Teaching Mindfulness to Teachers: a Systematic Review and Narrative Synthesis

Teaching is a highly stressful profession (Smith et al. 2000); approximately 40% of teachers in the USA leave the profession within the first 5 years of qualifying (Ingersoll 2002) and 73% of newly qualified teachers in the UK consider leaving (Association of Teachers and Lecturers 2015). Teachers report that current teaching climates generate high levels of stress, which lead to work-related fatigue, depression and anxiety, cynicism and low self-efficacy (NASUWT 2013). Furthermore, teacher stress and burnout (exhaustion with depressive symptoms; Swider and Zimmerman 2010) can negatively impact pupil engagement and learning through teacher absenteeism, reduced self-efficacy and diminished teaching effectiveness (for a review of the evidence, see Bricheno et al. 2009; Roeser et al. 2012). Given these potential negative and costly effects of teacher stress, there is impetus to identify effective interventions to support teachers to stay healthy and remain within the profession.

In the current paper, the term stress refers to physical symptoms, such as sleep disturbance, and psychological symptoms, such as depression, anxiety and burnout. Jennings and Greenberg (2009) argued that teachers’ capacity to cope with work-related stress relates to their social and emotional competence (SEC), defined as an awareness and ability to regulate emotions. In turn, emotion regulation facilitates teachers’ sense of mastery and protects wellbeing. Existing research on stress, emotion regulation and self-efficacy in teachers supports the proposal that these three constructs contribute to effective teaching practice.

Emotion regulation is the capacity to effectively manage one’s emotional reactivity, including the conscious and unconscious use of strategies, and internal or external resources, to decrease, maintain or increase positive and negative emotions, in either an anticipatory or responsive manner (Cole et al. 2004; Gross and Thompson 2007). Poor emotion regulation in teachers has been associated with more frequent and enduring negative affect, increased negative interactions with pupils, stress and burnout, and attrition from the profession (Darling-Hammond 2001; Montgomery and Rupp 2005). Specifically, two components of emotion regulation (poor emotional appraisal and self-regulation) significantly predicted burnout (Chan 2006), with effective emotion regulation predicting increased teacher self-efficacy (Chan 2004). Teacher self-efficacy is considered to be a protective factor against the effects of stress and burnout (Beltman, Mansfield, & Price 2011; Caprara et al. 2006). Self-efficacy is the belief in one’s ability to persevere with a course of action in pursuit of a valued goal (Bandura 1992). In the context of teaching, higher teacher self-efficacy has been linked to perseverance with challenging students and improved pupil behavior in the classroom (Robertson and Dunsmuir 2013), whilst lower teacher self-efficacy has been associated with increased stress and lower occupational commitment (Klassen and Chiu 2011).

Individual differences in emotion regulation and self-efficacy may contribute to a vulnerability to stress in teachers; therefore, interventions that target these constructs may protect teachers against potential stress. Mindfulness-based interventions have burgeoning promise for their applicability to occupational contexts, and there is emerging evidence for their efficacy in enhancing emotion regulation and self-efficacy. Mindfulness is the ability to focus attention on the present moment while possessing an orientation of openness, curiosity and non-judgement (Bishop et al. 2004). Meta-analytic reviews have reported medium to large effect sizes of MBIs on indicators of psychological health (Carmody and Baer 2009; De Vibe et al. 2012). Similarly, when delivered across occupational settings, there have been reported positive effects of MBIs on stress and wellbeing (health care providers, Escuriex and Labbé 2011; health care professionals, Irving et al. 2009; working adults, Virgili 2013). The traditional format for delivery of a MBI is eight weekly 2.5-h sessions, which incorporate formal mindfulness practices (e.g. sitting and movement meditation, body scanning and mindful eating), experiential group discussion, psycho-education and home practices. The two most well-established programmes are mindfulness-based stress reduction (MBSR; Kabat-Zinn 1990), which was originally developed for hospital patients experiencing conditions that were difficult to treat with medical interventions, and mindfulness-based cognitive therapy (MBCT; Segal et al. 2002), which was originally developed to prevent relapse of recurring depression and includes specific components from cognitive-behavior therapy.

In recent years, MBIs have been adapted and applied in the teaching context. Jennings and Greenberg (2009) have proposed their logic model, which suggests that MBIs work directly to help teachers recognise (i.e. greater self-awareness) and regulate stress reactions (i.e. emotion regulation; Jennings and Greenberg 2009). Consistent with this model, we propose that the main outcome of MBIs for teachers is reduced stress, with intermediary increases in emotion regulation and self-efficacy. Increased mindfulness and self-compassion, resulting from the MBI, serve as the mechanisms of these effects. Previous research has demonstrated that MBIs can improve emotion regulation (Chiesa et al. 2013) through greater self-awareness and attentional capacity, giving the individual an enhanced ability to detect (through greater awareness of thoughts and bodily sensations), decentre from, accept and modulate emotion in real time (Garland et al. 2011). MBIs can also positively alter one’s perspective on the self, specifically through increased decentering, reduced self-referential thoughts and increased self-compassion (Hölzel et al. 2011). Moreover, enhanced self-compassion can reduce negative appraisals of an individual’s teaching competence and thus increase self-efficacy (Neff 2003). Similarly, greater self-compassion acts as an effective emotion regulation strategy by responding to negative affective states (e.g. perceived inadequacy or failures as a teacher) with kindness rather than criticism (Hölzel et al. 2011). As such, research relating to MBIs often includes the measurement of trait mindfulness and self-compassion as indicators of cognitive changes that may precipitate other broader changes (e.g. decreased stress).

This systematic review aims to synthesise the current research evidence examining the effectiveness of MBIs for (i) reducing teacher stress as a main outcome, (ii) supporting gains in emotion regulation and self-efficacy as intermediary effects and (iii) mindfulness and self-compassion as mechanisms of action. The review includes a systematic search of published intervention studies where teachers of children aged 5–18 years have participated in an MBI. We will present a narrative review of the literature and discuss the results in the context of theoretical models of teacher stress. The limitations of the current evidence base and future directions for applying mindfulness in educational contexts will be considered.

Twelve publications reported findings from 13 studies (two studies in Jennings et al. 2011). Table 1 provides key information regarding each publication. Data were reported from a total of 589 participants, with the majority from qualified teachers in mainstream educational settings. Most research was conducted in the USA and a minority in Canada and the UK. Quantitative designs dominated, including controlled and non-controlled studies, with a minority reporting qualitative data. The quality assessment of the literature suggested that strengths lie in the reporting of study details, including aims, recruitment and data collection procedures. In addition, there was generally a good fit between the research questions and methods employed to assess these. Studies that faired better in quality included a theoretical framework, had a reasonable sample size and detailed the psychometric properties of the tools and analytical methods used. On the whole, the body of literature was weakened by little consideration of sample size in terms of analysis, or the reliability of the analytical process (with the exception of Taylor et al. 2016). In addition, only studies by Jennings et al. (2011, 2013) included user involvement (i.e. feedback from teachers on the intervention) in the design.

This review has examined the current evidence for the effects of MBIs on stress, emotion regulation and self-efficacy. The present review of evidence informs a proposed model of reduced teacher stress resulting from a MBI (Fig. 1). Through participation in mindfulness training, an individual may see gains in mindfulness (e.g. decentering, regulation of attention) and self-compassion (Hölzel et al. 2011) that lead to more effective emotion regulation strategies (Chiesa, Serretti, & Jakobsen 2013) and increased professional self-efficacy (Neff et al. 2005) and ultimately reduced stress (e.g. Carmody and Baer 2009; De Vibe et al. 2012). Importantly, within this model, there is a reciprocal interaction between increased self-efficacy and emotion regulation, as enhanced teacher self-efficacy increases effective emotion regulation and appraisal of effective regulation of affect (in a teaching context) enhances a sense of teaching efficacy (Bandura et al. 2003). Although the studies reviewed were not designed to directly test this model and as such the full model was not tested in this review, there is some support for gains in emotion regulation and reduced teacher stress following the MBIs. In addition, around a third of the increases shown in trait mindfulness and self-compassion were significant with variable effect sizes.

From the evidence reported across the 13 eligible studies, it is possible to conclude that MBIs for teachers show most promise for the proposed intermediary effect of emotion regulation. Effect sizes for emotion regulation tended to be larger with more of them showing statistical significance, although effects varied according to the measures used. Furthermore, qualitative data from teachers also demonstrated that the aspect of MBIs most commonly rated as helpful related to emotion awareness, recognition and understanding (Schussler et al. 2016; Taylor et al. 2016). Emotion regulation is proposed to be a key contributing factor toward teacher SEC (Jennings and Greenberg 2009); thus, the observed improvements in emotion regulation as a result of participating in a MBI would theoretically improve a teacher’s overall SEC. Jennings and Greenberg highlight the cascading benefits of improved teacher SEC for the classroom environment, and individual pupil wellbeing. However, none of the studies reviewed reported comparisons of emotion regulation at their chosen follow-up points (the longest of which was 3 months) making it difficult to ascertain lasting effects. Furthermore, there was a lack of data on the hypothesised cascading effects of mindful teachers in the classroom and on pupil outcomes. Long-term studies are therefore needed to assess changes in teacher effectiveness, classroom climate and pupil relationships.

Across studies, there was nascent promise for application of MBIs for reducing physical and psychological symptoms of stress, including burnout. Effect sizes were variable and the significance of results was inconsistent across studies. However, the greatest proportion of significant findings was for outcomes relating to teachers’ perceived stress. It may be that a distinction between perceived stress and subsequent distress (e.g. anxiety, depression, burnout) is necessary when studying the effects of MBIs in occupational settings. MBIs in occupational settings may be more fruitful in reducing stress, and thereby possibly preventing subsequent distress, rather than reduced current symptoms of distress. This proposal is consistent with the largest effects on emotion regulation, as a possible intermediary effect in reducing or preventing stress in the longer term. Longer follow-up and prospective studies are required to confirm these proposed preventative effects.

The inconsistency in findings about the impact of MBIs on psychological outcomes across studies could be due to intervention as well as methodological differences. There was a heterogeneous approach to delivering MBIs across the studies reviewed. It is beyond the scope of this review, and the stage of the evidence, to provide a detailed consideration of the content and theoretical underpinnings between the MBIs employed across studies; however, it is important to note the heterogeneity in terms of aims, content, duration and mode of delivery. The SMART-in-Education and CARE programmes include specific instruction on emotion management skills. Therefore, it is difficult to conclude whether the significant effects observed in the studies utilising these programmes are due to the mindfulness components of the course or taught elements on emotion regulation. It is possible that the effects of mindfulness training are enhanced when coupled with additional techniques, and the effectiveness of embedding mindfulness training alongside other stress reduction techniques for teachers merits investigation. Some studies reported that the mindfulness component of the intervention was adapted for teachers (Flook et al. 2013), whereas others were not (Gold et al. 2010). Based on current evidence, it is difficult to determine if tailoring for teachers promotes better outcomes. For example, the CARE intervention is highly tailored, yet did not consistently improve teacher self-efficacy across three studies (Jennings et al. 2011, 2013). In contrast, mindfulness-based wellness education (MBWE) has only minor adaptations for teachers and was associated with significant changes in self-efficacy (Poulin et al. 2008). These nuances in the design of the mindfulness programme are likely to have an impact on how effective the course is, and thus the outcomes reported. As the evidence base grows, it will be important to consider the effective components of MBIs for teachers.

Quantitative and qualitative data reported across studies suggest an improvement in teachers’ ability to be mindful, which indicates successful manipulation effects of the MBIs. However, in line with the broader work on MBIs, there is a need to better understand and isolate the mechanisms of change associated with participation in these interventions (Garland et al. 2011; Holzel et al. 2011). For example, group support is known to reduce work-based stress (Michie and Williams 2003) and is likely to be an important feature of the MBI experience (e.g. Irving, et al. 2014). Future research should focus on separating the effects of group attendance and intervention-specific effects by comparison with active group controls.

Overall, the body of literature reviewed demonstrated a number of limitations that need to be addressed in future research, including small sample sizes, insufficient statistical power and the absence of active controls. Where MBIs have been compared to an active control intervention for health care professionals, no significant differences were observed between groups, with each reporting greater relaxation and life satisfaction (Poulin et al. 2008), demonstrating the importance of this experimental control. As can be the case in intervention studies, teachers in the reviewed studies were not blind to the intervention aim and were also self-selecting. This is routine in psychological interventions (and in MBIs) where the readiness for intervention, and the personal decision to engage, is deemed important to outcomes as well as reflecting ethical practice (Lyubomirsky et al. 2011; Seligman et al. 2006; Seligman et al. 2005). It is unknown whether referring teachers to MBIs, or delivering them wholesale to a school staff, would be an effective, preventative approach to stress management.

Finally, only three studies reported fidelity checks (Benn et al. 2012; Jennings et al. 2013; Roeser et al. 2013). Checks of fidelity can be helpful for ensuring that reported effects result from the application of specific interventions and not other extraneous variables (Horner et al. 2006).