Table 1.
Empirical studies on MBI's in a school-setting.
| Study | N | Age range, mean (SD), grade and gender | School/participant description (country) | Study design | Measures and domain | gHedges Baseline equivalence | gHedges Within-group | gHedge Differences in change scores | Reported findings according to authors | |
|---|---|---|---|---|---|---|---|---|---|---|
| RANDOMIZED CONTROLLED TRIALS | ||||||||||
| 1. Desmond and Hanich, 2010 | 40 | 11–12, 6th grade | Urban, public middle school, low income (USA) | M-group (n = 15) vs. C (n = 25) | BRIEF (teacher) | T | 0.26 | 0.04 | 0.31 | MANOVAs: No sig. time by group interaction (all ps > 0.05). Multiple regression analysis: Sig. interaction between pre-test score and group membership for predicting differences in one of eight subscales, indicating that M-group showed greater improvement in ability to shift (p < 0.05). In general, M-group maintained or improved executive function skills, while C shows a decline. |
| 41% female | ||||||||||
| 2. Flook et al., 2010 | 64 | 7–9 | On-campus university elementary school, diverse ethical backgrounds (USA) | M-group (n = 32) vs. C (n = 32) | BRIEF (teacher) | T | 0.31 | 0.20 | 0.08 | MANCOVAs with post-test scores as outcome variables: No sig. group main effect, indicating no group differences for pre- to post-test (p < 0.05). Sig. interaction between baseline levels and group in teacher report (p = 0.005) as well as in parent report (p = 0.020). In M-group, children with poorer initial executive function showed greater improvement at Time 2 compared to C. |
| 8.23 (0.66) 2nd + 3rd grade | BRIEF (parent) | T | 0.27 | 0.39 | 0.12 | |||||
| 55% female | ||||||||||
| 3. Franco Justo, 2009 | 60 | 15–18 | 3 public secondary schools (Spain) | M-group (n = 30) vs. waitlist c (n = 30), follow-up after 3 months | TTCT (verbal) | C | Independent and dependent t-Tests: Sig. improvement from pre- to post-test in M-group in all subscales (Fluency, Flexibility, Originality; all ps < 0.01) and no improvement in C (all ps > 0.05). At post-test M-group shows significantly higher scores in all subscales compared to C (all ps < 0.01). Effects sustained at follow up compared to pre-test (all ps = 0.001), but not compared to post-test (all ps > 0.05). | |||
| 17.3 | -Fluency | −0.11 | 1.50 | 1.48 | ||||||
| 1st + 2nd year high school | -Flexibility | 0.05 | 1.53 | 1.87 | ||||||
| 72% female | -Originality | −0.05 | 1.61 | 1.67 | ||||||
| 4. Franco Justo et al., 2011a | 61 | 16–18 | 3 compulsory secondary schools, public (Spain) | M-group (n = 31) vs. waitlist c (n = 30) Schools were allocated at random | Grades | C | −0.27 | 1.52 | 1.43 | Dependent and independent t-Tests: Sig. improvement from pre- to post-test in M-group in all measures (all ps = 0.001) and no improvement in C (all ps > 0.05). Sig. difference between groups in post-tests (all ps > 0.01). Detailed analysis: students with middle range academic performance show the most improvement in Grades (Cohen's d = 3.05), Students with low self-concept show most improvement in self-concept (d = 5.12), students with high state anxiety benefited the most on state anxiety (d = 1.95) and students with medium trait anxiety benefited the most on trait anxiety (d = 1.44). |
| 16.75 (0.83) | Self-concept | R | 0.59 | 1.55 | 1.84 | |||||
| 1st year high school | STAI | E | 0.35 | 0.62 | 0.11 | |||||
| 48% female | ||||||||||
| 5. Franco Justo et al., 2011b | 84 | 16–19 | Various compulsory secondary schools (Spain) | M-group (n = 42) vs. waitlist C (n = 42) | AURE | R | −0.06 | 1.26 | 1.29 | Dependent and independent t-Tests: Sig. improvement from pre- to post-test in M-group for all 3 subfactors (1. Approaching and Coping with a Task 2. Self-Concept and Self-Esteem 3. Empathy and Social Relations; all ps < 0.05) and no improvement in C (all ps > 0.05). Sig. difference between groups in post-tests in the first 2 subfactors (ps < 0.001), but not in the third (p = 0.16). |
| 17.06 (2.44) | ||||||||||
| 1st + 2nd year high school | ||||||||||
| 72% female | ||||||||||
| 6. Mai, 2010 | 12 | 13–17 | Urban high school, low socio economic status, low performing (USA) | M-group (n = 7) vs. waitlist C (n = 5), follow-up after 6 weeks | DERS | E | 0.57 | −0.06 | −0.60 | ANOVAs (repeated measures): No sig. findings were found (all ps > 0.05). |
| 14.4 | BRIC (teacher) | T | −0.12 | −0.10 | −0.10 | |||||
| (Mdn = 14.0), 9th grade, 25% female | Grades | C | −0.55 | 0.02 | 0.30 | |||||
| School attendance | – | −0.05 | 0.29 | 0.10 | ||||||
| 7. Mendelson et al., 2010 | 97 | 10.15 (0.7), 4th + 5th grade | 4 urban public elementary schools, low income neighborhood with high levels of violence (USA) | 2 M-groups (n = 42–47) vs. 2 waitlist C (n = 40–43) | PANAS | R | −0,14 | 0.17 | 0.23 | Multiple regressions: M-group demonstrated sig. improvements on the overall scale of Involuntary Engagement compared to C (p < 0.001). Sig. differences were found on three of the five subcales (Rumination, Emotional Arousal, Intrusive Thoughts: p < 0.05) and a trend for Impulsive Action and Physiologic Arousal (boths ps < 0.07). No other sig. results were found. However, depressive symptoms and negative effect displayed a pattern consistent with predictions. |
| 61% female | 4 schools were allocated at random | SMFQ—C | E | 0.9 | 0.14 | 0.02 | ||||
| PIML | R | −0.21 | −0.02 | 0.09 | ||||||
| Involuntary Engagement (RSQ) | S | 0 | 0.41 | 0.90 | ||||||
| 8. Napoli et al., 2005 | 194 | 1st-3rd grade | 2 elementary schools (USA) | M-group (n = 97) vs. C (n = 97) | ACTeRS (teacher) | T | # | 0.20# | 0.24 | T-Tests for change scores between groups: Sig. improvement for M-group on attention and social skills subcale of ACTeRS (both ps = 0.001). Sig. reduction of Test Anxiety in M-group (p = 0.007). Sig. improvement of M-group on selective attention (p < 0.001) but not on sustained attention subscale (p = 0.350). |
| TAS | E | # | 0.38# | 0.39 | ||||||
| Selective Attention (TEA-Ch) | C | # | 0.48# | 0.60 | ||||||
| Sustained Attention (TEA-Ch) | C | # | 0# | 0.13 | ||||||
| 9. Potek, 2012 | 30 | 14–17 | 2 high schools in an urban or rural setting, diverse range of socioeconomic status (USA) | M-group (n = 16) vs. waitlist C (n = 14) | MASC | E | 0.01 | 1.12 | 0.85 | Repeated-measures ANOVAs: Sig. interaction between time and group on MASC scores (p < 0.0001), indicating that the anxiety level of M-group decreased more compared to C. No sig interaction effect on DERS and PSS scores (boths ps = 0.14). |
| 15 (0.98) | DERS | E | 0.32 | 0.27 | 0.33 | |||||
| 9th-12th grade | PSS | S | 0.25 | 0.49 | 0.42 | |||||
| 48% female | ||||||||||
| 10. White, 2012 | 155 | 8–11 | Public schools, 85% reported having no family stress or health problems, majority of parents went to college (USA) | M-group (n = 70) vs. waitlist C (n = 85) | FBS | S | 0.16 | −0.17 | −0.11 | Repeated-measures ANOVAs: Sig. time by group interaction on the SCSI subscale frequency of coping (p < 0.04), suggesting that M-group is coping more frequently after intervention. No sig. interaction for Global self-worth (p = 0.57) and an approached significance for FBS (p = 0.06), indicating increasing stress levels in M-group after intervention compared to C. Further analysis revealed that this was due to a sig. interaction for the stress appraisal subscale of FBS (p = 0.005). Compared to C, M-group was more likely to increase their appraisal of stress at post-test. |
| 9.9 (0.72) | SCSI | S | −0.05 | 0.05 | 0.16 | |||||
| 4th + 5th grade | Global Self-worth Scale (SPPC) | R | 0 | 0.17 | −0.18 | |||||
| 100% female | ||||||||||
| QUASI-RANDOMIZED CONTROLLED TRIALS | ||||||||||
| 11. Broderick and Metz, 2009 | 122 | 16–19 | Suburban, private catholic high school for female (USA) | M-group (seniors, n = 105, age: M = 17.43) vs. C (juniors, n = 17, age: M = 16.41) | PANAS | R | −0.21 | 0.24 | 0.55 | T-Tests for change scores between groups: M-group demonstrated sig. reduction in neg. affect and sig increase on the calm/relaxed/self-accepting scale (both ps < 0.05). No other measures showed sig. differences in gain scores (p > 0.05). |
| M-group: Seniors 17.43 (0.53) | Calm/relaxed/self-accepting scale | R | 0.03 | 0.33 | 0.55 | Dependet t-tests: M-group showed sig. decline in neg. emotions and somatic complaints, sig. increase in the calm/relaxed/self-accepting scale and emotion regulation (all ps < 0.01). No sig. findings on the RRS factors (p > 0.05). | ||||
| C: Juniors 16.41 (0.85) | DERS | E | 0.13 | 0.20 | 0.18 | |||||
| 100% female | Reflective pondering (RRS) | E | 0.18 | 0.01 | 0.08 | |||||
| Moody pondering (RRS) | E | 0.09 | 0.19 | 0.22 | ||||||
| SICBC | E | 0.10 | 0.24 | 0.13 | ||||||
| 12. Corbett, 2011 | 107 | 8–11 | Elementary school located at university campus, (Florida, USA) | M-group (n = 63) vs. C (n = 44), cortisol measures: M-group (n = 12) vs. C (n = 13) | State Anxiety (STAIC) | E | 0.70 | # | 0# | ANCOVAs with pretest scores as covariates: No sig. differences between M-group and C in test anxiety, cortisol release, positive, and negative affect after the Mindfulness training (all ps > 0.05). |
| 9.94 (0.76) | TAS-C | E | 0.52 | 0.11 | −0.63 | ANOVA on STAIC difference scores showed no sig. difference between groups in level of reported state anxiety (p > 0.05). ANOVA on pop quiz scores demonstrated no sig. difference between groups (p > 0.05). | ||||
| 4th + 5th grade | PANAS-C | R | 0.37 | 0.07 | −0.43 | |||||
| 47% female | CCTT | C | −0.50 | 0.84 | 1.18 | |||||
| Pop quiz | – | −0.37 | 1.06 | −0.44 | ||||||
| Salivary cortisol | – | −0.74 | 0.02 | 0.14 | ||||||
| 13. Frenkel et al., in press | 47 | 13–15 | Private secondary school (Germany) | M-group (n = 24) vs. waitlist C (n = 23) Classes had been assigned randomly to conditions, follow up after 6 weeks. | Test d2 | C | 0.04 | 1.48 | −0.06 | MANOVAs: marginally sig improvement in combined parents ratings (p = 0.071) and measures of cognitive performance (p = 0.067). |
| 14.59 (0.54) | Unnoticed Mind Wandering | C | 0.20 | 0.13 | 0.15 | ANOVAs: M-group demonstrated sig. decrease in mind wandering noticed by others (p < 0. 05) which sustained in f –up (p < 0.10). Subjects in M-group were more likely not to notice their Mind Wandering (self-noticed Mind Wandering p < 0.10). | ||||
| 9th grade | Mind Wandering noticed by others | C | −0.86 | 0.84 | 1.26 | |||||
| 46% female | Self-noticed Mind Wandering | C | 0.11 | 0.35 | 0.38 | |||||
| PSQ | S | 0.42 | 0.22 | −0.12 | ||||||
| Kiddo-KINDL-R | R | −0.23 | 0.06 | −0.11 | ||||||
| PANAS | R | 0.03 | 0.11 | −0.18 | ||||||
| KINDL (parents) | T | 0.38 | 0.35 | −0.35 | ||||||
| 14. Hennelly, 2011 | 99 | 11–17 | 3 typical, mixed-gender state secondary schools (UK) | M-group (n = 53) vs. C (n = 46), follow-up after 6 months | WEMWBS | R | −0.11 | 0.19 | 0.41 | ANOVAs and pairwise comparisons by age, gender and group: Sig. effects on well-being due to decreasing scores of C, while participants scores remained steady (p < 0.05). In Ego-Resilience only the oldest students of M-group (12 Grade) reported sig. improvement (p < 0.05). Female participants ego-resilience increased compared to female controls whereas male participants ego-resilience reduced. At post-test, female participants scored sig. higher on ERS than male participants (p < 0.01). Compared to post-test, M-group showed a further increase of well-being and a slight decrease of ego-resilience at follow up. |
| 7th-12th grade | ERS | R | 0.53# | 0.04# | 0.08# | |||||
| 50% female | ||||||||||
| 15. Huppert and Johnson, 2010 | 134 | 14–15 | 2 independent, fee-paying boys schools, 5% ethnic minorities (UK) | M-group (n = 78) vs. C (n = 56) | ERS | R | −0.08 | 0 | 0 | Multiple regressions: no sig. overall differences between M-group and C for resilience (p < 0.05). Condition was found to contribute marginally significantly to change in well-being (p < 0.01). Sig. improvement of well-being related to the degree of individual practice (p < 0.05). |
| 100% male | WEMWBS | R | −0.09 | 0.26 | 0.34 | |||||
| 16. Metz et al., 2013 | 216 | 16,45 (0.95) | 2 high schools in a suburban district (USA) | M-school (n = 129) vs. C—school (n = 87) | DERS | E | −0.11 | 0.42 | 0.26 | MANOVA on mean gain scores: Sig. difference between groups (p = 0.003) and approximately 12% of multivariate variance of the dependent variable is associated/can be explained by with the group factor. |
| 10th-12th grade | Psychosomatic complaints | E | 0.03 | 0.37 | 0.20 | ANOVAs: compared to C, M-group demonstrated improvement in emotion regulation (p = 0.021), self-regulation efficacy (p = 0.001) and a lager reduction in psychosomatic complaints (p = 0.043). Sig. effect for several subscales of DERS and psychosomatic items (all ps < 0.05). M-group reported 10% decrease in amount of stress, whereas C stated no change (p = 0.005). | ||||
| 36% female | ASRES | R | −0.16 | 0.56 | 0.48 | |||||
| Stress level Item | S | 0.19 | 0.43 | 0.40 | ||||||
| 17. Kohls and Sauer, unpublished raw data | 87 | 9th–12th | Public secondary school | M-group (n = 29–31) vs. C (reading training: n = 24–26; passive: n = 22–30) | Attention test | C | −0.34 | 0.34 | 0.27 | Analysis of Effect sizes: M-Group demonstrated improvement in Attention compared to C. Well-being scores in M-group remained stable, whereas scores in C were decreasing. No difference between groups in vulnerability to stress and physical symptoms. In psychological symptoms, M-group proved the smallest increase. Compared to C, M-group showed strongest improvement in emotion regulation in response to stress. |
| 5th grade | (Germany) | KINDL | R | −0.19 | −0.02 | 0.47 | ||||
| Vulnerability (SSKJ) | S | −0.36 | 0.07 | −0.03 | ||||||
| Stress symptoms (SSKJ) | S | −0.32 | −0.33 | 0.02 | ||||||
| Emotion-Regulation Items (SSKJ) | S | 0.08 | 0.12 | 0.25 | ||||||
| 18. Schonert-Reichl and Lawlor, 2010 | 246 | 9–13 | 12 public elementary schools, | M-group (n = 139) vs. waitlist C (n = 107) | Optimism (RI) | R | # | 0.02# | 0.27# | ANCOVAs on change scores: M-group showed increase in optimism (p < 0. 05) and positive affect (p < 0.10), but no decrease in negative affect. No main effect for Group on the two self-concept subscales, but sig. interaction effect for Group and Age for general self-concept: Participants in grade 4 and 5 reported sig. improvement in general self-concept, whereas controls in this age showed sig. decreases. In contrast, M-group in grade 6 and 7 demonstrated sig. decrease in self-concept and students in control condition increased. |
| 11.43 (1.07) | 57% identified English as their first language, diverse range of socioeconomic status (Canada) | Teachers, instructing M in their classes had been assigned randomly | PANAS | R | # | 0.02# | 0.10# | ANCOVA on post-test scores: teacher ratings yielded an sig. intervention effect on total score in all subscales (all ps < 0.001). | ||
| 4th-7th grade | School self-concept (SD) | R | 0# | 0# | 0# | |||||
| 48% female | General self-concept (SD) | R | 0# | 0# | 0# | |||||
| TRSC (teacher) | T | # | 0.73# | 0.73#° | ||||||
| TWO ARMED COHORT STUDY | ||||||||||
| 19. Lau and Hue, 2011 | 48 | 14–16 | 2 Public schools for students with lower performance (Hong Kong) | M-group (n = 24) vs. C (n = 24) | SPWB | R | 0.25 | 0.44 | 0.52 | MANOVAs, ANOVAs and post-hoc tests: No sig. effect on well-being total score (p = 0.22), although M-group had significantly higher levels at personal growth dimension in post-test compared to C (p = 0.04). Sig. Time and Group interaction for combining depressive symptoms and perceived stress (p = 0.01). C's level of depression increased at post-test (p = 0.01), whereas in M-group there was no increase (p = 0.13). |
| DASS | E | −0.49 | 0.26 | 0.84 | ||||||
| PSS | S | −0.35 | 0.47 | 0.88 | ||||||
| NON-CONTROLLED TRIALS | ||||||||||
| 20. Anand and Sharma, in press | 33 | 14.23 | Public high school, middle socio-economic status, urban background (Bangalore, India) | Pre-post, follow-up after 3 months | SSS | S | — | 1.64 | — | ANOVAs: participants reported sig. reduction in perceived stress and sig. improvement in well-being from pre-test to post-test and from post-test to follow-up. Detailed analysis revealed sig changes in 5 of 7 subscales of SSS and in all of PWI-SC (no ps reported). |
| 46% female | PWI-SC | R | 1.51 | |||||||
| 21. Beauchemin et al., 2008 | 34 | 13–18 | Private residential high school specialized in serving students with learning disorder (Vermont, USA) | Pre-post | SSRS (student) | R | — | 0.53 | —– | T-tests: Students reported sig. reduction in state and trait anxiety, and sig. increase in social skills (all ps < 0.05). Sig. improvements emerged for teacher ratings on all 3 subscales (social skills, problem behavior, and academic performance; all ps < 0.05). |
| 16.16 | SSRS (teacher) | T | 0.74 | |||||||
| 29% female | STAI | E | 0.66 | |||||||
| 22. Biegel and Brown, 2010 | 79 | 6–8 | Elementary school (California, USA) | Pre-post, follow-up after 3 months | BEEDS | R | — | 0# | —– | ANOVAs and post-hoc tests: Sig. improvement in one aspect of attention (executive control; p < 0.01) form pre-test to post-test. Score stabilized from post-test to follow-up (p = 0.86). Sig. improvement in teacher rating of social skills from pre-test to post-test (p < 0.05), which stabilized at follow-up (p = 0.75). |
| 2nd + 3rd grade | Sense of Relatedness scale | R | 0# | No other results reported. | ||||||
| Altering (ANT-C) | C | 0# | ||||||||
| Orienting (ANT-C) | C | 0# | ||||||||
| Executive Control (ANT-C) | C | 0.41# | ||||||||
| SSRS (teacher) | T | 0.16# | ||||||||
| 23. Joyce et al., 2010 | 141 | 10–13 | 2 primary schools in Melbourne's outer suburbs (Australia) | Pre-post, sample size varied between Questionnaires | Total Difficulties (SDQ) | E | — | 0.26 | — | T-tests: Participants showed sig. reductions in total difficulties score of SDQ (p < 0.00). On the prosocial scale, only students with initially low scores demonstrated sig. enhancement (p < 0.05). Further, students proved sig. reductions in depression levels due to large changes in high-scoring individuals (p < 0.01). |
| 11.4 | CDI: 120; | Prosocial behavior (SDQ) | R | 0.15 | ||||||
| 5th + 6th grade | SDQ Diff.: 129; SDQ Prosoc.: 141 | CDI | E | 0.27 | ||||||
| 44% female | ||||||||||
| 24. Wisner, 2008 | 28 | 15–19 | Public alternative high school in a small city. | Pre-post | BERS-2/Teacher rating scale | T | — | 0.83 | — | T-tests: According to teacher ratings, students showed sig. improvement on behavioral and emotional functioning (p < 0.001). A sig. increase was also revealed in each subscale (all ps < 0.05). ANOVAs: No interaction effects on gender, grade level, and age. |
| 17.86 | At risk of dropping out of school. (USA) | |||||||||
| 10th-12th grade | ||||||||||
| 38% female | ||||||||||
#
Data essential for exact calculation of effect sizes were not provided. If possible we appraised effects based on information given, as graphs for example.
°
Teachers rated improvement form pre- to post-test after the training in M-group and Control. Between group differences were used to estimate within effect sizes as well as effect sizes of change scores.
SD, Standard deviation; M-group, Mindfulness-group; C, Control; RCT, Randomized controlled trial; ANOVA, Analysis of variance; ANCOVA, Analysis of covariance; MANOVA, Multivariate Analysis of Variance; MANCOVA, Multivariate analysis of covariance Domains: C, Cognitive Performance; E, Emotional Problems, R, Factors of Resilience; S, Perceived Stress and Coping; T, Third Person Rating. Measures: ACTeRS, ADD-H Comprehensive Teacher Rating Scale; ANT-C, Attention Network Test for Children; ASRES, Affective Self-Regulatory Efficacy Scale; AURE, Self-Concept and Self-Actualization Questionnaire; BEEDS, Behavioral and Emotional Engagement vs. Disaffection scale; BERS-2, Behavioral and Emotional Rating Scale; BRIC, Behavior Rating Index for Children; BRIEF, Behavior Rating Inventory of Executive Function; CCTT, Children's Color Trail Test; CDI, Children's Depression Inventory; DASS, Depression Anxiety Stress Scale; DERS, Difficulties in Emotion Regulation Scale; EP, Emotion Profile Inventory; ERS, Ego-Resiliency Scale; FBS, Feel Bad Scale; KINDL, QoL Questionnaire for Children and Adolescents; MASC, Multidimensional Anxiety Scale for Children; PANAS-C; Positive and Negative Affect Scale for Children; PIML, People in My Life; PSS, Perceived Stress Scale; PWI-SC; Personal Wellbeing Index—School Children; RRS, Ruminative Response Scale; RSQ, Responses to Stress Questionnaire; SCSI, Schoolagers' Coping Strategies Inventory; SD, Self-Description Questionnaire; SDQ, Strengths and Difficulties Questionnaire (Diff., difficulties subscales; Prosoc., prosocial behavior subscale); SICBC, Somatization Index of the Child Behavior Checklist; SMFQ-C, Short Mood and feelings Questionnaire—Child Version; SPPC, Self-Perception Profile for Children (Global Self-Worth Subscale); SPWB, Scales of Psychological Well-Being; SSKJ, Stress and Coping Questionnaire for Children and Adolescents; SSRS, Social Skills Rating System; SSS, School Situation Survey; STAIC, State-Trait Anxiety Inventory for Children; TASC, Test Anxiety Scale for Children; TEA-Ch, Test of Everyday Attention for Children; TIPI, Ten Item Personality Inventory; TTCT, Torrance Test of Creative Thinking; WEMWBS, Warwick-Edinburgh Mental Well-being Scale.