Research Analysis paper

JOURNAL OF APPLIED SPORT PSYCHOLOGY, 22: 106–121, 2010
Copyright C© Association for Applied Sport
ISSN: 1041-3200 print / 1533-1571 online
DOI: 10.1080/10413200903474472

Relationship Maintenance Strategies in the Coach-Athlete
Relationship: The Development of the COMPASS Model

DANIEL J. A. RHIND

Brunel University

SOPHIA JOWETT

Loughborough University

The investigation of relationship maintenance strategies has received considerable attention in
various types of dyads including romantic, marital, and familial relationships. No research,
however, has yet investigated the use of maintenance strategies in the coach-athlete partnership.
Thus, this study aimed to investigate coaches’ and athletes’ perceptions of the strategies they
use to maintain relationship quality. Twelve one-to-one interviews with coaches (4 males and
2 females) and athletes (2 males and 4 females) were conducted. The interviews were struc-
tured based on the factors within Jowett’s (2007) 3+1C conceptualization of the coach-athlete
relationship (i.e., closeness, commitment, complementarity, and co-orientation). Deductive
and inductive content analysis revealed seven main categories: conflict management, open-
ness, motivation, positivity, advice, support, and social networks. The COMPASS model was
developed based on this analysis and was offered as a theoretical framework for understanding
how coaches and athletes might maintain the quality of their relationships.

Significant strides have been made in our understanding of the nature and role of inter-
personal relationships in sport over the past decade. Research in this domain has considered
athlete-athlete, parent-athlete, and coach-athlete partnerships (e.g., Jowett & Chaundy, 2004;
Ullrich-French & Smith, 2006). Recently, Jowett and Wylleman (2006), in considering re-
search on the coach-athlete relationship, suggested that “ . . . we have started crossing the
chasm and started approaching an exciting research territory that needs exploration with care-
ful navigation” (p. 123). Research progress in this field employs numerous theoretical and
methodological approaches (see e.g., Chelladurai & Riemer, 1998; Jowett & Poczwardowski,
2007 for overviews). Research indicated that relationship quality was positively associated with
outcomes such as self-concept (Jowett, in press), satisfaction (Jowett & Ntoumanis, 2004),
passion for one’s sport (Lafreniere, Jowett, Vallerand, Donahue, & Lorimer, 2008), and team
cohesion (Jowett & Chaundy, 2004). Nonetheless, no research has focused on how coaches
and athletes maintain the quality of their athletic relationships.

Received 8 April 2009; accepted 8 September 2009.
Address correspondence to Daniel J. A. Rhind, Centre for Youth Sport and Athlete Welfare, Brunel

University, Kingston Lane, Uxbridge, Middlesex, UB8 3PH. E-mail: [email protected]

106

MAINTENANCE OF THE COACH-ATHLETE RELATIONSHIP 107

Commenting on all dyads, Canary and Stafford (1994) stated that “most people desire
long-term, stable, and satisfying relationships” (p. 4). Moreover, it has been argued that unless
people use effective maintenance strategies, their relationships will weaken and ultimately
end (Canary & Stafford, 1994). A great deal of time must be invested in personal relation-
ships to maintain them (Duck, 1986). In the context of sport, coaches and athletes establish
and maintain their athletic relationships motivated by such wide-ranging goals as improving
performance, achieving success, maintaining fitness, or simply enjoying participation. Clearly
such partnerships occur in a range of situations (e.g., different competitive levels and types of
sport) and are subject to organizational constraints (e.g., organizational culture, goals, fund-
ing pressures). Thus, maintenance of the coach-athlete relationship is not simple and often
necessitates conscious effort from both parties.

The coach-athlete relationship has been defined as “ . . . the situation in which coaches’
and athletes’ emotions, thoughts, and behaviors are mutually and causally inter-connected”
(Jowett & Ntoumanis, 2004, p. 245). This definition highlights the bi-directional nature of
such relationships in that the feelings, thoughts, and behaviors of the coach are both affected
by, and also affect, those of the athlete and vice versa. This definition also supports the belief
that relationship quality is multi-dimensional and hence one needs to consider the affective
(emotions), cognitive (thoughts), and behavioral interpersonal aspects of relationships (e.g.,
Kelley et al., 1983).

Jowett (2005, 2007) developed the 3+1Cs conceptualization of the coach-athlete relation-
ship based on a series of qualitative studies (e.g., Jowett & Cockerill, 2003; Jowett & Meek,
2000). This conceptualization refers to four constructs: closeness, commitment, complemen-
tarity, and co-orientation. This model argues that the quality of the relationship between a
coach and an athlete is formed through these four key constructs. Closeness was defined as the
affective meanings that the coach and athlete assign to their relationship (e.g., respect, trust,
liking). The cognitive aspect, operationalized as commitment, relates to the members’ inten-
tions to maintain the relationship now and in the future. The behavioral aspect, operationalized
as complementarity, refers to the relationship members’ co-operative and corresponding be-
haviors of affiliation (e.g., being responsive and friendly). Finally, the “+1” element of this
conceptualization was co-orientation, falling under the cognitive construct because it is per-
ceptual in nature (Laing, Phillipson, & Lee, 1966). It was labeled the “+1” element because
it runs through each of the other affective, cognitive, and behavioral elements. Co-orientation
concerns the degree to which an athlete and coach are able to accurately infer how his/her
coach/athlete is feeling, thinking, and behaving.

This body of research shed light on different factors that both affect, and are affected by, the
coach-athlete relationship. It also developed our understanding of the nature and content of
this important dyadic relationship within the realm of sport development. Nevertheless, there
remains a great scope for research in this field, as displayed by Jowett and Poczwardowski’s
(2007) research model. This model calls for research regarding the role played by interpersonal
communication, because communication is viewed as a process from which coaches and
athletes can either become close (united) or distant (divided). Thus, the argument is that
relationship maintenance strategies are embedded within the interpersonal communication
“layer” of Jowett and Poczwardowski’s (2007) research model.

Dindia and Canary (1993) described relationship maintenance as the strategies used to
keep a relationship in a specified state or condition. Some examples of the use of relationship
maintenance may include discussing an area of disagreement and coming to a joint decision
of how it can be resolved (i.e., conflict management) or going out together for the evening
(i.e., socializing). Wiegel and Ballard-Reisch (1999) suggested that maintenance strategies are
the primary method via which people within close relationships maintain relationship quality.

108 D. J. A. RHIND AND S. JOWETT

The initial studies of relationship maintenance centered on dating or married couples (e.g.,
Dainton & Stafford, 1993; Stafford & Canary, 1991). More recently the principles and concepts
of relationship maintenance have been applied to parent-child relationships (e.g., Punyanunt-
Carter, 2006) and friendships (Bippus, & Rollin, 2003). As yet, however, no research has
addressed this topic with reference to the coach-athlete relationship.

In a seminal paper within the relationship maintenance literature, Stafford and Canary
(1991) identified, based on factor analysis, five primary relationship maintenance strategies
that were used by dating or engaged couples: positivity, openness, assurance, social networks,
and shared tasks. Positivity concerned acting cheerful and upbeat when around one’s partner.
Openness related to the direct discussion of the relationship and disclosing what one wants
from it. Assurance referred to the sending of messages that imply one’s commitment to the
relationship. Social networks involved spending time together and interacting with mutual
friends. Finally, shared tasks related to the partners performing assigned chores around the
house. Canary and Zelley (2000) added two additional relationship maintenance categories:
Conflict management (i.e., co-operating when discussing disagreements) and advice (i.e.,
giving one’s opinions regarding the partner’s problems).

Canary and Stafford (1992) suggested that, in general, the use of maintenance strategies
also “affects the nature of the relationship” (p. 9). Over the years, research studies found
positive associations with relationship properties such as liking (Canary & Stafford, 1992),
trust (Stafford, Dainton, & Hass, 2000), affinity seeking (Bell, Daly, & Gonzalez, 1987),
control mutuality (Canary & Stafford, 1992), love (Stafford, et al. 2000), and commitment
(Canary & Stafford, 1992; Stafford et al., 2000). Research also has indicated that maintenance
strategies need to be continually performed because their absence could lead to a rapid decline
in the quality of a relationship (Canary, Stafford, & Semic, 2002).

Overall, it was believed that these types of relationship maintenance acts were used to ensure
the continuation of valued relationships through three distinct routes: (a) the prevention of their
decline, (b) their further enhancement, and/or (c) their repair and re-establishment (Canary &
Stafford, 1994). Canary and Stafford (1994) argued that the use of these strategies could lead
to a number of positive relationship-related outcomes including intimacy, commitment, and
satisfaction.

Although no sport psychology research has directly considered relationship maintenance
within the coach-athlete relationship, some research appeared to address issues related to
maintenance strategies. For example, Gould, Lauer, Collins, and Chung (2007) interviewed
10 American football coaches who all received awards for their abilities to facilitate their
athletes’ personal development. In the interviews, these coaches emphasized the importance
of communication (i.e., having open lines of communication with their athletes, possessing
clear expectations, and holding their players accountable). These coaches also avoided using
punishment or criticisms that were directed toward their players’ characters or personalities,
and showed that they cared, trusted, and respected their players as people. These enacted
communicative acts paralleled the relationship maintenance strategies labeled as positivity,
openness, and assurance (Stafford & Canary, 1991). Moreover, research focusing on coaches’
behaviors consistently has illustrated that supportive and encouraging coaches were likely to
have a positive influence on their athletes’ development (Coatsworth & Conroy, 2006). This
supporting coaching was particularly effective when their athletes were less confident about
themselves (Smith & Smoll, 1990). Thus, the use of maintenance strategies in sport has been
indirectly associated with positive outcomes.

Although this limited body of research indirectly addressed some of the processes leading
to satisfying coach-athlete interactions, it did not directly examine relationship maintenance
strategies. Thus, the aim of this study was to assess coaches’ and athletes’ perceptions of the

MAINTENANCE OF THE COACH-ATHLETE RELATIONSHIP 109

strategies that might be used to maintain the quality of the coach-athlete relationship. Even
though it was acknowledged that all relationships are unique, the present research aimed to
identify strategies that were salient across different relationships.

The need for the present study could be justified on three fronts. First, because the quality
of a coach-athlete relationship has been associated with a range of positive outcomes, there
is a demand to understand how relationship quality could be optimized. Second, relationship
maintenance is a significant area of research regarding other relationships but it, as yet, has not
been studied within sport. Third, coach-athlete relationships are significantly different to the
types of romantic, marital, and familial relationships studied thus far. Clearly, these alternative
dyads are linked biologically or legally. In contrast, a coach and an athlete are generally linked
contractually, professionally, or voluntarily with a view to achieving set goals (e.g., fitness,
financial rewards, enjoyment). Professional relationships, such as that between a coach and an
athlete, are fundamentally different to romantic, marital, and familial relationships and hence,
they may be maintained using alternative maintenance strategies. As a result, research was
merited to specifically investigate how coach-athlete relationships were maintained.

METHOD

Participants

A purposive sample (N = 12) consisted of 6 coaches and 6 athletes who worked inde-
pendently and hence, were not existing coach-athlete dyads. Efforts were made to recruit
participants who had experienced a range of different coach-athlete relationships. To maintain
the anonymity of the participants, we identified them as Athlete 1–6 (A1, A2 . . .) and Coach
1–6 (C1, C2 . . .) in the results section.

A1 (25 year-old female) was an international ice skater who had been training since she
was 12 years old. She worked with a single coach throughout her career, but this relationship
ended after a period of significant conflict. A1 subsequently continued for a short time without
a coach before retiring from the sport. A2 (25 year-old male) was an international discus
thrower and had been involved with the sport for 13 years. He worked with his first coach
between age 13 and 18 years old before changing coaches when he moved away to university.
A2 competed at many international youth sport events and currently trained with his coach for
10 hours each week.

A3 (24 year-old female) competed as a rower at the national level and had been rowing
for 8 years. She had worked with her current coach for 1 year, training for around 10 hours
each week. She was also successful at the elite university level. A4 (23 year-old female) had
been playing women’s football (soccer) for 7 years. She played for an elite university football
team. A4 had been working with her coach for 2 years, training for approximately 4 hours
each week.

A5 (18 year-old female) represented her university at field hockey and had been playing
the sport for around 5 years. She trained for 6 hours per week, and had two main coach-
athlete relationships with a coach at home and one at university. A6 (22 year-old male) was a
successful karate player and won medals at the national level. He had worked with his coach
for 20 hours each week over the previous year.

C1 (25 year-old male) coached at the county level (i.e., regional representative teams). He
generally worked with children under 16 years old and had been coaching for 3 years. He
trained 3 hours per week. C2 (32 year-old male) was a football (soccer) coach with a Union of
European Football Associations (UEFA) B license, which is a level of coaching certification. He
coached at the national level, both in England and in Spain. He had been coaching for 10 years.

110 D. J. A. RHIND AND S. JOWETT

C3 (24 year-old male) coached the university archery team. He had been involved in archery
for 4 years. He had only been coaching for the past year and mainly worked with beginners
for around 2 hours each week. C4 (27 year-old female) coached the university 1st football
(soccer) team. The team was successful at the university level. She used to play for the team
before injury ended her career. She then obtained her UEFA B license, which is a level of
coaching certification, before training the team for 10 hours per week over the past 3 years.

C5 (23 year-old female) used to compete in trampolining at the national level. She had
been coaching at the elite university level for 2 years, training for around 3 hours per week.
C6 (57 year-old male) coached squash at the county level (i.e., regional representative teams)
for around 30 years. He coached a range of players from the beginner level through to county
players.

Instrumentation

One-to-one interviews were conducted to gain in-depth data about the strategies used to
maintain the quality of the coach-athlete relationship. An interview schedule was developed
based on relevant theory and literature (e.g., Jowett, 2007; Jowett & Cockerill, 2003). The
original interview schedule was piloted with an athlete and a coach. These interviews were
conducted to gain feedback on the content of the interview and to check its approximate
length and duration. As a result of the pilot study, minor modifications were made to enhance
clarity, coherence, and discourse. The final interview schedule contained a total of 10 open-
ended questions and was divided into 3 sections. (The interview schedule is available from the
authors.)

The first section outlined the purpose of the study before covering demographic information
(e.g., the participant’s main sport, length of participation, and competitive level). The second
section was the main part of the interview and contained standardized questions that addressed
various aspects of interpersonal interactions. Specifically, the focus of this section was on
collecting responses related to what the participants considered to be important in a good
coach-athlete relationship, as well as how they thought such a relationship was developed and
maintained. Participants were also presented with simple definitions of the 3+1Cs and were
invited to discuss the time/s when they experienced each of these relationship properties and
to consider efforts made to maintain the quality of the relationship (e.g., Think of a time when
you were committed to your coach/athlete. What do you feel helped to develop and maintain
this?). Prompts and follow-up questions were used as necessary to facilitate and encourage the
flow of information. The third and final section of the interview asked participants to make
any additional comments, clarify, or expand on any comments made during the interview
(i.e., Are there any other factors which you now think help the development and maintenance
of a good coach-athlete relationship, which we have not already discussed?). All interviews
were conducted by the first author. All interviews were audio-taped with the permission of the
participants and were transcribed verbatim immediately after each interview.

Procedure

Sport clubs within or close to the investigators’ university were e-mailed to inform coaches
and athletes about the purpose of the study and to invite them to participate. Shortly after
the e-mail was sent, a phone call was made to the head coach at the club to identify any
potential participants. If there was interest in participating in the study, a convenient time
and location was arranged for the interview. In an attempt to maximize the diversity and
breadth of the obtained qualitative data, a conscious effort was made to recruit independent
coaches and athletes (i.e., no dyads) from a range of sports and competitive levels. Each

MAINTENANCE OF THE COACH-ATHLETE RELATIONSHIP 111

interview was conducted in a private room with only the interviewer and interviewee present.
All participants were informed that the interview was anonymous and confidential and that
they were free to end the interview at any time. Prior to any data being collected, approval for
the study was obtained from the university’s ethical advisory committee. Interviewees were e-
mailed a copy of their transcribed interviews so they could check and verify the accuracy of the
transcription.

Data Analysis

The use of content analysis after an interview is the most common qualitative approach
within sport psychology research (Côté, Salmela, Baria, & Russell, 1993). Content analysis
is a process by which large amounts of qualitative data are organized through coding the
information into categories that concern similar themes. It was deemed to be the most ap-
propriate approach in the present research because it would allow the identification of salient
themes across participants’ responses. It also helped to highlight potential strategies that might
be relevant to all coach-athlete relationships. Both deductive and inductive forms of content
analysis were employed in this study.

The analytical process commenced by reading all 143 single-spaced pages of the transcribed
data to increase our understanding of the information that was obtained from the participants.
Subsequently, a review panel (comprised of the authors and an independent researcher) ana-
lyzed the data on the basis that a raw data unit represented a “quote” (i.e., a complete sentence/s
that referred to a distinct strategy and made sense as a stand-alone unit). A total of 401 meaning
units were highlighted. In the first stage of the analysis, quotes were categorized deductively
into one of the seven maintenance strategies highlighted within the literature review: conflict
management, openness, advice, positivity, assurance, shared tasks, and social networks. Any
relevant quotes that did not represent these categories were placed in an “other” category.

In the second stage, the quotes within the ‘other’ category were then inductively analyzed
to underline new potential maintenance strategies specific to the coach-athlete relationship.
Finally, the quotes within the existing and new categories were reviewed to identify lower and
higher-order themes.

The review panel then independently reviewed the initial categorizations to confirm that
all of the quotes were correctly classified. The panel went through each quote to confirm the
agreement on the classification. In particular, quotes were reviewed in terms of whether they
better reflected the newly created motivation category. Thus, some of the quotes were moved
from the positivity category to the motivation category at this stage. All areas of discrepancy
were resolved by dialogue and re-assignment.

As a final check, once the review panel was satisfied with all the classifications, the
categories and sub-categories were given to an independent psychology researcher who was
asked to assign each sub-category (lower order theme, specific) to the most appropriate category
(higher order theme, general). Then, the same researcher categorized the quotes into the most
appropriate lower-order theme. The responses supplied were 90% in agreement with the
authors’ categorizations, demonstrating some evidence of inter-rater reliability.

RESULTS AND DISCUSSION

Seven higher-order themes emerged from the data: Conflict management, openness, mo-
tivation, positivity, advice, support, and social networks. These seven dimensions contained
21 sub-categories. The results are presented using a frequency table and illustrative quotes
as recommended by Culver, Gilbert, and Trudel (2003). Overall, each of the dimensions was

112 D. J. A. RHIND AND S. JOWETT

referenced by the majority of the participants. The largest discrepancies between the coaches
and the athletes were found regarding conflict management (which was mentioned by all
coaches, but only by four of the athletes). Interestingly, both athletes who did not mention
conflict management were from team sports (i.e., field hockey and rowing). Similarly, support
was discussed by all athletes (but only by four of the coaches). Both coaches who did not
mention support were involved at the county level indicating that competitive level might
influence the use of maintenance strategies.

Table 1 summarizes the categorization of raw data points into the various dimensions and
themes. Results were broken down to show the distribution of the coaches’ responses, the
athletes’ responses, and the sample as a whole. This information is presented to give an
overview of the data and to outline the distribution of the quotes, rather than to provide a basis
for any comparative or statistical analysis.

Conflict Management

The first dimension that emerged from the data concerned conflict management, which was
cited by 10 of the participants (6 coaches and 4 athletes) with 9.5% of the raw data points

Table 1
Frequency of the Main Categories and Subcategories of Maintenance Strategies in the

Coach-Athlete Relationship

Frequencies

Coaches Athletes Total
Categories
Sub-categories N % N % N %

Conflict Management 26 6.5 12 3 38 9.5
Proactive 11 2.75 6 1.5 17 4.25
Reactive 15 3.75 6 1.5 21 5.25

Openness 20 5 28 7 48 12
Non-sport communication 9 2.25 9 2.25 18 4.5
Talk about anything 6 1.5 9 2.25 15 3.75
Other awareness 5 1.25 10 2.5 15 3.75

Motivation 63 15.75 64 16 127 31.75
Effort 16 4 14 3.5 30 7.5
Motivate other 13 3.25 15 3.75 28 7
Fun 9 2.25 5 1.25 14 3.5
Show ability 25 6.25 30 7.5 55 13.75

Positivity 22 5.5 18 4.5 40 10
Adaptability 6 1.5 7 1.75 13 3.25
Fairness 7 1.75 1 0.25 8 2
External Pressures 9 2.25 10 2.5 19 4.75

Advice 32 8 34 8.5 66 16.5
Sport communication 10 2.5 5 1.25 15 3.75
Reward feedback 10 2.5 6 1.5 16 4
Constructive feedback 12 3 23 5.75 35 8.75

Support 12 3 24 6 36 9
Assurance 7 1.75 6 1.5 13 3.25
Sport support 5 1.25 14 3.5 19 4.75
Personal support 0 0 4 1 4 1

Social Networks 25 6.25 21 5.25 46 11.5
Socializing 21 5.25 17 4.25 38 9.5
Shared networks 4 1 4 1 8 2

MAINTENANCE OF THE COACH-ATHLETE RELATIONSHIP 113

being categorized in this dimension (6.5% coaches and 3% athletes). Conflict management
reflected expectations, consequences of unmet expectations, and cooperation in the discussion
of conflict. It contained the themes of proactive strategies (e.g., taking steps to clarify expec-
tations and avoid conflict) and reactive strategies (e.g., co-operating during the discussion of
disagreements). The following quote is an example of proactive conflict management. It shows
how this particular athlete continually discussed expectations with his coach, helping him to
avoid the development of any conflicts that could arise from expectations not being met:

You can discuss at the beginning of the relationship what you both expect from one another
and have an understanding, a consensus, between the two from the start of the relationship,
but I think that it is important during the relationship as years go on to reassess that (A2: Male
international track and field athlete).

This dimension was similar to that highlighted within close relationships (Canary & Zelley,
2000; Stafford et al., 2000). Its conceptualization, however, has been expanded to include
not only co-operative acts during disagreements, but also pre-emptive strategies such as
clarifying the expectations and the consequences when these are not met (e.g., fines for being
consistently late for training). The inclusion of these pre-emptive strategies might highlight the
distinctive elements of the coach-athlete relationship relative to other dyads. Previous research
emphasized the importance of discussing expectations (e.g., Gould et al., 2007). It might be
that sporting dyads are used to setting goals and planning for future events and hence, these
pre-emptive strategies play a particularly significant role within the sporting arena.

Openness

Openness was mentioned by all 12 participants and 12% of the raw data units were catego-
rized within this dimension (5% coaches and 7% athletes). Openness related to the disclosure
of one’s feelings. It contained three themes: Non-sport communication (e.g., discussing issues
that are not directly related to training or competition), talk about anything (e.g., making it
clear that the coach/athlete can talk about any topic with you), and other awareness (e.g.,
making an attempt to understand how the coach/athlete is feeling). The following quote was
an example of the use of the ‘talk-about-anything’ form of openness:

You are confident enough to have two-way communication and feel assured that if you bring
something up outside of sport, then it is not just going to be battered away and it is not going
to affect anything. It is just positive from both sides (A4: Female university football player).

A coach focused on the importance of the non-sport communication form of openness
within the coach-athlete relationship and showed the way it could help to maintain the quality
of the relationship by saying, “Talking a lot, but not just talking about your sport and your
coaching but sometimes taking it further than just being coach and athlete, that can build up a
better relationship” (C5: Female university trampolining coach).

This dimension directly related to the openness category within Stafford and Canary’s
(1991) model of relationship maintenance. Openness strategies within the coach-athlete re-
lationship involved the discussion of topics outside of the sporting environment (e.g., work
and family life). Communication related to training or competition was categorized within the
advice dimension discussed below.

114 D. J. A. RHIND AND S. JOWETT

Motivation

The motivation dimension was discussed by all 12 participants. With 31.75% of the raw
data …