saury and Durand, 1998 and Cushion et al., 2006,). In these constructed relationships neither coach, the player, nor the context has the capacity to unilaterally determine action; the key to understanding the coaching process lies in the relationships between the three variables (Cushion 2008). However, while coaching is best conceptualised and understood as a series of inter-related and interconnected relationships (Lyle, 2002 and Cushion et al., 2006), at the heart of these relationships, and as a consequence the coaching process, the coach plays an essential and highly influential role. It seems a statement of the obvious, but it is a statement worth making nonetheless; the coach occupies a position of centrality and considerable influence in efforts to improve sporting performance (Cushion et al., 2006 and Smith and Smoll, 2007).
Indeed, the coach has been identified as a powerful socialising agent in the physical domain (e.g. Amorose, 2007 and Horn, 2002). The coach’s behaviour impacts on athletes’ behaviour, cognitions and affective responses, and coaches can influence whether athletes learn and achieve at a high level, enjoy their experience, demonstrate effort and persistence, and develop a sense of confidence and a self-determined motivational orientation (Smoll and Smith, 2002, Mageau and Vallerand, 2003 and Amorose, 2007). Moreover, coaching behaviour delivered or interpreted incorrectly or inappropriately can lead to negative outcomes (e.g. poor performance, low self-esteem, high levels of competitive anxiety, burnout) (Amorose 2007). Given this, it follows that through their words and actions (i.e. their behaviour), coaches will not just impact performance but also the social and emotional well-being and perspectives of their athletes (Miller, 1992, DeMarco et al., 1996a, DeMarco et al., 1996b, Jones et al., 1997 and Horn, 2002).
With this in mind we can see that coaching practice has affective, cognitive, behavioural and social elements. On a strictly behavioural level coaching can also be considered in terms of antecedents (the stimuli prior to the event); the behaviour (what is done or said); and consequences or what follows from the behaviour. The traditional or common sense view of coaching has tended to focus solely on the observable behavioural elements, and has paid much less attention to the ‘what’ or ‘why’ of the behaviour. Indeed, the most common focus for coaching research since the 1970s has been concerned with the coach’s behaviour. As a result this research has emerged as one of the largest single categories within the general body of coaching knowledge (Gilbert & Trudel 2004). This approach to coaches and their behaviour has resulted in coach behaviour being an area of significant research activity (Nash & Collins 2006). Research into this aspect of the coach’s influential role has been rightly warranted. However, rather than embracing the complexity of the coaching role, both with the coach as a person and with how he or she engages with the wider pedagogical process, the majority of coaching research to date has been content simply to describe how coaches behave during practice, and to a lesser extent, competitive situations (Kahan 1999, Lyle 1999, Cushion and Jones, 2001 and MacPhail and Kirk, 2001). Even within the behavioural literature it is unlikely that we would find an exploration of the antecedents or consequences of coach behaviour. It is against this backdrop that the coach behaviour literature should be considered.
This approach can perhaps be attributed to coaching’s traditional location within a dominant psychological discourse, which, in turn, has its epistemological roots in the positivistic natural sciences (Ward & Barret 2002). A key characteristic of behavioural analysis is the production of objective, reliable and valid data, free from the distortion of suggestion and perception, and as a result systematic descriptive-analytic systems have been widely used as research instruments to gather information on coaches while coaching (Kahan 1999). This approach can also be attributed to an initial desire by many early scholars for some dispassionate baseline data to find out ‘what (good) coaches do’ in what was, not so long ago, a very under-researched field. Despite its original popularity, in more recent times such methods and the resultant findings have somewhat fallen into disrepute as many authors have tended to highlight their weakness (e.g. Abraham & Collins 1998).
Clearly, coaching behaviours per se do not stand alone as predictors of effective coaching (Douge & Hastie 1993) nor do they ‘embrace the entirety of the coaching process’ (Lyle 1999a p 14). Since this was never their purpose, it would be wrong to criticise them for not doing so. However, given this, what is the value of a chapter considering coach behaviour and its related research? The initial rationale for this area of endeavour remains as valid today as it did in the 1970s; coaches are central to the coaching process and what they say and do continues to impact performers’ achievement and well-being. Therefore, understanding which behaviours translate into positive experiences and functioning on the part of the athletes is critical for researchers and practitioners alike (Amorose 2007). Moreover, while coach behaviour may be portrayed as ‘superficial’ in one sense, we still know relatively little about coaching practice, good or bad, across a variety of settings (Lyle, 2002 and Cushion et al., 2006, Portac et al 2007), and thinking about what coaches do and why they do it, still offers much in developing our understanding about coaching.
A useful database of coach behaviour does exist, and what this body of work has done is to identify (within certain constraints) ‘tried and tested coaching behaviours’ (Douge & Hastie 1993 p 54). This has resulted in claims that ‘instructors within sport have available to them an extensive and growing knowledge base from which to make decisions about their practice’ (Hughes & Franks 1997 p 190). The term ‘instructors’ with its connotations of a more restricted performance improvement role is perhaps a revealing comment about the value of ‘behavioural’ evidence. Despite this knowledge base, however, coaching practice remains largely ‘belief’ as opposed to ‘evidence based’ (Rushall 2003). It is important therefore, that as our understanding of coaching becomes more sophisticated, and research priorities change, we should not disregard existing accumulated knowledge but, rather, consider ways to integrate new knowledge with what is already known (Cushion 2007). It is in the interests of coaching and its development that we integrate existing work and ideas into a more sophisticated knowledge base (Rink, 1993 and Cushion, 2007). In essence, how can we use and integrate what we know about coach behaviour into contemporary debates about coaching practice and the nature of coaching? This chapter goes some way to addressing this question.
Finally, and perhaps most importantly, coaches are notoriously poor at describing their own behaviour. Indeed, research has demonstrated that coaches have limited awareness of how often they behave in various ways, and that performer’s ratings correlate more strongly with observed behaviours than the coaches own self-ratings (Smoll & Smith 2006). Therefore, illustrating the actions performed by the coach in practice and competition situations (Borrie & Knowles 2003) can contribute to raising the awareness of coaches to what they are actually doing. The most sophisticated understandings of coaching practice and advances in coach education would seem fruitless if coaches lack seemingly basic levels of self-awareness.
In terms of this chapter, it is not the intention to provide a full comprehensive ‘review’ of all coaching behaviour research to date, a task which, if done to an insightful level of critical depth, would inevitably fall outside the constraints of a single chapter. Rather, the objective here is first, to locate and critique key elements of coach behaviour, to attempt to identify principles for practice, and to identify where further research is required. Second, it is intended to examine critically the core concepts and assumptions associated with, and inferred through, such research. Such critical mapping provides us with a progressive way both to evaluate past and to process future research (Hart 1998). The significance of this work is also grounded in an attempt to improve the clarity and applicability of past and future findings for those who can then use them in the practical setting. Indeed, the utility of a body of knowledge such as coach behaviour lies in its usage in assisting everyday coaching practice. Hence, this chapter marks an attempt to develop further the relationship between academic presentation and dissemination and the concrete (reality) of coaching experience and need.
The chapter is divided into two parts. The first part examines current coach behaviour research; the objective is to identify key consistent behaviours identified in the research (what coaches do) and develop some explanations for this behaviour (why coaches behave this way) before going on to consider the gaps inherent in the knowledge produced to date. The second part of the chapter attempts to synthesise this discussion with more contemporary debates about coach behaviour and considers the dualism in coaching created by directive versus non-directive approaches. Finally, an attempt is made to draw these arguments together and develop some ideas about the nature of coaching behaviour with some consideration given to potential coach behaviour strategies.
The study of coach behaviour
Background
Initially, behaviour analysis and the use of descriptive analytical instruments were prominent in teaching where research focusing on the description and analysis of physical education instruction gathered momentum during the 1970s (Lawson 1990). This research was perceived as ushering sport pedagogy into ‘an era of legitimacy, innovation, and unparalleled activity’ (DeMarco et al., 1996b and Kahan, 1999) with the research resulting in a wealth of information on the type and quality of practitioners’ instruction (DeMarco et al 1996b). Investigations into coach behaviour grew from this foundation. These were perceived by some to have increased legitimacy as the behaviour of coaches was viewed as being more directly related to outcomes, such as win/loss (Claxton 1988). Considerable enthusiasm greeted this development, with many authors stressing its importance in establishing an empirical base for a future ‘science’ of coaching particularly related to coach behaviour (Lacy and Darst, 1985, Lacy and Goldston, 1990, Seagrave and Ciancio, 1990 and Trudel et al., 1996).
A prominent feature of this work was and continues to be, the examination of observable coach intervention, in particular, coaches’ instruction. Tharp and Gallimore (1976) were pioneers in this field, conducting a ground-breaking and extensive study of legendary UCLA basketball coach John Wooden (re-visited recently, see Gallimore & Tharpe 2004). This research was replicated and built upon by many others and further inspired the widespread development and use of systematic observation instruments for examining coaches’ work in training and competition environments (Williams, 1978, Langsdorf, 1979, Smith et al., 1979 and Lucas, 1980). Subsequently, a number of manual and computerised observation systems have been developed specifically to analyse coaching behaviour and have been used across a number of sports and at various levels of competition (e.g. Tharp and Gallimore, 1976, Rushall, 1977, Smith et al., 1977, Langsdorf, 1979, Smith et al., 1979, Lucas, 1980, Quarterman, 1980, Metzler, 1983, McKenzie and Carlson, 1984, Crossman, 1985, Lacy and Darst, 1985 and Franks et al., 1988).
The constant production of articles since 1975, from different authors, using various observation instruments, and focusing on different aspects of coach behaviour, would appear to indicate that such a method is indeed appropriate for describing coaches’ behaviour in training and competition (Trudel et al., 1993, Kahan, 1999 and Gilbert and Trudel, 2004). Undoubtedly, the studies conducted have yielded insights that have contributed greatly to the body of knowledge in sport pedagogy (DeMarco et al., 1996a, DeMarco et al., 1996b and Jones, 1997), providing a wide range of literature describing coach behaviour in an array of sports including for example: basketball, soccer, athletics, archery, tennis, American football, swimming, volleyball, and rowing (Dubois, 1981, Lacy and Darst, 1985, Claxton, 1988, Lacy, 1989, Lacy and Goldston, 1990, Seagrave and Ciancio, 1990, van der Mars et al., 1991, Miller, 1992, Salminen and Liukkonen, 1995, Millard, 1996, Potrac et al., 1997, Vangucci et al., 1997, Bloom et al., 1999 and Cushion and Jones, 2001). Such studies represent the beginnings of a database of coaching behaviours, fulfilling in part the goal set by Lacy and Goldston (1990) that, to be meaningful, observation should be conducted in a variety of settings.
Part 1: Key behaviours (what coaches do and why)
This section considers coach behaviours that consistently appear in the findings of coach behaviour research. Of course the complex nature of coaching and the coaching process means that the different elements of coach behaviour are in fact interrelated and interdependent (Cushion et al 2006, Smith & Cushion 2006), and coaches will engage with and deploy a range of discrete behaviours. Moreover, in reality, behaviours overlap and do not occur in neat ordered ways, and whilst they are described as such here, these behaviours should not be thought of as separate or isolated.
The section is organised around three categories of behaviour: (1) instruction, (2) praise/scold, and (3) silence; and together these behaviours account for approximately 80% of what research has identified that coaches do. This analysis is by no means definitive, but attempts to locate these key behaviours within a number of analytical frameworks. What it does do, however, is to attempt to look beyond the superficiality of ‘just behaviour’ to offer insight into why these behaviours occur and go some way to demonstrate the complexity of the coaching task.
Instruction
There is evidence to link effective coaching and athlete learning to the quality of the coaches’ instructional behaviour (Carreira Da Costa and Pieron, 1992, DeMarco et al., 1996a, DeMarco et al., 1996b and Gallimore and Tharp, 2004, Hodges & Franks 2004). However, as the discussion here will illustrate, coaches validate the use of instructional techniques based on a variety of reasons, not solely upon their effectiveness for athlete learning. Moreover, there are principles that underlie the provision of effective instruction (Hodges & Franks 2004), but these principles alone would seem unable to account for the dominance of such behaviour in coaching, both in terms of individual coaching sessions and across coaching sessions over time.
With this in mind, instruction remains the most frequently used ‘active’ coaching behaviour; a finding that consistently permeates research literature. Moreover, behaviours that relate to task accomplishment, e.g. training and instruction and positive feedback, are generally speaking the most preferred by athletes (Chelladurai & Riemer 1998, Reimer 2007). How can we account for this?Jones et al (2004) contend that coaches use behaviours that they perceive are congruent with the coaching role. Prior socialisation (as an athlete and a coach), along with established beliefs and traditions that validate and acknowledge certain behaviour as ‘effective’ serves to reinforce this image (Cushion and Jones, 2006 and Potrac et al., 2007). Therefore, high levels of instruction reflect beliefs about effective and appropriate coaching behaviour that derive from previous playing (athlete as recipient) and coaching experiences, and reproduce and reinforce an ‘instructional’ and directive approach (Potrac et al 2007). As Goffman (1959) suggests, this behaviour is a performance shaped by the environment and audience, and constructed to provide others with impressions that are consonant with the desired goals of the actor (Potrac et al 2002).
This emphasis on instructional behaviour can also be seen in terms of fulfilling the requirements of the role, particularly when they are strongly associated with performance success. The pressure to succeed in sport and the accountability and kudos for such success (real or perceived) sees the coach attempting to control as many variables as possible through high levels of instructional behaviour (Bloom et al., 1999, Coakley, 2004 and Potrac et al., 2007). ‘Responsibility for the outcome of performance is a notable determinant in understanding a coach’s desire to be in control of his or her respective athletes and coaching situations’ (Eitzen and Sage, 1989 and Potrac et al., 2002).
To achieve success, coaches perhaps not only perceive the need to have the respect of their athletes, but also the capacity to impact significantly on their athletes’ performances thus achieving the desired outcomes (Potrac et al., 2002, Potrac et al., 2007 and Jones et al., 2004). This means not being viewed as weak, indecisive or lacking in knowledge or expertise (i.e. asking athletes for solutions). As such they tend to rely on safer, tried and tested traditional methods that prove their knowledge and expertise, namely instruction (Coakley 1994, Potrac et al., 2002, Potrac et al., 2007 and Jones et al., 2004). ‘The consequence of such action is that athletes are, in turn, increasingly socialised into expecting instructional behaviours from coaches, and thus resist other coaching methods’ (Potrac et al 2007 p 40), as these behaviours are deemed consciously or subconsciously to be associated with performance accomplishment. As a result, coaching behaviours like instruction become a historical and traditional thread passed on through an ‘apprenticeship of observation’ (Schempp, 1989 and Cushion et al., 2003) with experiences becoming a powerful, long-lasting, and continual influence over pedagogical perspectives, practices, beliefs and behaviours.
Silence
Silence has been identified as a significant behaviour within coaching practice and can account for up to 40% of a coach’s total behaviours in both training and competition (Potrac 2001, Smith & Cushion 2006). This would seem logical as clearly coaches cannot constantly be engaged in ‘active’ coaching behaviours (Miller 1992). However, coaches whose observation has been interpreted as ‘passive’ have been described as ‘off-task’ (Claxton 1988). This suggestion has been superseded, with silence becoming increasingly understood as a deliberate coaching strategy, being used as a tool for promoting learning (Cushion & Jones 2001, Smith & Cushion 2006, Potrac et al 2007).
Indeed, research (Cushion & Jones 2001, Smith & Cushion 2006, Potrac et al 2007) describes a pattern of coaching behaviours in which periods of silence are punctuated with verbal cues, short reminders and specific commands or correction. When the coaches were not ‘actively’ providing feedback, they were in fact ‘on-task’, intently watching the action in silence. Therefore the research evidence would clearly indicate that the coaches’ silence was indeed an intentional modus operandi.
Interestingly, the coaches in the Smith and Cushion (2006) study expressed concern that too much verbal intervention would deny the athletes not only opportunities to learn but also the opportunity to demonstrate what has already been learnt. Clearly, during moments of silence, coaches are involved in a number of cognitive processes. These include observing and analysing the athletes’ performance, allowing opportunities for the athletes to learn for themselves, and checking learning and athletes’ decision-making.
Positive reinforcement; praise and scold
Research suggests that coaches consider a positive working climate a vital part of delivering high-quality coaching practice (Jones et al., 2004 and Potrac et al., 2007). Indeed, Smith and Smoll (2002) have identified a positive approach to coaching where ‘effective’ coaching behaviours include: high frequencies of reinforcement for effort and performance, encouragement following errors, mistake contingent and general instruction, while minimising punitive behaviours and non-responses. This approach is supported by Amorose and Horn (2000), who champion behaviour that contains high-frequency positive and informational feedback and low frequencies of punishment-oriented feedback.
With this in mind it appears that coaches frequently provide liberal support and encouragement to their performers. This is evidenced by behavioural research that reports praise as a substantial element in coaching practice, often being the second or third largest category of behaviour overall (e.g. Lacy and Darst, 1985, Miller, 1992, Bloom et al., 1999 and Cushion and Jones, 2001, Portac et al 2007). In attempting to account for this, Potrac et al (2002) contend that coaches understand the significance of establishing a positive learning environment, recognising that athletes are more responsive to a positive coach and that positive behaviour can impact motivation and self-efficacy in athletes. More subtly than this, the coaches could use this behaviour to reinforce desired athlete behaviour. As Benfarri et al (1986) note, the method and style of transmission is critical in forming the recipients’ perception of it. However, too much praise can be seen as losing its value and thus becoming meaningless to athletes. This suggests that the overuse of praise, especially general praise, can be interpreted as non-specific feedback that dilutes its effects (Schmidt 1991). This can usefully be understood in terms of reward power; the power that comes from a person’s control over another’s rewards (Slack, 1997 and Potrac et al., 2002). Maintaining the legitimacy of reward power, in this case praise, will require that the recipient perceives the behaviour as having meaning. Thus, overusing praise or giving praise when not deserved, will reduce its value (Potrac et al 2002).
Alongside high levels of praise and contributing to a notion of a ‘positive’ coaching environment is a low level of scold behaviour, with some recorded praise to scold ratios recorded as high as 33:1 (Potrac et al 2002). The ability to punish another can be regarded as ‘dysfunctional because it alienates people and builds up resentment’ (Slack 1997 p 181). Hence coaches’ restricted use of scold behaviour is perhaps evidence of recognition of the unproductive nature of such behaviour and the potential damage that can be caused to coach–athlete relationships (Jowett 2007).
What and why: expectancy effects
One factor that has been shown to influence coach behaviour is coaches’ expectations or judgements of their players, which has been shown to ‘influence the athletes’ performance and behaviour’ (Horn & Lox 1993 p 69). This is an important point because it illustrates the bi-directional or dialectic nature of coach behaviour (LaVoi 2007). It is often assumed that the athlete is a passive recipient, that the coach’s behaviour is successfully received, and that neither coach nor athlete is affected by the interaction. However, Poczwardowski et al., 2002 and Cushion, 2008 and his colleagues (Cushion et al 2006) illustrate that coach behaviour is not benign and that there is a dialectic relationship with both coach and player influenced by their interaction.
Given the competitive and evaluative nature of sport, it is common for coaches to form expectations of athletes at the beginning of a season (Solomon, 1998 and Wilson et al., 2006). These are initial judgements or assessments about the physical competence or sporting potential of each athlete and are based on certain information available to the coach. A coach’s expectations develop from a variety of factors but generally fall into two categories; person cues and performance information (Horn & Lox 1993). Person cues include, for example, the player’s age, social background, ethnicity, and physical attributes. These cues are not used exclusively but are combined with performance information about the player that could include: past performance, physical test scores, ability, and comments from other coaches (Horn and Lox, 1993 and Solomon et al., 1996). Coaches also use initial impressions of athletes in practice situations informed by, for example, observation of the player’s motivation, enthusiasm, pleasantness, response to criticism, and interaction with staff and team mates (Horn & Lox 1993). In addition, it is possible, if not likely, for two coaches to form different sets of expectations for an athlete based on which aspects of the available information they value most (Horn & Lox 1993). These expectations are expressed to the athlete through verbal and non-verbal behaviours (Solomon et al 1996).
The major sport research linking feedback and expectations has found that differential feedback is issued to high- and low-expectancy athletes but is not consistent. For example Cushion and Jones, 2006 and Wilson et al., 2006 identified significant differences in the nature of coach behaviour based on an athlete’s position in the group with clear ‘favourites’ and ‘rejects’ suffering differential treatment within the coaching process. The conclusions from these and other works, have again shown that some coaches appear to be ‘Pygmalion-like’ with differences between the quality and frequency of athlete–coach interaction (Horn & Lox 1993). Expectancy theory, applied to the sport setting, assumes that coaches’ feedback patterns will differ because athletes are perceived as high or low expectancy (Solomon et al 1996). However, this assumption is worth considering further. Put simply, expectation effects are responsible for some differences in coach behaviour, but not all.
Perception of ability may not be the only basis for differential treatment and behaviour patterns. In accordance with the principles of behaviour modification and reinforcement theory, coaches ordinarily use performance feedback to cultivate and shape desired athlete behaviours (Sinclair & Vealey 1989). Markland and Martinek (1988) noted that players of more successful coaches (defined by win/loss record) received more feedback than their less-successful peers. This finding was mirrored by Solomon et al (1996) who, while investigating expectancy effects, found that head coaches and assistants differed in the amount of feedback given. Here, head coaches gave feedback based on mistakes whereas assistant coaches delivered more general positive feedback. Moreover, Claxton (1988) found that expert coaches used more questioning behaviours and would give less instruction than their non-expert colleagues (Claxton 1988).
This research demonstrates other possible reasons for differential feedback, namely the coaches’ ‘success’ or expertise. Indeed, differential feedback and differences in coach behaviour have been attributed to numerous factors related to the coach. These include for example job expectation, interaction opportunities, differentiation of roles, the philosophy of the programme, and the goals of the sessions (Solomon 1998).
Thinking about coach behaviour
The existing research evidence suggests that ‘what coaches do’ in terms of coaching behaviour is a judicious mix of the instructional and positive and that this is a considered approach with silence used as a deliberate behavioural strategy to enhance learning. Indeed, several authors advocate a positive approach to behaviour in general and link this positively with improved levels of performance, team cohesion, and player self-esteem (Dubois, 1981, Wandzilak et al., 1988, Amorose, 2007 and Smith and Smoll, 2007). This notion is supported by Saury and Durand (1998), who further reported that expert coaches are empathetic, tactful, give autonomy and share knowledge and experience. There is an intuitive appeal to this positive approach but unfortunately effective coach behaviour is more complex than implied distinctions between positive or negative, directive and non-directive (Amorose 2007). Indeed, considering the application of these behaviours in practice provides a glimpse of coaching’s complexity.
First, there is evidence that suggests that if it is used inappropriately negative affective outcomes can occur from praise and instruction, while if perceived incorrectly socially supportive behaviours can lower motivation (Amorose and Horn, 2001, Mageau and Vallerand, 2003 and Amorose, 2007). Thus, the skill or craft of the coach in applying appropriate behaviours is likely to be hugely influential.
Second, there are numerous ‘negative cases’ where the coaches’ behaviour may be considered contrary to the principles espoused in the prevailing literature. For example, the coaches in research by d’Arrippe-Longueville et al (1998) investigating elite judo were highly autocratic, showed low levels of social support and used their behaviour as part of a ‘toughening up’ process for their players. The researchers found that the athletes did not like this autocratic coaching behaviour but acknowledged it to be the most effective. d’Arrippe-Longueville et al (1998) explain this by proposing that the culture of the sport strongly influences coach behaviour. Indeed echoing these sentiments, Parker (1996), Tomlinson (1993), Roderick, 2006 and Cushion and Jones, 2006 argue that there remains an underlying authoritarian character in some sporting sub-cultures and this has a pervasive and influential effect in the coaching process and on coach behaviour. Consequently, differences in coach behaviour may indeed be reflective of ‘deep seated cultural differences, inherent differences in belief systems, value structure and underlying cultural patterns that are embedded’ (Greendorfer 1982 p 198) within it.
Therefore, there would be merit in understanding how much coaching behaviour is ‘traditional’ and how the coach’s current practice is related to a selectively remembered past, which acts to validate and acknowledge it as ‘effective’ (Cushion & Jones 2006). Moreover, we perhaps need to probe more deeply and examine processes of socialisation among other influencing factors, before trying to draw conclusions about coaches’ good practice and the practice of the ‘good’. In addition, despite their historical and sub-cultural connections within the context of athlete development and well-being, it might be pertinent to consider more critically the evidence supporting certain behaviours and whether these practices might be legitimately justified.
Identifying gaps and problematising existing knowledge
Clearly, the range and variety of opportunities to observe coach behaviour is considerable. Not only does this relate to observations within differing cultures but also to the number of different sports available and to the various levels or domains in each sport. Even within a particular sport and at specific levels of competition, coach behaviour could be examined in both contest and practice environments and at different times in the season. As these variables have been demonstrated to have an impact on coaches’ actions, and despite the general volume of work undertaken (Kahan, 1999 and Potrac et al., 2002), there appear to be several ‘gaps’ in existing coach behaviour knowledge.
For example, the majority of coaching research has been conducted in predominantly educational settings within North America. It would be naive to accept that this body of knowledge can be unproblematically applied to coaches in Europe as a consequence of an assumed transatlantic ‘validity’. Coaching knowledge, however labelled, can never be so culture-proof (Nicholas 1983). Yet, despite this rather obvious limitation, there has been insufficient account taken of the culture and contextual differences in how the body of knowledge has been interpreted (Brewer & Jones 2002). Furthermore, professional as opposed to youth or university sport as a context for analysis remains very much under-researched. Clearly, the demands and pressures upon professional coaches in ‘performance’ sport (Lyle, 1999a and Lyle, 2002) are different to those upon other coaches who operate within a more recreational ethos. Indeed, Lyle (2002) argues that performance coaches fulfil a distinctive role and are not part of a sporting continuum from participation to performance; hence, their functions and actions will reflect that uniqueness. Although the clarity of this distinction is open to debate (Jones 2006), certainly evidence exists that the performance coach is engaged differently within the coaching process in terms of time, detail and emotion than the ‘participation’-orientated coach, who is not so nearly concerned with the result of the next competition. Gaps are clearly evident in the existing knowledge base, as coach behaviour within many coaching environments remains largely under-researched.
A principal shortcoming in this respect lies in the lack of research on coaches’ in-competition behaviour. It is understandable that the actions of the coach have been observed more frequently in practice settings than in games or competitions. Simply there are fewer games than practices, and the coaches’ behaviour in competition may be more variable, context-specific, and difficult to capture. The practice environment therefore, provides greater scope and opportunity to observe coach behaviour. This issue of availability has been reflected in the number of studies undertaken, with observations of how coaches manage the competitive ‘game’ environment being conspicuous by their absence (Kahan 1999, Smith & Cushion 2006). Those projects that have been carried out under both settings, have found that certain behavioural categories register differently under game as opposed to practice conditions, hence the findings from one can not be assumed for the other (Horn, 1984, Chaumeton and Duda, 1988, Wandzilak et al., 1988 and Trudel et al., 1996). For example, coaches engage in less instruction and overall interaction with their athletes during competition (Salmela et al 1993). Speculative factors responsible for these differences include the greater time spent in practice which impacts on the nature of coach–athlete interaction, in addition to a general ‘hardening’ of the emotional climate during competition (Liukkonen et al 1996). Competition is also a time for predominantly reactive decision-making by coaches, as she or he must evaluate progress within the on-going event and act accordingly. It would appear that the nature of the competition experience and the pressures contained within it could well be an important factor that impacts on a coach’s actions; a time that not only involves physical game engagement, but also that period spent in immediate preparation for it, time-outs, half-time talks, between attempts or points, full-time briefings, and managing other stoppages where opportunities exist for coach–athlete interaction. The examination of practice time alone then would appear to exclude an important and by its very nature spontaneous and ‘creative’ part of the coach’s work (Gilbert et al 1999, Smith & Cushion 2006).
Despite the general volume of research carried out and an apparent ‘pattern’ of coach behaviour that emerges, when divided into context-specific studies, the scope for drawing meaningful conclusions from the work seems to be limited. However, this limitation does not seem to have inhibited the tendency towards proclaiming generalisations from the particular. Indeed, many conclusions about coach effectiveness have been drawn from only a few hours of observation of a very small number of coaches practising in a specific context (Horn, 1984 and Kahan, 1999). These conclusions have often been assumed to be valid across all contexts and applicable to all coaches. Similarly, research that has compared behaviours across a season has been based on a limited number of observed hours of practice, which inevitably casts doubt on the validity of any generalisations drawn from it (Potrac et al 2002). The research evidence therefore needs to be treated with considerable caution.
Indeed, it could be argued that the resulting recommendations and subsequent strategies for coaches have been ‘too absolutist for the complex coaching environment’ (Abraham & Collins 1998 p 65). This echoes the point made earlier in relation to generalisations drawn from the particular, and perhaps highlights a fundamental problem with coaching knowledge so far and its accompanying ‘models’ approach. Indeed, it is becoming increasingly recognised that coaching scholars have not taken the time to acknowledge adequately and empirically explore the complex nature of coaching, before developing generalist explanations of, and recommendations for, ‘good practice’ (Jones, 2006 and Cushion, 2008).
Part 2: How should coaches behave?
Given the discussion so far, what research-led behaviour guidelines can we give coaches in relation to their practice? This is a difficult question to address because, as highlighted, there are no definitive answers about how much of which behaviour a coach should use in any particular situation. Similarly, just because a ‘successful’ coach uses much of a specific behaviour or coaches in a particular ‘style’, does not mean that it will be either applicable or effective for another coach in a different context. Coaching prescriptions cannot be so simplistic.
The problem with the inability of the research to prescribe appropriate behaviour is partly rooted in the nature of behaviourist and reductionist study as it cannot by itself capture the complex and dynamic essence of coaching. Indeed, as already discussed most research only provides a very limited snapshot of coaches’ behaviours, while much of the interaction which typifies the wider process including what transpires prior to, during, and after practice sessions or games, goes unrecorded (Jones et al 1997). A further problem in this respect lies in the assumption that the methodologies used have the ability to differentiate good coaches from bad (Abraham & Collins 1998), which has been found largely not to be the case.
These inconsistencies have led Abraham and Collins (1998) to suggest that coach behaviour relates to the specific situation rather than to global rules. Moreover, attempting to apply the findings of behavioural analysis in prescriptive ways ignores the particular context under which studies have been conducted (Douge and Hastie, 1993 and Kahan, 1999); an important issue when considering the design and implementation of coach education. Importantly, however, certain behaviours do consistently permeate the findings of behavioural studies and provide a commonality of practice that cannot be ignored. Bearing this in mind, it is worthwhile considering some of the factors that can inform a considered approach to coach behaviour, and in doing so, to consider critically contemporary issues relating to this topic from a coach education perspective.
Coach behaviour: thinking philosophically
Although coaches may not necessarily articulate clear beliefs about it, their practice invariably rests upon basic often-unquestioned beliefs about learning (Light 2008). Indeed, all coaching is based upon some theory about how we learn. As has been discussed in this chapter so far, behaviourism strongly informs coaching (and its research) with a resulting instructional approach that emphasises the use of feedback and reward behaviour.
Epistemology refers to questions about the nature of knowledge and the relationship between the inquirer and the known (Sparkes, 1992, Denzin and Lincoln, 1994 and Denzin and Lincoln, 2000). Epistemological assumptions determine the issue of ‘whether knowledge is something which can be acquired on the one hand, or something which has to be personally experienced on the other’ (Burrell & Morgan 1979 p 2 cited in Sparkes 1992). Epistemologically, a behaviourist approach assumes that knowledge is objective and ‘out there’ and filtering or internalising objective knowledge requires a highly structured and technical pedagogical approach (Light 2008).
Recent developments in coaching research and coach education have tended to promote a more constructivist epistemology, where knowledge is assumed to be socially constructed in interaction, and has to be experienced rather than acquired (e.g. Cassidy et al., 2004 and Cushion, 2006). This approach emphasises the coach’s facilitative behaviour, not instructing per se but constructing experiences for athletes, an example being Teaching Games for Understanding (Bunker & Thorpe 1982).
It is not the intention here to square the philosophical circle, or indeed promote one approach as superior to another. Instead, as a starting point to a discussion around recommendations/ guidelines for coach behaviour, there is a need to emphasise that coaches should be aware of the assumptions about learning that underpin their behaviour and practice (Light 2008). This is crucial in raising coach self-awareness, a quality that the evidence suggests is lacking, but seems essential if coaches are to grasp the implications (good or bad) of their behaviour. More importantly perhaps, Butler (2005) identifies an ‘epistemological gap’ or ‘cognitive dissonance’ (Light 2008) where there is a difference between an embodied and unarticulated belief that informs behaviour and practice and an alternative set of assumptions, resulting in coaches struggling to adopt any alternative behaviour. This has manifested itself in coaches’ ‘impression managing’ on coach education courses and paying lip service to alternative behaviour (Cushion et al 2003) by adopting the language of a particular approach but continuing to coach and behave in ways that are informed by a traditional objectivist approach to learning. Coaches can develop better conceptual understanding by reflecting on why they coach as they do and what assumptions underpin this. This, alongside critical reflection on their socialisation experiences and the culturally accepted coaching behaviours of their sport, puts coaches in a strong position to transcend ‘traditional’ coach behaviour and develop their own informed approach.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
