Over the past 30 years, two researchers from the United States, Ronald Smith and Frank Smoll, with various colleagues, have conducted an impressive line of research on the effects of a cognitive-behavioural approach to coach training. Given that their research is unique and continues to evolve today, a brief overview of their intervention program and their findings is presented.

Initially, two studies were completed to establish a baseline of coaching behaviours and athlete attitudes and perceptions. The two studies were completed in the late 1970s and early 1980s with youth baseball (n= 51) and basketball (n = 31) male coaches of boy’s teams (n = 724 athletes) ranging in age from 8–15 years old (Smith and Smoll, 1990, Smith et al., 1978 and Smith et al., 1983). Coaches’ behaviours were coded during at least two games (range 2–5 games) using the Coaching Behavior Assessment System (CBAS), and in the first study the coaches also completed a coaching philosophy questionnaire. Athlete perceptions, attitudes, and self-esteem were assessed at the end of the season using two questionnaires, one of which eventually became known as the Washington Self-Description Questionnaire (WSDQ). Although the specific results varied between the two studies, the findings showed that coach behaviours strongly influence athlete attitudes toward their coach, as well as their perceptions of themselves, their coaches, and the actual sport experience. The results were most striking for young participants with low self-esteem, showing that this group of athletes responded most positively to coaches high in reinforcing and encouragement behaviours. Coaches scoring low on support and high on punishment were least liked by their athletes. The results also showed a large discrepancy between coaches’ perceptions of their behaviours and their actual behaviours. These studies laid the foundation for the creation of the Coach Effectiveness Training (CET) program. The CET is a brief (2.5 hours) workshop designed to promote principles of positive control, to help coaches conceptualise winning as giving maximum effort, and to nurture self-awareness and self-monitoring in coaches.

The effectiveness of the CET program has subsequently been tested in four separate studies. The first study was conducted with 31 baseball male coaches and 325 athletes who were 10–15 years of age (Smith et al 1979). A multi-method approach was used to provide the coach training. At the end of a 2-hour workshop, coaches were given a manual with behavioural guidelines and a personal behavioural profile based on observation (using the CBAS) of two games at the beginning of the season with norms for comparison. Coaches were also asked to complete self-monitoring forms after the first ten games of the season and reminder phone calls were made periodically. Athletes’ perceptions, attitudes, and self-esteem were assessed with questionnaires. Results indicated that compared to the untrained coaches, coaches in the experimental group provided greater amounts of reinforcement to their athletes. Athletes of the experimental group coaches evaluated both their coach and their team’s interpersonal climate more positively. Trained coaches were perceived to be more reinforcing, more encouraging, more technically instructive, and less punitive in response to mistakes. It was also found that the low-self-esteem athletes were the group of athletes who exhibited the greatest positive change in attitudes toward their coaches.

The results of the second intervention study are reported across three publications (Barnett et al., 1992, Smoll et al., 1993 and Smith et al., 1995). The intervention was tested with 18 male baseball coaches and 152 athletes 10–12 years of age. The researchers did not collect data on the coaches’ behaviours but asked the athletes to answer a questionnaire on their perceptions of the coaches’ behaviours, and on their attitudes toward the coaches and other aspects of participation. Athletes also completed three tests on self-esteem (WSDQ) and anxiety (Sport Anxiety Scale [SAS] and Sport Competition Anxiety Test [SCAT]) immediately before and after the season. Athletes of trained coaches perceived their coaches to engage more frequently in desirable behaviours, liked their coaches and teammates more, rated their coaches as better teachers, and had more fun than players of untrained coaches. In terms of anxiety and self-esteem, athletes of trained coaches showed a decrease in anxiety over the course of the season but no general self-esteem trend was evident. However, there were significant increases in self-esteem for the sub-sample of boys who started the season with low self-esteem. Lastly, only 5% of the boys who played for trained coaches did not return the following season, compared to 26% of the boys who played for untrained coaches.

Smith, Smoll and colleagues revised their coach training protocol in their most recent study (see Table 9.1, Smith et al., 2007 and Smoll et al., 2007). The CET program was modified and is now referred to as MAC (Mastery Approach to Coaching). It is briefer in duration (75 minutes) and the content is delivered through a lecture approach rather than a discussion. The MAC intervention is based upon two major themes: (a) positive coaching behaviours and (b) defining success as giving maximum effort. A 28-page researcher-designed coaching manual (Smoll & Smith 2005) and a self-monitoring form are still provided to the coaches. Using a quasi-experimental design, the MAC was tested with 37 community-based basketball coaches and 216 athletes (mean age = 11.5, SD = 1.63), which for the first time included girls (n = 99). The athletes completed four different measurement instruments (anxiety, motivational climate, achievement goal for sport, and academic achievement goal) at the beginning and at the end of the sport season (12 weeks later). Athletes who played for MAC-trained coaches perceived their coaches to be more mastery-oriented (as opposed to ego-oriented), increased their mastery goal orientation scores, lowered their ego orientation scores, and exhibited decreases in anxiety from preseason to late season. No changes were found for the control group athletes. These group differences were noted equally for boys and for girls.

To our knowledge, only one study has been completed on Smith and Smoll’s coach education interventions by someone other than the original authors. David Conroy and J. Douglas Coatsworth (Coatsworth and Conroy, 2006 and Conroy and Coatsworth, 2004) tested the effectiveness of the CET intervention with seven early-career swimming coaches (mean years of coaching experience = 2.43 years) and 135 swimmers (52 boys and 83 girls, mean age = 11.4, SD = 2.23, range = 7–18). They deliberately selected a sample very different from those used in the Smith and Smoll line of research in an attempt to explore the effectiveness of the CET intervention in diverse youth sport settings. Four coaches received the CET intervention and three coaches received a control training program consisting of 2 hours on injury prevention and emergency first aid. The intervention group coaches received a 2-hour workshop based on the CET, and were given a copy of Smoll and Smith’s coaching manual. Coaching behaviours were coded in two 1-hour practices for each coach using the CBAS and the athletes completed a self-esteem scale and a performance failure appraisal (Performance Failure Appraisal Inventory) three times across a 7-week swim season (beginning, middle, and end). The results, published in two separate articles, indicated no significant effects on the participants’ fear of failure, a general tendency to increase positive self-esteem over the season – particularly for girls, and only moderate or no change in coaching behaviours. This is the first published study showing very limited or no significant impact, either for coaches or their athletes, following participation in a CET-based workshop. These findings were partially attributed to several methodological limitations noted both by Conroy and Coatsworth themselves and by Smith and Smoll (Smith et al 2007). These limitations include a small sample size (only seven coaches), suitability of the fear of failure instrument for this age group, insufficient data points for the coaching behaviours, homogeneous sample for race and SES, and the relatively brief nature of the workshop. As Conroy and Coatsworth concluded, how realistic is it to expect ingrained coaching behaviours to change a few weeks after exposure to a 2-hour workshop?

Similar to the studies presented so far, Trudel et al (2000) developed an intervention strategy that focused on a specific coaching element. Their intervention strategy was developed especially for ice hockey coaches and applied to 28 coaches and their players (14–15 years old) in competitive leagues. The three-part intervention research started with a 2-hour individual meeting with each coach. Using specially made video recordings, the meeting sought to (a) make coaches aware of the problem of injuries and penalties, (b) demonstrate the importance of teaching body checking, (c) furnish teaching materials, and (d) present the concept of self-supervision and how to use this technique during the season. Coaches also had to present a video montage to their players explaining the importance of making legal body checks and the techniques involved. Finally, the coach had to teach body checking during at least four on-ice training sessions at the start of the season. The hypothesis was that if coaches applied the principles of self-supervision and used the teaching materials provided, they would be able to help players use body checks correctly, and fewer minor aggression penalties and injuries would be noted. When the researchers compared the data collected during that hockey season with the data of the previous season, no significant differences were found. To the question ‘how do we explain that the intervention strategy pleased coaches but did not produce the desired results?’ the authors stressed the fact that it is almost impossible to control all variables when research is conducted in the sporting milieu and no pressure, other than what was said during the workshop, is put on the coaches to be compliant. For example, some coaches admitted that they had not completely followed the intervention strategy, giving reasons such as the lack of time, and coach and player changes during the season. The authors concluded that, ‘considering that minor league hockey exists with the participation of many individuals (organizers, officials, parents, coaches, players) it is probably unrealistic to believe that a short term intervention implicating only the coach, will suffice to reduce violence’ (p 246).

The study by Cassidy et al (2006) with eight rugby coaches in a developmental sport context is very different from the studies presented so far because the researchers did not try to measure the effectiveness of a behavioural approach. Using a qualitative methodology, they examined the effectiveness of a theory-based coach education program (CoDe) that can be qualified ‘as a boutique, community-orientated, short-term (28 hours over 6 months), classroom based, theoretical, educational/personal development coaching program with no assessment component that was offered free of charge to the volunteer coaches’ (p 148). Once the program was completed, in-depth semi-structured interviews were conducted with each coach to collect data on their perceptions of participating in the CoDe. Results indicated that coaches appreciated the fact that, contrary to training courses they had attended in the past that focused on technical sport components, CoDe helped them to see the complexities associated with the coaching process and critically reflect upon their own approach. This was achieved mainly through formalised opportunities where the coach educator facilitated discussion, interaction, and negotiation of meaning among the coaches instead of lecturing and prescribing coaching and theoretical principles.

From this brief overview of studies published on small-scale coach education training programs it is important to highlight a number of issues. First, it is extremely difficult to determine the effect of a coach education training program even when the researchers (a) have full control of the intervention content, (b) can select a trained researcher as facilitator to conduct the workshop, (c) decide the sport and competitive context, and (d) separate coaches into control and experimental conditions. For example, Trudel et al (2000) found no significant differences even if coaches said they appreciated the intervention strategy. Second, studies such as those conducted by Smith and Smoll can sometimes show significant differences in favour of the trained coaches but are these differences large enough to be meaningful, and can these results be generalised to all coaches and athletes? For example, on 12 behaviours measured by the CBAS, trained coaches scored better than the untrained coaches only on the reinforcement behaviour (Smith et al 1979). For athletes the biggest difference reported on their perception of the sport context was just over one-half of a point on a 7-point scale (trained coaches M = 5.59, untrained coaches M = 4.95) for the question, ‘do you like your manager/head coach more or less than you did at the beginning of the season? (Smith et al 1995 p 134). For the Mastery climate, trained coaches scored on average 26.23 (SE = 0.36) and the untrained coaches 25.08 (SE = 0.48) on a maximum of 30 (Smith et al p 50). We also do not have any data on the long-term impact of these training programs. Will the few, relatively small, changes found still be evident in following years when coaches do not have self-monitoring forms to complete, do not receive reminder phone calls, and their athletes are not assessed on the research variables? Third, studies on small-scale coach education training programs have provided information on only a very limited segment of youth sport; generally competitive team sports.

Our conclusion is that, at the moment, there is no substantial body of evidence to support the widespread or long-term effectiveness of coach education training programs, even in highly controlled and small-scale quasi-experimental settings. After nearly three decades of sophisticated research on the CET/MAC, Smoll et al (2007) concluded that, because of the lack of a true experimental design, or comparison to alternative coach education interventions, ‘we cannot rule out the possibility that simply receiving an intervention (regardless of its content or nature) helped change the coaches’ behaviour’ (p 40). We might also wonder if we should strive to find the ‘best’ model (i.e., best practices). The search for such a model may be futile as ‘coach training programs are clearly not equally effective in all situations’ (Conroy & Coatsworth 2004 p 211).

Contrary to the studies presented in Table 9.1, studies on university-based coach education programs (a) do not focus on a specific sport, (b) the duration of the intervention is longer, (c) there are fewer participants, and (d) the focus (Cassidy et al 2006 being the exception) is not to change certain specific coaching behaviours but, to use Jones and Turner’s (2006) expression, to ‘help coaches acquire the “quality of mind” necessary to deal with, and excel at, the dynamic nature of their work’ (p 185).

In the first example, Demers et al (2006) discuss the development of their university undergraduate coach education program designed to develop reflective practitioners. For them, there is a ‘need for a professional education program that could reach beyond the many limitations that a community-based program must accommodate, such as cost, time involved in training, as well as assessment requirement for coaches to demonstrate competency’ (p 164). The researchers reported that to develop and implement their curriculum based on competency-based and problem-based learning (PBL) approaches, faculty members have to work together and time might become a barrier. It was also stressed that students might have difficulties transferring the knowledge presented in courses to their practice (internship). It was suggested that one way to help students link the concepts to their coaching practice is to create specific assignments that require students to complete critical reflection reports.

For Jones and Turner (2006) coaching must be ‘intellectualised’. This requires ‘developing theory related to its complexity, while appreciating the need to grow the aforementioned “quality of mind” in coaches through habits of reflection, problem-solving, and critique; elements that are integral to PBL’ (p 183). These researchers introduced the principles of PBL during the final year of a Bachelor of Arts degree (Coach Education and Sport Development). Results showed that students had begun to think differently about coaching but the process itself was extremely challenging. For example, tutors need to be well trained in order to provide a balance between allowing student discussion time and intervening to make sure that important learning issues are raised. Students not familiar with a PBL approach might be somewhat antagonistic to this approach and therefore will need additional support and clear information and expectations. The problematic scenarios have to be selected carefully and support from a resource technician (e.g., a librarian) is important to help direct students to the relevant information. Finally, in a PBL approach, participants often must work in small groups, and therefore finding enough time to meet could become a significant barrier to consider.

Two studies have reported on the impact of a program to develop and assess reflective skills among students of a Bachelor of Science (Honours) coaching science degree (Knowles et al., 2001 and Knowles et al., 2006). During the first semester of their second academic year, students (coaches) attended theoretical lectures that focused on conceptual and practical issues associated with reflective practice. In the second semester, students were engaged in 60-hour coaching placements and also had to participate in five workshop sessions to discuss coaching topics, keep a reflective journal, and complete an academic year (annual) report. Results showed that, for some coaches, the workshop sessions allowed collective discussion and the generation of action plans while other coaches felt the sessions were not useful, as they preferred to work alone. It was also found that students needed extra support during the early stages of their placement, and the role of the workshop facilitator is complex and multi-faceted. For example, the facilitator often had to conduct one-on-one sessions with the participants to address fully certain coaching issues. This highlights the potential need for workshop facilitators to not only have sport science, pedagogy, and reflection skills, but also effective interpersonal skills. The coaching placements differed in terms of their potential to impact the development of reflective practice. Assessment of the level of reflection is a daunting task because students typically will vary in their writing skills. The authors concluded, ‘coach educators cannot therefore assume that development of reflective skills will be a naturally occurring phenomenon that runs parallel to increasing coaching experience’ (p 204). In an attempt to better assess the effectiveness of their program Knowles and colleagues later studied how graduates of this coaching science degree deployed reflective processes in their coaching practice. Using interviews with six graduates, no evidence was found of reflecting at a critical or a practical level. When coaches did report reflection, they tended to reflect primarily on coaching problems. Although the coaches acknowledged that written reflection was important in the reflective process, none of them made time for reflective writing. Reflection was limited to mental notes and peer discussion. One possible reason for these findings is that in the field, coaches often work in isolation. This situation is in stark contrast to what they experienced when they were undergraduate students working in an environment with skilled tutors and reflective workshops. In their conclusion, the authors made reference to the difference between the academic experience and actual coaching practice: ‘In summary, the in-built reflective rigour present in the undergraduate program is at variance with the post-graduation reality of sports coach employment’ (p 176).

These four studies share the common objective of developing reflective coaches who will increase their capacity for creating effective solutions to their coaching issues. As shown by the results of these studies, it is not easy to teach or assess the reflective process. Participants were university students involved full time in coaching programs delivered by accredited institutions. Even under these intensive and controlled conditions, very little evidence of growth in coach reflection skills was noted, and even less evidence was found for use of reflection post-graduation. There seem to be challenges both for those who deliver the program and for the participants. In a program aimed at preparing reflective practitioners (coaches) the instructor has to be well prepared to use non-traditional approaches like problem-based learning. For the student-coaches the challenge might be to adapt to a new teaching/learning approach in which the evaluation criteria are less specific and require them to share orally or in writing their thoughts about coaching issues and possible solutions. These four studies, when viewed in conjunction with the Knowles et al (2005) review of national governing bodies’ coach education program inclusion of reflective skills, paint a dismal portrait of coach education’s ability to develop reflective practitioners. Knowles et al (2005) found that none of the programs directly taught or nurtured reflection in their curriculum. Clearly much work is required if reflection is expected to become a primary characteristic of coaches who complete episodes of coach education. We think it is important to also note that very few university-based coach education programs focus on developing reflective coaches. A review of university-based coach education programs in the United States shows a very traditional, and uniform, curriculum (McMillin & Reffner 1999). All programs include a mix of coaching theory, sport science sub-disciplines (i.e., biomechanics, sport psychology) and some form of coaching practicum. This trend is also evident in a review of online coach education curriculum descriptions at universities that have established coach education programs (e.g., http://www.bu.edu/online/online_programs/certificate_programs/physical_education.html, www.cui.edu/AcademicPrograms, http://www.georgiasouthernhealthscience.com, http://www.teambath.com and http://www.essex.ac.uk/sport/coach/index.shtm). Unfortunately, no studies on the effectiveness of these types of traditional programs were found while preparing this chapter.