As a coach, licensed clinical social worker, and researcher, I have been curious about the definition of and boundaries around coaching since I first started learning about coaching over 10 years ago. There are a number of definitions of coaching that have been circulated, without any clear consensus that any of the definitions is superior to the others. But if we can’t agree on what coaches do, it seemed important to find a way to describe coaching so that we and future researchers would, at least, have a common language when talking about coaching.
So, in 2009, Tatiana Bachkirova, a professor and researcher at Oxford Brookes University in the UK, and I submitted a research proposal to the Institute of Coaching at Harvard University and were awarded a grant to develop an instrument that could be used to describe what happens in a coaching session – the Coaching Process Q-set.
Together with a student at Oxford Brookes, Adrian Myers, we developed an initial set of items to describe what might happen in a coaching session, keeping in mind the coach, client, and coach-client dyad. We were inspired by a series of research studies from the psychology process research literature. Enrico Jones had developed a set of 100 items to describe psychotherapy sessions and went on, with colleagues, to use these items:
• to describe differences between psychotherapy approaches (descriptions of prototypes for various forms of therapy)
• to explore differences between how psychotherapists said they practiced (e.g., expert cognitive-behavioral therapists vs. psychodynamic therapists) and the actual sessions of these therapists
• to evaluate which items and approaches seemed to lead to better outcomes
Coaching is a relatively new field and the psychological process literature had a much longer history than the coaching process literature and we hoped to provide future coaching researchers with a tool that could describe coaching at a more granular level than previous approaches we had found.
We convened a group of expert coaches in the UK to explore and revise our initial list of items, followed by groups in the US and Canada. Each time, we incorporated the feedback we received from the group. Following this, we posted the list of items on the internet and gathered feedback from 207 visitors from 26 countries about these items. At this point, based on this feedback, it appeared that our list was reasonably complete and ready to be piloted.
Again, we turned to the internet to ask coaches to use our list of 80 items to describe an imagined, typical mid-engagement coaching session, putting each item into a bucket ranging from “most characteristic” of a session to “most uncharacteristic”. Using a set of techniques called “Q methodology”, 41 participants from 5 countries sorted the list of items online after which the researchers used a form of factor analysis across participants to see whether we could identify sub-groups of participants who described their way of coaching as similar to other coaches in the sub-group and as different from the coaching described by coaches in other sub-groups.
After analyzing the results, no clear subgroups could be identified. Although, of course, there was not complete agreement about the ranking of the 80 items, there appeared to be relatively strong consensus among the 41 participants about which items were most and least characteristic of a typical coaching session.
These items / themes were considered most characteristic:
• A focus on the client’s agenda, not the agenda of the coach or of a third-party
• The coach pays attention to the client’s overall goals and goals for the session
• A sense of optimism empathy and rapport
• A focus on exploration, including “questions to open new possibilities” for the client and, for many, an exploration of the client’s mindset (beliefs and assumptions) and values, as well as an exploration of the client’s resources (e.g., strengths, accomplishments, and/or external resources) and, to some degree, of the deeper meaning of a presenting issue
These items / themes were considered to be least characteristic:
• Advice giving
• Helping the client to deepen his or her emotions
• Exploring the client’s “apparent defensiveness” or “unconscious motives”
• The coach sharing his or her own feelings and bodily sensations that are evoked during the session
• The coach is verbose, uses an intervention mechanistically, coach and/or client interrupt one another
We chose to use the full list of 80 items for this pilot, although some of the items are likely to be more useful when used to describe an actual coaching session. The advantage of using the full list, particularly when items are sorted or ranked in comparison to one another, is that these results could be more easily compared to results based on using the same 80 items to describe actual coaching sessions. The disadvantage to using the full list is that items such as “coach is verbose” or “coach appears to be using an intervention mechanistically”, while potentially important in describing an actual session, are likely to be ranked as least characteristic by nearly all coaches when they describe an imagined, typical session.
Although the results from this sample of 41 coaches did not allow us to distinguish clear subgroups of coaches and coaching styles, it is possible that future research involving more and/or different coaches would lead to different results. For example, coaches who practice eclectically, combining different methods and frameworks, may make it difficult to tease out subtle differences in coaching styles. Comparing a group of psychodynamic coaches to a group of gestalt or behavioral coaches might lead to clearer differences between those groups of coaches. It is also possible that some of the more seemingly apparent differences between coaching approaches involve the language and labels used by each approach. In attempting to remove “jargon” from our items, we may have removed some of the potential differentiators between approaches. At the same time, if the differences are, in fact, based more on language and labels than on what coaches actually do, this raises the question of how important these differences actually are in coaching sessions and outcomes.
Although we piloted using the instrument using Q methodology and participants sorted the 80 items into a forced distribution, future researchers might wish to use the same items (or a subset of these items) with ratings on a Likert scale (e.g., from -5 to +5) for greater simplicity and ease of use. It would also be interesting to evaluate the 80 items psychometrically to see how well these items reflect what is and isn’t happening in a coaching sessions (or across coaching sessions) and whether the list of items needs any fine-tuning.
Going forward, we hope that this tool can be used in future research to better define and understand coaching, as well as by coaches, teachers, and supervisors to explore how coaches think about and practice coaching. Coaches involved in the development and piloting of the Coaching Process Q-set often told us that they found the process useful in thinking about their own coaching and we hope that this can help to move coaching research and practice forward as our field continues to mature.