Jannik Zimmermann ✉ (University of Kassel) Silja Kotte (University of Applied Sciences Aschaffenburg) Isabell Diermann Heidi Möller (University of Kassel)
This study investigates the supervision of workplace coaches and factors that may influence its use. Drawing on the theory of planned behaviour and focusing on German-speaking countries, we conducted an online survey with open and closed questions to gather the experiences and perspectives of 121 coaching practitioners. Most coaches used supervision, though somewhat irregularly, in different forms and with different focuses, and less than what they thought ideal. A positive attitude impacted positively upon supervision use. Perceived time and cost involved, and prior experiences with supervision also appear to influence supervision use, while coaching association membership had no impact.
Coaching supervision
Accepted for publication: 11 July 2023 Published online: 01 August 2023
© the Author(s) Published by Oxford Brookes University
Supervision has become an established practice for supporting coaches in their work, developing their skills, and increasing the professionalism of coaching, especially in recent years (Bachkirova, Jackson, Henning & Moral, 2020). Experts and practitioners widely agree that supervision has many benefits (e.g., Clutterbuck, Whitaker & Lucas, 2016; Hawkins, Turner & Passmore, 2019). Their assumptions are based on theoretical explanations, practical experience and initial empirical evidence (Bachkirova et al., 2020; Birch & Welch, 2019; Hawkins & Smith, 2006). Consequently, coaching associations, training institutes and commissioning organisations require the use of supervision as a central element of coaches’ continuous professional development beyond the initial coach training (e.g., De Haan & Brich, 2010; Gannon & Myers, 2018; ICF, n.d.; Passmore & Sinclair, 2020). It seems to be common sense that supervision of workplace coaching is essential for good practice.
But what about in practice? Although many coaches agree that supervision can be beneficial to their work and are aware of the external demands for using it, they seem reluctant to act accordingly. This discrepancy was already evident in one of the earliest studies on supervision of coaches: 86% of coaches surveyed agreed that regular supervision should take place, but only 44% stated that they practised it (Hawkins & Schwenk, 2006a). Later studies found similar discrepancies (Jepson, 2016; Salter, 2008). The question remains: How to narrow the gap between attitude and behaviour and promote the use of supervision? More detailed knowledge concerning coaches’ supervision practice and the factors that impact upon it could provide significant insights that shed light on this question as well as foster the professionalisation of coaching in the longer term. Such knowledge can also provide practitioners and supervision providers with helpful guidance, support and stimulation for their own practice, enabling them to reflect critically on how they approach supervision.
The aim of this article is therefore to take a closer look at the supervision practice of coaches and to gain further insights into the factors that may influence its use. We present a brief review of the literature, a description of the methodology and the results of our study. We then discuss its limitations as well as its contributions and practical implications.
In this section, we provide a brief overview of relevant aspects of the current state of the literature concerning the supervision of coaches. We conclude with an outline of the key assumptions of the Theory of Planned Behaviour, using it as a lens through which to view behaviour (i.e., supervision practice) and its antecedents.
There are several different definitions for supervision of coaches (Tkach & DiGiralmo, 2017). Bachkirova (2008) defines it as a
process of professional support, which ensures continuing development of the coach and effectiveness of their coaching practice through interactive reflection, interpretative evaluation, and the sharing of experience. (pp. 16, 17)
As Bachkirova and colleagues (2020) note, studies focusing on the use of supervision often do not distinguish between different modalities of supervision, such as a dyadic or group setting, or the presence/absence of an external supervisor. Likewise, practitioners also differ in their understanding of supervision (Lawrence & Whyte, 2014). Previous studies have shown that the different modalities can differ, for example, in terms of frequency of use or perceived value (e.g., McAnally, Abrams, Asmus & Hildebrandt, 2020; Homer, 2017). Following Grant (2012), we distinguish three forms of supervision, namely informal supervision (or collegial exchange), peer supervision (or intervision), and coaching supervision in the narrow sense (one-to-one, group), depending on the degree of formalisation. Informal supervision denotes an informal and unstructured exchange among practitioners (dyadic or multiple people) (Grant, 2012). Peer supervision refers to a form where a group or dyad regularly and independently discusses its coaching practice without an external supervisor (Grant, 2012; Kotte & Zimmermann, 2022). In the case of coaching supervision, which can take place in either a dyadic or group setting, work is carried out with a supervisor for a fee (Grant, 2012; Kotte & Zimmermann, 2022).
Since its beginnings in the early 2000s (e.g., Hawkins & Smith, 2006), research on supervision of coaches has increased, especially in recent years.
Although different forms of supervision (as outlined above) are present in practice, the literature/research tends to focus predominantly on coaching supervision, i.e., with a formal supervisor. Accordingly, review papers address coaching supervision and summarise existing studies on business coaching supervision (Joseph, 2016), discuss recent research and identify and map themes in the contemporary literature (Bachkirova et al., 2020; Moyes, 2009), as well as examine the literature on the supervision of coaching in general (Kotte & Zimmermann, 2022; Tkach & DiGirolamo, 2017). In the most recent systematic literature review, Bachkirova and colleagues (2020) identify three key themes beyond the use of supervision that emerged from the literature on coaching supervision: its value, clarification of the concept, and development of the theory. Overall, studies to date often focus on coaches’ subjective evaluations of the benefits of supervision, the factors influencing its effectiveness and the implementation of coaching supervision (e.g., Grant, 2012; Jepson, 2016). However, empirical investigations of relationships between influencing factors and effectiveness via experimental designs or regression-analytical examinations of various predictors are the exception (e.g., Graßmann & Schermuly, 2018; Müller, Kotte & Möller, 2020). Most studies rely on coaches’ subjective perceptions of influencing factors and effectiveness. As a consequence, many questions remain unanswered (Bachkirova et al., 2020) and research has not greatly advanced (Kotte & Zimmermann, 2022).
While robust empirical evidence of the effectiveness of supervision is still lacking (Bachkirova et al., 2020), theoretical reasoning supports its value. Hawkins and colleagues (2019) argue, for example, that supervision is essential for working on the self, the most important instrument of a coach. In this way, a coach can continuously adapt to the demands of coaching, gain new perspectives and ensure quality (Hawkins et al., 2019). Coaching practitioners describe a similar range of benefits from supervision (e.g., development of capabilities and new perspectives; Lawrence & Whyte, 2014; McAnally et al., 2020). However, as highlighted above, not all coaches put their positive attitude (namely that the use of coaching supervision is beneficial) into practice by actually applying supervision to their own coaching practice. This notable discrepancy between attitude and behaviour regarding supervision is not limited to early studies of actual supervision practice (Hawkins & Schwenk, 2006a; Salter, 2008). Jepson (2016) found that almost all coaches who reported not using regular supervision themselves believed that coaches in general should use it regularly.
Moreover, the extant research on actual supervision practice addresses geographical regions in an unequal way. We do have some knowledge about supervision practice in certain geographical regions, particularly in the UK (e.g., Jepson, 2016; Hawkins & Schwenk, 2006a). At the same time, other regions are underrepresented in research (Tkach & DiGirolamo, 2017). German-speaking countries, for example, have received only superficial attention in international studies (McAnally et al., 2020; Passmore, Brown, Csigas & the European Coaching and Mentoring Research Consortium, 2017; Passmore, Brown, Greif, Rauen & the European Coaching and Mentoring Research Consortium, 2018). Nevertheless, this region has a long tradition of supervision as an independent form of work-related counselling (Kotte & Zimmermann, 2022) and constitutes a relevant market with an estimated 9000+ business coaches in Germany, Austria and Switzerland (Stephan & Rötz, 2018). Initial findings suggest, however, that actual supervision practice can vary considerably from region to region (McAnally et al, 2020; Hawkins & Turner, 2017). There are various reasons for this, for example the historical development of markets or accrediting-body requirements (Tkach & DiGirolamo, 2017). All in all, the research does not yet paint a clear picture because detailed findings about several relevant regions are lacking and possible regional differences limit the generalisability of results (Bachkirova et al., 2020). We therefore focus on the German-speaking countries in our study. We provide a point of comparison for future research which will incorporate results from other regions so that a better understanding of regional differences can be reached.
Existing studies on actual supervision practice look at a wide range of different aspects of use such as functions, experiences, frequency, barriers and issues addressed. In reviewing these studies, we find that similar studies tend to co-exist somewhat “loosely” side by side, for example because of different methodological approaches and the use of different data collection instruments (Tkach & DiGirolamo, 2017). We thus build on existing studies (e.g., Grant, 2012; Hawkins & Turner, 2017) to derive our questionnaire and promote better comparability of studies in the longer term. We also consider two aspects that are called for in the literature: critical moments in coaching supervision and reasons against the (more frequent) use of coaching supervision. In regard to critical moments, we take up Moyes' (2009) impulse to gain insight into learning moments in coaching supervision and to transfer the study of such critical moments in coaching (De Haan, 2010) to coaching supervision. We also explore reasons for not using coaching supervision to capture the underrepresented views of coaches who do not use supervision (Bachkirova et al., 2020).
Lastly, previous studies have often been criticised for lacking a theoretical framework (Bachkirova et al., 2020; Tkach & DiGirolamo, 2017) for examining supervision-related behaviour and its determinants in more detail. To give our study on coaches’ actual supervision practice and their attitudes towards supervision a more stable theoretical grounding and to contextualise our research, we use the Theory of Planned Behaviour, one of the most popular and established models, to explain behaviour (Ajzen, 1985; 1991).
The Theory of Planned Behaviour posits that an individual’s behavioural intentions, and in turn their behaviour, depend on three key determinants (Ajzen, 1985; 1991): the attitudes toward the behaviour (i.e., a positive or negative evaluation of the behaviour and reasons associated with this evaluation, e.g., whether coaches consider coaching supervision useful or not and why); subjective norms (i.e., the level to which a person assumes relevant others expect them to perform the behaviour, e.g., coaching associations expecting coaches to take part in coaching supervision); and perceived behavioural control (i.e., perceived constraints to performing a particular behaviour, e.g., perceived time and financial constraints that prevent coaches from engaging in coaching supervision). Perceived difficulties associated with the implementation of a behaviour (i.e., low behavioural control) are assumed to influence the probability of the behaviour independently of the attitude towards it. For this reason, discrepancies can arise between attitudes and actual behaviour (Ajzen, 1985; 1991). Numerous studies suggest that the theory can be used to predict behaviour across contexts as well as to inform useful interventions that may change behaviour (Steinmetz, Knappstein, Ajzen, Schmidt & Kabst, 2016).
Against this background, our study seeks to investigate the following three research questions.
Firstly, we focus on actual behaviour: RQ 1. How do German-speaking coaches behave in regard to supervision? The following aspects are of interest here: frequency of supervision use, topics discussed during supervision and boundary conditions such as selection of the supervisor, finding a peer supervision group and fees.
Secondly, as outlined above, behaviour is decisively, and independently, influenced by attitudes towards it and by perceived behaviour control. Following the Theory of Planned Behaviour, both attitudes and perceived behavioural control are based on practitioners’ salient beliefs (Ajzen, 1985; 1991). For example, attitudes are based on beliefs about the likely consequences, both positive and negative, of performing the behaviour. Perceived behavioural control is based on beliefs about the existence of factors that may help or hinder the performance of the behaviour. We therefore look at attitudes, perceived behavioural control and salient beliefs that may influence actual behaviour: RQ 2. What attitudes do coaches have towards supervision and how do they perceive other aspects that may shape salient beliefs? The factors that potentially shape salient beliefs that we explore are the functions of coaching supervision, critical moments in coaching supervision, possible discrepancies between ideal use and reality, reasons for not using coaching supervision and disappointments with supervision.
Finally, we look at the subjective norms that may shape behaviour. Coaching associations potentially exert (social) pressure on their members to use supervision (DBVC, 2022; DGSV, 2012; Bso, 2021; EESC, 2021; ICF, n.d.; ÖVS, n.d.). This leads us to ask: RQ 3. Do coaches who are organised in coaching associations differ from those who are not a) in their assumption of how often peer and coaching supervision should be used, b) in their attitude towards peer and coaching supervision, and c) in their actual use of supervision?
An online survey was created for our study[1] which contained both closed and open questions. The only requirement for participation was that the participant worked as a coach in a German-speaking country (i.e., Germany, Austria or Switzerland). With this mixed-method approach, it was possible for the participants to add alternative answers to those provided and thereby provide us with in-depth insights, for instance into the way that peer supervision groups come together.
The survey was carried out online between August and October 2017. It was delivered in German, compiled using SoSci Survey (Leiner, 2017) and made available via www.soscisurvey.de. To reduce possible self-selection among the participating coaches, “supervision” was not directly mentioned in the title of the survey, this being “How do coaches reflect upon their work? Quality-assurance strategies in coaching”. The link to the online survey was disseminated via social networks (LinkedIn, Xing), coaching associations (including the German Association for Supervision and Coaching (DGSv), the German Federal Association for Coaching (DBVC), the Austrian Association for Supervision and Coaching (ÖVS)), coach-training institutes (including the German Psychologists Academy) and the authors’ professional networks. The link was clicked on 745 times, 212 people began the survey and 124 completed it (an overall completion rate of 58.5%). As three data sets showed that the survey was only clicked through and not answered, 121 complete data sets were ultimately included in this survey.
The sample comprised 121 coaches from German-speaking countries. The average age was 50.24 (SD = 11.41, n = 99). There were 67 female and 35 male coaches, while 19 participants gave no information on their sex. Almost half of the coaches stated that they belong to a coaching association. Almost two thirds worked as external coaches. On average, 30.3% (SD = 22.3, n = 108) of the annual working time of all respondents was spent working as a coach. The coaches charged an average of €142.60 (SD = 64.4, n = 100) for an hour of coaching and had carried out an average of 15 (SD = 12, median = 14) coaching processes in the year prior to the survey. The coaches reported that they had been using peer supervision for an average of about 12 years (SD = 10.22, n = 88, median = 10) and coaching supervision for an average of almost 14 years (SD = 10.41, n = 86, median = 10). Compared to German coaching-market surveys from the reference period (Middendorf 2017, 2018; Stephan & Rötz, 2018), the sample is largely comparable in regard to age structure, the average number of coaching processes carried out, the proportion of coaching hours within annual working hours and coaches’ sex distribution.
When compiling the survey, we gave consideration to previous studies on coaching supervision, e.g., by transforming some qualitative answers into closed question answers (e.g., Grant, 2012; Hawkins & Turner, 2017; Hodge, 2016), including on the topics covered during the supervision of coaching and on the reasons against (more frequent) use of it. In accordance with the research questions, the survey covered the following areas:
We used two items each to measure the attitude towards peer (α = .86) and coaching (α = .83) supervision. A sample item was: “I am generally very positive towards peer/coaching supervision”. To capture the perceived barriers to supervision, we formulated closed answer options on the basis of the results obtained by Grant (2012) and Hawkins and Tuner (2017).
Inspired by the research on critical moments in coaching (e.g., De Haan, Bertie, Day & Sills, 2010a), we explicitly asked about moments which were experienced during the supervision process as exciting, tense, challenging or meaningful. We also collected sociodemographic and professional background information by means of questions on age, sex, the respondent’s own coaching activities and the demands of supervision.
The analysis of the quantitative data was carried out using IBM SPSS Version 27.0 (IBM Corp., 2020) and Microsoft Excel©. For the analysis of the qualitative data, we used MAXQDA 2018 (VERBI Software, 2017) in order to conduct a qualitative content analysis (Mayring, 2014; Schreier, 2012). Regarding critical moments (1) and functions of coaching supervision (2), we used existing categorisations (namely (1) De Haan, Bertie, Day & Sills, 2010b; (2) Hawkins & Smith, 2006) and supplemented these deductive categories with inductive categories drawn from the data. To summarise the content of the answers to the remaining open questions, we inductively developed categories from the data. The coding was carried out by the first author and one of the co-authors recoded 10% of the material (Kuckartz & Rädiker, 2019) to determine the interrater reliability. The kappa coefficient of agreement (as per Brennan & Prediger, 1981) was calculated and achieved a very high level (κn = .81; Landis & Koch, 1977).
The respondents most frequently used informal supervision (median = monthly), followed by peer and coaching supervision (median = once every four months) to reflect upon their work. An aggregated overview can be seen in Figure 1. Overall, a majority (73%) used both of the structured peer and coaching supervision forms; only 7% eschewed both forms.
Absolute frequency of the use of supervision. Shown in aggregated form; Supervision practice was collected according to the following divisions: daily, weekly, fortnightly, monthly, once every four months, half-yearly, yearly and never/no response; n = 121.
To check possible relationships between the use of different forms of supervision, we calculated the Kendall Tau-b coefficients (two-sided). After Bonferroni correction, two significant positive correlations with medium effect sizes (Cohen, 1988) resulted: Between informal and peer supervision (r = .29, p < .001) and between peer and coaching supervision (r = .29, p < .001). To further test whether how often someone coaches affects supervision use, we calculated the Kendall Tau-b coefficients (two-sided). After Bonferroni correction, no significant relationship was found between coaching frequency and the use of specific forms of supervision.
Table 1 shows which topics the coaches worked on per form of supervision (multiple options were possible; answer options were provided). A Chi-squared test (two-sided) was carried out. The frequency of mentioning certain topics differed significantly, depending on the respective form (χ2 (12, N = 965) = 23.69, p < .05), with a small effect size (w = .157). A post-hoc test showed that Methodological approaches (even after Bonferroni correction) were addressed less frequently than statistically expected in coaching supervision while they were addressed more frequently than statistically expected in informal supervision (without Bonferroni correction). On the other hand, Challenging cases or clients were addressed more frequently than statistically expected (without Bonferroni correction) in coaching supervision.
Multiple answers were possible, N = 965 for 121 surveyed coaches.
* Distribution of the percentage of mentions of a topic on the respective forms of support.
Among the coaches, 37 stated that they varied the use of peer supervision according to their own needs. More frequent use was reported on specific occasions, for instance when encountering “personal limits” (15.25[2] ) or “points requiring discussion” (15.12). In these cases, it was also crucial whether the peer supervision group was immediately available, i.e., whether urgency and opportunity went hand in hand and whether one had sufficient time to participate. “Depending on the circumstances”, some individuals reported that “different peer supervision partners” (15.11) were consulted.
The same was true for coaching supervision: Here, 42 respondents stated that they varied their use of coaching supervision according to their needs. On specific occasions such as when a “[high] number of cases” (21.12) or “difficult cases” (21.20) were seen, many reported more frequent use. Some also referred to seeking out a different setting if the occasion required it, for example “individual supervision [instead of group supervision] according to need and the problem/case” (21.6). The reduction of use due to a lack of time was also reported.
When asked about the criteria which the coaches used when selecting a suitable supervisor, the respondents most frequently indicated (multiple options were possible; answer options were provided) the following aspects: experience as a supervisor (n = 64), professional background (n = 61) and experience as a coach (n = 52). Also frequently mentioned were personal experience with the supervisor (n = 48), recommendation by colleagues/acquaintances (n = 47) and one’s own gut feeling/feeling of affinity (n = 41). Less frequently mentioned were affiliation with a specific coaching/supervision approach (n = 21) and the supervisor’s sex (n = 5).
More specifically, when asked about the professional background of their own supervisor, the coaches most frequently indicated (multiple options were possible; answer options were provided) a psychological (n = 67) or pedagogical/educational (n = 66) background. Less frequently mentioned were supervisors with an educational background in economics (n = 13) or medicine (n = 11).
87 coaches responded to the question on how they became aware of their peer supervision group. According to the descriptions, the most frequent response (n = 41) was through contact with a professional network, mostly of an informal nature. In other cases (n = 26), the contact came about within the context of their own training or further education (e.g., further coaching training, conferences). Some (n = 12) reported starting the group themselves while others (n = 5) stated that they were asked to join.
The respondents paid an average of just under €121 (SD = 44.98, n = 81) for an hour of coaching supervision in an individual setting and just under €74 in a group setting (SD = 43.44, n = 39). This compares to an average income of €142.59 (SD = 64.42; n = 100) per self-conducted coaching hour.
The recorded attitude of coaches towards peer and coaching supervision tended to be very positive (peer supervision: MD = 5.78, SD = 1.34; coaching supervision: MD = 6.24, SD = 1.04; seven-point scale from 1 = strong aversion to 7 = strong approval). Based on the scale values, we examined the possible relationships between attitude and use of peer and coaching supervision by calculating the Kendall Tau-b coefficients (two-sided). In conformity with the Theory of Planned Behaviour, a more positive attitude towards peer supervision was associated with more frequent use of peer supervision (r = .24 p = .002). After Bonferroni correction, the effect (p < .0125) is significant and small. The same was shown for coaching supervision and its more frequent use (r = .27, p < .001, small effect). A positive attitude was also associated with a more frequent use of peer supervision (r = .22, p = .004).
In an open question, the coaches indicated which functions coaching supervision satisfied for them personally (n = 212). We categorised the answers deductively along the functions of coaching supervision (developmental, qualitative and resourcing-related) as described by Hawkins and Smith (2006). Most frequently mentioned were aspects of the developmental function (e.g., support for reflection and further development). We also inductively added the social function (5%) which particularly focuses on exchange and the sense of community. An aggregated overview can be seen in Figure 2.
To gain insights into learning moments, we asked an open question about critical moments. The coaches indicated 67 moments which they had experienced as critical (exciting, tense, challenging or meaningful) in a past supervision process. We categorised the answers according to the 12 codes for critical moments in coaching identified by De Haan and colleagues (2010b). The frequencies can be seen in Figure 3.
Relative frequency of the coaching supervision functions mentioned; based upon 212 open answers; categories aligned with Hawkins and Smith (2006; developmental, qualitative, resourcing); the social function was inductively derived.
Frequencies of critical moments in coaching supervision, structured on the basis of the codes from De Haan et al. (2010b). N = 67; Codes were partially adapted to the material, adaptations resulted from the different context (coach as client in coaching supervision) and the divergent emotions which were expressed. The category “cases rather than moments” was added.
It turns out that – similar to the research on critical moments in coaching – moments of learning in particular were mentioned here (for frequencies see Figure 3 “Frequencies”), such as:
Moments of emotional experience were much less common (for frequencies see Figure 3: Significant emotional experience): These were characterised by emotions that were perceived to be rather unpleasant such as anger, sadness and doubt.
We asked the respondents who actually used the structured forms of peer and coaching supervision (individual and group setting) to indicate how often coaches should ideally use each form. The Wilcoxon signed-rank test (two-sided) was used to calculate whether these evaluations differ from the coaches’ own actual use. The results clearly show the tendency for practice to lag behind the ideal: After Bonferroni correction, the use of peer supervision (p < .0125) is significantly lower than the use estimated for the ideal case (z = -3.26, p = <.01, n = 98; r = .33, medium effect size, Cohen (1988)). The same applies to the use of group coaching supervision (median = yearly; z = -4.69, p < .001, n = 55; r = .63, strong effect size). The tendency is weaker for the use of individual coaching supervision (z = -2.12, p = .03, n = 72). The difference after Bonferroni correction is not significant here.
22 coaches stated that they do not use any coaching supervision at all. When asked about the reasons, they agreed (multiple options were possible; answer options were provided) most frequently with the statement “I belong to a peer network in which I discuss my cases” (12x), followed by “My clients don’t require it of me” (7x) and “It’s too expensive for me” (6x).
The coaches who used coaching supervision at least sometimes named the following reasons (multiple options were possible; answer options were provided) as those which prevented them from using coaching supervision more often: 23 stated that “More supervision is too expensive for me”, 12 that “More supervision is too time-consuming for me”, three that “I can’t find (a) good supervisor(s)” and two that “It affects my coaching cases too much”. Overall, 12 gave free text answers which can be summarised in the following reasons: “A lack of resources”, “A lack of transparency about the supervisors’ background”, “Not needed” and “Too much of a public figure”.
Almost 28% of coaches (n = 33) stated that they had in the past been annoyed with the supervision process because their expectations for the respective form had not been met. In regard to peer supervision, coaches described annoyance at the working style (e.g., entrenched routines) or at other group members (e.g., a lack of willingness to engage in continuous professional development). In regard to coaching supervision in an individual setting, the respondents described annoyance at the supervisor’s working style (including the methods used), at the supervisor as a person, at the content (priorities set) or a lack of success. In the group setting (coaching supervision) there was also dissatisfaction with the working style (e.g., a lack of structure), with the competence of the supervisor, with the content (a lack of fit with one’s own problem, level too low), with little personal progress or with difficulties with the group working together (e.g., a lack of trust and openness, a lack of a sustainable working alliance, group-dynamic processes).
Following the Theory of Planned Behaviour, namely subjective norms as an influencing factor, we examined possible differences between coaches who belonged to a coaching association and those who did not. Interestingly, no significant effect was seen: The group of coaches organised in coaching associations did not differ in the acceptance of how often peer and coaching supervision should be practised (Mann-Whitney U Tests, two-sided; peer supervision: p = .07, individual coaching supervision: p = .51, group coaching supervision: p = .99). The same applies to the attitude towards peer supervision and coaching supervision (t-tests for independent samples; t(110) = -.690, p = .8; t(115) = -1,34, p = .578) and to the actual use of supervision of coaching (Mann-Whitney U Tests, two-sided; informal supervision: p = .14, peer supervision: p = .23, coaching supervision: p = .25).
The results of this study provide a deeper insight into the supervision practice of coaches as well as factors that may influence its use.
Generally, the majority of respondents used one or more of the forms of supervision, i.e., informal, peer, and coaching supervision. A higher degree of formalisation tended to go hand in hand with less frequent use. When peer and coaching supervision were used, in half of the cases this was every four months or less, i.e., at wide intervals.
In an international study, Hawkins and Turner (2017) found high use rates of supervision that are comparable to our study (globally: 83%, Europe: 81%). However, the rate is significantly higher than the rate found by Passmore and colleagues (2018) for Germany (approx. 61%; compared to 80%[3] in our study). One explanation may be that, by recruiting our sample partly through coaching associations anchored in the German-speaking tradition of work-related supervision for social sectors, the coaches participating in our study may have worked more frequently in the social or healthcare sectors than “mere” business coaches. In these sectors, supervision is very common and well established as a tool for quality assurance and continued professional development. As we did not control for the sector in which the coaches worked, future studies need to determine the influence of the sectors in which they work on their supervisory practice. However, the reported average hourly wage for coaching, which is lower than in comparable studies (Middendorf, 2017, 2018), points in a similar direction.
Our results also show that certain forms of supervision are preferred for specific topics. While methodological questions were more likely to be discussed in informal supervision, one’s own limitations, in the form of difficulties experienced with challenging cases or clients, were more likely to present themselves in coaching supervision. In future research, it would be fascinating to consider the “degree of difficulty” experienced in order to check whether a tendency to introduce “more difficult cases” into more structured forms can be confirmed.
In the case of the more strongly formalised forms of supervision, i.e., peer and coaching supervision, around one third of users varied their use depending upon the occasion, particularly when they perceived an additional need, for instance due to challenges. Taking into account the actual use, which is spaced at relatively wide intervals, it can be assumed that supervision was understood more as a “problem solver”. The use of a shorter-interval routine-based support (e.g., Hawkins & Smith, 2006) was seen less often. This is also supported by the fact that a higher annual working time as a coach was not associated with a more frequent use of coaching supervision.
As a rule, coaches spent less on a coaching supervision time unit than they earned with coaching. This shows an imbalance already reported by Hawkins and Turner (2017) and suggests a lower value reflecting on one’s own work than on the coaching work itself (Hawkins & Turner, 2017). Additionally, this indicates that one reason for not using coaching supervision more frequently appears to be a perceived excessive expenditure of resources (money, time). This conforms with perceived situational constraints of the Theory of Planned Behaviour and previous studies (e.g., Jepson, 2016).
Overall, the respondents tended to be very positive towards peer supervision and coaching supervision, and a more positive attitude went hand in hand with more frequent use, in line with the Theory of Planned Behaviour
Nevertheless, even a very positive attitude contrasts with a relatively restrained use. Moreover, as found in earlier studies (e.g., Hawkins & Schwenk, 2006a, 2006b), coaches overall fell short of the ideal that they themselves posited when assessed in terms of the frequency of their actual use of supervision. Possible reasons for this discrepancy can be seen in the individual. Previous research suggests that coaches may face fear, shame and their own narcissism in supervision (e.g., De Haan, 2017; Hodge, 2014, 2016; Salter, 2008) which can undermine a positive professional self-image and trigger fear of negative perception by others. This raises the question of whether, behind the positive attitude, there is an inhibiting fear of being questioned and feeling insecure as a supervised coach. It therefore seems crucial to promote trust and psychological safety in supervision (De Haan, 2017; Schein & Bennis, 1965) and to foster an overall positive error culture in the coaching community, for example at coaching conferences. Beyond coaches’ possible fears associated with supervision, boundary conditions (Ajzen, 2011) could also play a role in the relatively restrained use of supervision, namely the above-mentioned reported excessive expenditure of resources which could act as a form of behavioural constraint in terms of the Theory of Planned Behaviour.
Coaching supervision served a variety of positive functions for the respondents, essentially corresponding to the categories suggested by Hawkins and Smith (2006). The most frequently mentioned function of supervision was the developmental function; previous studies have reported something similar when examining other regions (e.g., Jepson, 2016 (UK); Lawrence & Whyte, 2014 (Australia/New Zealand)). In addition, supervision served a social function as indicated by previous individual studies (e.g., Grant, 2012; as reviewed by Tkach & DiGirolamo, 2017). This can be interpreted as an attempt to mitigate the isolation associated with being a coach.
The reported critical moments provide additional information on the significance of the different functions: In particular, coaches retained the memory of moments of learning about themselves (e.g., their own blind spots) or of discussed cases (e.g., new perspectives) – an indication of development, breakthroughs and changes being experienced (De Haan, 2010; Greif & Riemenschneider-Greif, 2018).
On the other hand, results showed that almost every third person had already experienced annoyance during supervision due to unmet expectations, most frequently towards the working method. Many of the negative experiences that were reported may be understood as ruptures in the working alliance between coach and supervisor (Ehrenthal, Möller & Zimmermann, 2020). Accordingly, there are indications that the working alliance should be well regulated (Graßmann & Schermuly, 2020) and that common bond, goal and task should not be neglected.
It is interesting that the coaches organised in coaching associations do not differ significantly from the non-organised in the use of the three forms of supervision or in their attitudes toward supervision. The Theory of Planned Behaviour allows various reasons to appear plausible. It is possible that active supervision practice is also widespread beyond the requirements of coaching associations and has taken firm hold in the self-concept of coaches. Alternatively, it may not (yet) have been possible for the coaching associations to establish such a practice as a binding subjective norm.
This study investigates the supervision practice of coaches in German-speaking countries on the basis of self-reported data; possible bias thus cannot be ruled out (e.g., memory errors, social desirability). It may be questionable to what extent it was possible for coaches to provide realistic information on the frequency of use as this was sometimes spaced at very wide intervals and rather event-dependent. The essential features of our sample correspond to German coaching market surveys. We therefore assume that our sample presents a sufficiently accurate picture of the German-speaking coaching market, even though a selection effect cannot be ruled out.
In future studies, it would be desirable to broaden the perspective beyond coaches, to involve the coaching clients and to investigate the extent to which supervision practice actually affects the practice of coaching as experienced by the clients. The benefit/effectiveness of supervision should also be conceptualised in a more differentiated way, ideally with consideration given to the different functions supervision fulfils. Future studies which build upon our study should thus investigate different functions, the factors that influence these functions (e.g., the form of the supervision or the frequency of its use) as well as the benefit/effect of supervision on both coach and clients in a more systematic manner.
The present study provides an overview of supervision practice and factors that may influence it in German-speaking countries. It is based on international supervision research to allow comparability with other studies. This results in a benchmark for further research, making it possible to identify regional differences as well as changes over time. The study also classifies these findings theoretically with the help of the Theory of Planned Behaviour and thereby provides an important building block for further research on the supervision of coaches.
What can coaching practitioners take away from this study? Overall, the results suggest that coaches are heading in the right direction. It still appears necessary to continue discussing the added value of structured supervision in the coaching community, especially in the form of regular and routine supervision. Not least, in order to increase the value of peer and coaching supervision that is experienced. This is not a mere marketing ploy, but rather about conveying a realistic picture of what coaching supervision can achieve and what functions it can fulfil for coaches.
The differentiated examination of the topics discussed in the various forms of coaching supervision as well as the disappointments associated with these forms enables coaching practitioners to reflect upon and, if necessary, to question their own use of the various forms. Negative experiences were also reported with the formalised forms of supervision. There is therefore a need for greater quality assurance and process reflection here. A focus on the working alliance and an accompanying joint process evaluation between supervisor and coach or among peers would be two starting points. Further potential lies, among other places, in targeted further supervision training, accreditation by coaching associations and in greater research attention.
Furthermore, the results underline the importance of networks of an informal and formal nature for coaches, whether for finding peer supervision groups or for collegial exchange. As indicated, this not only supports supervision practice but also counteracts the isolation (Satler, 2008; Tkach & DiGirolamo, 2017) associated with the activity of coaching.
We wish to thank Marlene Eigel for her great commitment in supervising the data collection during her master’s thesis.
The figures in parentheses refer to the coding of the respective case.
Based on the available data (n = 112), missing data is not taken into account (n = 9).
Jannik Zimmermann is a research associate at the University of Kassel, Germany. His research focuses on coaching, coaching supervision, and leadership development.
Silja Kotte is a professor of business psychology at the University of Applied Sciences Aschaffenburg, Germany. Her research focuses on coaching, human resource development and leadership.
Isabell Diermann is a freelance trainer and coach working on topics in the field of applied psychology such as communication, conflict management, groups, and leadership.
Heidi Möller is a professor of counselling psychology at University of Kassel, Germany. Her research focuses on workplace coaching, coaching supervision, and leadership development.