What research says about coaching, and where it stops

As a qualified social worker and a manager in social services, I am used to the question of what we actually know about whether something works. I ask the same of my own work. That is why the research behind Leading from within has its own page, How we know. Here I go through it with a bit more room, and with as much attention to where it ends as to what it shows.

First, a few words on reading this kind of thing. A meta-analysis pools the results of many studies. A randomised trial assigns participants to groups at random, which makes it easier to say something about cause. An association means two things tend to go together, and it says nothing about which affects which.

What coaching seems to do#

Three meta-analyses point the same way. Coaching has clear effects on how well people reach their goals and on self-efficacy, their belief in their own ability. They are an analysis of psychologically informed coaching approaches, an analysis limited to randomised trials and an earlier analysis of coaching in organisations.

The first of them also found that approaches combining several methods had a slightly larger effect than single-method ones. That matches how I work. One session might hold systemic questions, some narrative work on how you tell your own story, and a short exercise, depending on what you need that day.

More sessions are not automatically better#

I think this finding matters. The effect of coaching does not seem to grow with the number of sessions or with how long the coaching goes on. That is the direction of a meta-analysis of learning and performance outcomes, the trials-only analysis above, and an analysis of randomised trials of executive coaching.

For me it means I can't point to the research to sell more sessions. Your goals decide how long we keep going, and every three months we review progress against them.

Mindfulness and yoga#

A review of 23 randomised trials found that workplace mindfulness training reduces perceived stress and anxiety and increases well-being. A meta-analysis of workplace yoga found that yoga reduces perceived stress.

For leaders in particular there is a randomised study of 128 leaders, in which six weeks of mindfulness-based self-leadership training increased self-efficacy, resilience and self-leadership. Three reviews, of mindfulness and meditation for managers, of mindfulness in leaders and of self-leadership and mindfulness training, report better self-regulation, though the studies they draw on vary in quality.

Two things are worth knowing. The stress studies measure perceived stress, meaning how participants rate it themselves. And there is no evidence that mindfulness training for leaders affects their staff or teams. So I promise nothing about your team.

Living in line with what you value#

A meta-analysis of 75 studies with 36,533 participants found that perceived authenticity goes with well-being (r = 0.40) and with engagement at work (r = 0.37). That is an association, and it says nothing about what causes what. A classic study of goals found that goals that fit your own values are reached more often and bring more well-being.

The concept of authenticity itself is debated among researchers. So I would rather talk about living in line with your values than about finding a true self.

Where the research stops#

There are things the studies can't answer, and I want to be clear about them.

I haven't found any study that compares coaching and mentoring head to head. On mentoring there is a large meta-analysis that linked it with many good outcomes, but the effects were generally small. The studies were mostly comparisons after the fact, and I write more about that in Mentor or coach?.

There is also little research on leadership programmes aimed at staff mental health in healthcare. A systematic review found only seven studies it could include. Four of them showed a statistically significant improvement, and the most promising ones had reflective group elements spread over several seminar days. The authors themselves call the evidence limited.

Managers' own well-being and the way they lead go together. A meta-analysis found links between how leaders lead and how they feel, but the studies can't say which way it runs. It may just as well be well-being that shapes leadership.

None of these studies is about Swedish social services. I read them as support for how I set up the work, and as a reason to be careful about what I promise.

How I use it#

In practice, this is what the research gives me. We set clear goals from the start and follow them up, and the goals decide the length. Exercises such as breathing, yin yoga and meditation are there when they help you settle, and I don't claim they will change your team. When something you tell me sounds like a matter for healthcare, I say so.