Dutch trauma-informed schools show early teacher gains without clear causal effect

Teachers at eight Dutch schools reported modestly higher confidence and more positive classroom conditions during the first year their schools implemented a school-wide trauma-informed approach, but a two-year mixed-methods study could not establish that the program caused those changes. The research, published in Frontiers in Education, followed 110 teachers across two mainstream primary schools and six primary and secondary special education schools. Teachers’ perceived trauma knowledge rose immediately after initial staff training, while end-of-year ratings of class atmosphere and relationships among students were higher after implementation began. However, none of the estimated differences between baseline-year and implementation-year trends for self-efficacy or classroom climate reached statistical significance, limiting causal interpretation.

The Dutch program, called Trauma-sensitive Education, was designed to change how an entire school team understands and responds to trauma. It included a train-the-trainer course, a book for all school staff, a 12-hour training program and a manual for school-wide implementation. The training addressed how trauma may affect students’ physical, behavioral, emotional and academic functioning, as well as stable relationships, self-regulation, collaboration, staff self-care and implementation of trauma-informed policies and procedures. The approach was not simply an awareness session for individual teachers: schools also developed action plans adapted to their own staff and student needs.

Quantitative data came from two cohorts whose data-collection periods overlapped across the 2018–2020 and 2019–2021 academic years. Teachers completed questionnaires at three waves in each year, covering the beginning, middle and end of the academic year. The first year served as the baseline, while schools introduced the approach during the second year. Investigators used piecewise latent growth curve models to estimate both levels of change and rates of change within each year, rather than relying on a single before-and-after comparison. A total of 118 teachers consented, although eight did not complete the measures concerning themselves, leaving 110 in the analytical sample. Of these, 17 worked in mainstream primary education, 22 in primary special education and 71 in secondary special education.

Perceived trauma knowledge was assessed only immediately before and after the opening training modules, which combined theoretical content with practice. On a seven-point scale, the mean score increased from 4.90 before training to 5.21 after the initial modules. The paired test produced a small standardized effect, d = 0.33, with t(43) = 2.22 and p = .03. The instrument asked teachers to rate their perceived ability to recognize trauma-related manifestations and respond appropriately, so it measured confidence in knowledge and skills rather than objectively demonstrated competence. It was not measured again during the implementation year. The result therefore documents an immediate response to training, not evidence that the increase was sustained or that teachers consistently applied what they had learned.

Teachers’ self-efficacy and classroom climate were assessed repeatedly. The self-efficacy scale covered adapting instruction, maintaining discipline, motivating students, general instruction, coping with change and challenges, and cooperating with parents and colleagues. During the implementation year, researchers found significant increases in overall self-efficacy, efficacy in coping with change, and cooperation with parents and colleagues. Classroom ratings also showed significant implementation-year increases in atmosphere, student relationships and classroom order. At the end of the year, teachers rated class atmosphere as 0.122 points higher on the four-point scale and student relationships as 0.097 points higher than at the end of the baseline year. The researchers described the statistically significant changes as small in standardized magnitude.

The temporal pattern was not uniform. Class atmosphere, teacher-student alliance and classroom order were already improving during the baseline year, before implementation began. Motivating students and cooperating with parents and colleagues also showed significant baseline-year increases in self-efficacy. Student relationships produced the clearest implementation-linked pattern: they were the only classroom-climate measure that did not improve significantly during the baseline year but did increase during the implementation year, and their end-of-year scores were significantly higher after implementation began. Even so, formal comparisons of baseline and implementation slopes found no significant differences for self-efficacy or classroom climate. The findings therefore cannot establish whether the approach produced the observed gains or whether they reflected pre-existing trends or other concurrent influences.

Five focus groups involving 18 staff members—12 teachers and six support staff members—provided qualitative context at the end of the first implementation year. Participants described greater awareness of how trauma might shape classroom behavior, a shift toward interpreting behavior through what the training called a “trauma lens,” and increased confidence when responding to difficult situations. They also described calmer classes, students requesting time-outs to manage distress, stronger peer consultation and greater attention to collective and individual self-care. Some schools introduced staff check-ins and debriefings, peer consultation and supervision, revised student support plans, and classroom changes intended to make environments calmer and safer. A recurring theme was that awareness did not automatically produce consistent practice: staff still found it difficult to translate the principles into everyday interactions and shared procedures.

Implementation took place during the COVID-19 pandemic, which affected the format and timing of training and other school activities. Two schools replaced face-to-face sessions with online training, three delayed most modules until the latter part of the implementation year, and two had not completed all eight modules by the final assessment. Staff experienced illness, bereavement and other pressures that interrupted self-care and implementation activities. The study did not isolate the effects of the pandemic, so it could not determine which outcomes it affected or how strongly it influenced the findings. The researchers also noted that stress can hinder adoption of new practices, making the disrupted first year an important context for interpreting the results.

Further limitations include the small number of schools, uneven representation across educational settings, reliance on teacher self-report and the absence of a comparison group. Cluster-robust statistical methods accounted for teachers being nested within eight schools, but the researchers cautioned that cluster-robust standard errors can be downward biased when based on a small number of clusters, potentially overstating statistical significance. Model fit was mixed for several outcomes, and the classroom-discipline model performed poorly on two fit indices. No formal correction was made for multiple statistical tests, increasing the possibility that isolated significant findings reflect chance. The qualitative sample may also have overrepresented staff willing or able to attend, and implementation fidelity was not measured.

The most defensible conclusion is therefore promise rather than proof. The first implementation year was associated with an immediate rise in teachers’ perceived knowledge, favorable end-of-year ratings in selected measures, and staff reports of changed interpretations and classroom practices. These findings describe what teachers perceived and how they described changing their work, not whether the trauma-informed approach alone caused the changes. The sample was predominantly drawn from special education schools serving students with severe emotional and behavioral difficulties, which limits generalization to mainstream education and other student populations. Longer studies with comparison groups, repeated trauma-knowledge measurements, objective knowledge assessments, classroom observations and direct assessment of implementation fidelity are needed to determine whether the reported developments persist and can be credibly linked to trauma-informed schooling.

Subject of Research: Teacher-reported outcomes during the first year of a school-wide trauma-informed approach in eight Dutch primary and special education schools

Article Title: Evaluation of a school-wide trauma-informed approach in primary and secondary schools in the Netherlands: teacher outcomes in the first year of implementation

Article References: Nelen, W. B. L., Wassink-de Stigter, R., Homan, E., de Berk, A., Delsing, M., Kooijmans, R., Asselman, M., Offerman, E., Lindauer, R., & Helmond, P. (2026). Evaluation of a school-wide trauma-informed approach in primary and secondary schools in the Netherlands: teacher outcomes in the first year of implementation. Frontiers in Education, 11, Article 1932980. https://doi.org/10.3389/feduc.2026.1932980

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1932980

Keywords: trauma-informed education, teacher outcomes, teacher self-efficacy, classroom climate, trauma knowledge, school implementation, special education, Netherlands, mixed-methods study, longitudinal research, Evaluation, school-wide