Safety¶

Learner Variable · All learner variables
Evidence · 90 claims (64 show a difference, 18 mixed, 8 no difference) · 52 studies (18 associational, 12 review, 7 causal, 6 qualitative, 4 quant-synthesis, 4 theoretical, 1 design), q1–q4 · 4 of 52 report an effect size · 90 claims rest on one study

Description¶

Safety here means both being and feeling physically and psychologically secure in the places a learner spends time: home, school or work, the neighbourhood and the wider community. It covers violence and crime, bullying and cyberbullying, and, from adolescence on, a learner's own risk-taking behaviour such as substance use. It matters for learning because a learner who feels safe can focus on the task, while one who feels unsafe tends to miss school and take part less in class, losing learning opportunities. For adults, the LVN adds that learners who had bad experiences of school need to feel safe enough to learn again. It is measured with self-report bullying and victimisation scales, neighbourhood questionnaires and screening tools for risky behaviour. It overlaps with adverse experiences, since feeling unsafe can be traumatic, and with belonging.

Implications¶

Context¶

School climate, anti-bullying practice and predictable routines bear directly on whether learners can attend and focus.
Adult programmes should allow for learners whose participation in education changes relationships at home.
Attendance and participation are the visible signs of a safety problem before achievement is.

Target Learners¶

Learners in unsafe neighbourhoods or homes, and learners who are bullied or cyberbullied.
Adolescents, whose own risk-taking becomes part of the picture.
Adult learners re-entering education after negative school experiences.

Target Learning Objectives¶

Attendance, participation and engagement.
Focus on learning, and through it achievement in reading, writing and mathematics.

Claims¶

A more positive school climate is related to improved academic achievement beyond socioeconomic status [+W] — learners who differ on it differ in outcomes
A positive school climate improves academic achievement and reduces the negative effects of poverty on achievement [+M] — learners who differ on it differ in outcomes
A regional collective initiative (Teaching is My Favorite Color) provided teachers of color a safe space, belonging, and authentic identity expression [+W] — instruction changes it
AI Prompting Literacy is indirectly and significantly associated with Deep Revision Engagement through Perceived Competence, Intrinsic Motivation, and Psychological Safety [+M] — learners who differ on it differ in outcomes
AI Prompting Literacy significantly positively predicts Perceived Competence, Intrinsic Motivation, and Psychological Safety among Chinese university EFL learners [+W] — instruction changes it
An open classroom climate has a positive impact on adolescents' civic knowledge and appreciation of political conflict, controlling for individual, classroom, school, and district characteristics [+M] — learners who differ on it differ in outcomes
Beating-the-odds schools have substantially more positive school climate than the other California middle and high schools in a 1,715-school survey sample [+M] — learners who differ on it differ in outcomes
Black and Latinx students generally feel less safe at school than their White peers [+M] — instruction changes it
Common physical and exclusionary safety measures can cause harm to the very students schools are trying to protect [+M] — instruction changes it
Community members reported that participating in a Community Socratic Circle helped them feel comfortable with teachers facilitating classroom racial discourse [+M] — instruction changes it
Contextual themes (extracurricular activities, environmental/school safety) are associated with school belonging [+M] — learners who differ on it differ in outcomes
Effectiveness of anti-bullying programs varies significantly across evaluations (heterogeneity) [~M] — instruction changes it
Equalizing perceived supports and safety would reduce gay/lesbian/bisexual versus straight disparities in connectedness by 90-100 percent and other disparities by half [+M] — learners who differ on it differ in outcomes
GenAI can foster psychologically safe team interaction while risking reduced human interaction over time [~M] — instruction changes it
Honours community members perceive a safe and supportive environment that facilitates exploring new ideas [+W] — learners who differ on it differ in outcomes
Large majorities of Gateway students report feeling cared for, part of the school, safe, and connected [+W] — instruction changes it
Lockdown training and drill participation improved students' knowledge and perceptions of preparedness without increasing anxiety, per the studies the brief reviews [+M] — instruction changes it
More intensive anti-bullying programs are more effective [+M] — instruction changes it
Most students, even in bottom quartile schools, report feeling mostly safe at school [+M] — instruction changes it
Multiple dimensions of school climate were associated with lower school absences, with stronger relationships after the pandemic than before [~W] — learners who differ on it differ in outcomes
No association found between 2021/22 school climate scores and the amount of virtual and hybrid instruction in 2020/21 [-W] — instruction changes it
NoTrap! was the most effective program for reducing bullying victimization, while ViSC and fairplayer.manual showed non-significant negative effects [~M] — instruction changes it
Parent meetings, firm disciplinary methods, and improved playground supervision are program elements associated with decreases in bullying and victimization [+M] — instruction changes it
Perceived Competence, Intrinsic Motivation, and Psychological Safety each significantly positively predict Deep Revision Engagement [+W] — learners who differ on it differ in outcomes
Perceived safety changes under reduced suspension use varied by school racial composition: a small decline in predominantly Latino schools and a similar-sized but non-significant improvement in predominantly Black schools [~M] — instruction changes it
Playworks improved teachers' perceptions of students' safety and feeling included during recess in a randomized experiment [+M] — instruction changes it
Quality of staff-student and staff-parent relationships most strongly defines safe schools, with disadvantaged high-relationship schools feeling safer than advantaged low-relationship schools [+M] — instruction changes it
RJ-based responses to bullying may help, but require trained adult facilitators and may leave victims uncomfortable facing bullies [~W] — instruction changes it
SBLE shows no statistically significant relationship with perceptions of school, but may improve perceived safety in school-level studies only [~M] — instruction changes it
School climate differences between BTO and chronically underperforming schools are twice as large as differences between BTO and other schools [+M] — learners who differ on it differ in outcomes
School climate differences between BTO and other schools remain large and significant after adjusting for student demographics and personnel resources [+M] — learners who differ on it differ in outcomes
School climate functions as a setting-level indicator: students entering schools where they feel safe and supported attend more and earn higher grades, and Chicago's five essential supports survey responses are highly predictive of teacher and school performance [+M] — learners who differ on it differ in outcomes
School climate is more strongly associated with the likelihood of beating the odds than student demographics or personnel resources [+M] — learners who differ on it differ in outcomes
School-based anti-bullying programs are effective overall, decreasing bullying by 20-23% on average [+M] — instruction changes it
School-based anti-bullying programs decrease victimization by 17-20% on average [+W] — instruction changes it
School-based anti-bullying programs reduce bullying perpetration and victimization globally [+W] — instruction changes it
School-based law enforcement is not associated with statistically significant changes in students' perceptions of safety at school [-M] — instruction changes it
Schools that increased restorative practice utilization saw declines in misbehavior, gang membership, victimization, depressive symptoms, and substance abuse, and improvements in GPA and climate [+M] — instruction changes it
Schools where students report feeling more safe, supported, and challenged show stronger academic and social-emotional growth, and student reports predict future school improvement [+M] — learners who differ on it differ in outcomes
Schools with higher 5Essentials Survey scores, especially more Supportive Environments, are more likely to reduce students' fear of bullying [+M] — instruction changes it
Security cameras show no evidence of improving school safety, and evidence does not support expanding metal detector use [-W] — instruction changes it
Student perceptions of safety and teacher perceptions of order improved in CPS high schools during the same period that OSS rates declined [+M] — instruction changes it
Student reports of safety did not improve faster in Culture of Calm schools than in other similar high schools [-M] — instruction changes it
Student-reported safety and classroom behavior are more positive in top-quartile schools, and students feel less safe outside school than inside [+M] — instruction changes it
Students feel safest in classrooms and least safe in areas just outside the school, with safety tracking the level of adult supervision [+W] — instruction changes it
Students report feeling less safe and more likely to perceive violence and disorder in schools with metal detectors [+W] — instruction changes it
Students report fewer violent incidents and a personalized, supportive environment in CHSRI small schools [+M] — instruction changes it
Teacher-rated bullying and peer rejection rose over four years of elementary school in both SWPBIS and comparison schools [-M] — instruction changes it
The AI's nonjudgmental conversational tone and adaptive prompting fostered risk-taking and productive struggle in a psychologically safe environment [+M] — instruction changes it
The effects of attending a bullying-reducing school on graduation are larger for students with a history of victimization [+M] — an instructional effect differs with it
The Olweus Bullying Prevention Program produced the largest effect sizes for bullying perpetration, with effectiveness varying between Norway and USA evaluations [~M] — instruction changes it
The school climate gap between beating-the-odds and chronically underperforming schools is more than twice as large as the gap between BTO and other schools [+M] — learners who differ on it differ in outcomes
There is no evidence that arming school staff improves school safety, and one study found an armed guard present was associated with more deaths in school shootings [~W] — instruction changes it
Work with peers in anti-bullying programs is associated with an increase in victimization [+M] — instruction changes it
Bullied children with NSSI or suicidality show greater behavioral problems and poorer grades than bullied children without these behaviors [+W] — learners who differ on it differ in outcomes
Bullied girls show lower episodic memory and crystallized composite cognitive scores than non-bullied girls [+W] — learners who differ on it differ in outcomes
Bullied preadolescent children have lower cognitive scores than non-bullied peers [+W] — learners who differ on it differ in outcomes
Bullied preadolescent children have poorer academic grades than non-bullied peers [+W] — learners who differ on it differ in outcomes
Bullied preadolescent children show greater behavioral problems across CBCL domains than non-bullied peers [+W] — learners who differ on it differ in outcomes
Caregiver-reported bullying victimization is associated with higher odds of NSSI/passive suicidality and active suicidality in preadolescent children [+W] — learners who differ on it differ in outcomes
In very safe neighborhoods, curiosity does not differ between lower-resourced and higher-resourced children, while a gap appears in fairly safe neighborhoods [~W] — learners who differ on it differ in outcomes
Living in a very safe neighborhood is associated with higher kindergarten curiosity, but only for lower- and middle-resourced children [~W] — learners who differ on it differ in outcomes
Peer-problem mediation of the ADHD–depression link is not driven entirely by victimisation [~W] — learners who differ on it differ in outcomes
Among youth from low-income families, perceived neighborhood safety is negatively associated with GPA, whereas among non-low-income youth no association is observed [~W] — learners who differ on it differ in outcomes
Bullied adolescents are about half as likely to reach the national academic achievement benchmark for their age group two years later [+W] — learners who differ on it differ in outcomes
Bullying predicts depressive symptoms two years later for boys but not for girls [~W] — learners who differ on it differ in outcomes
High levels of support from friends protect bullied adolescents from poor academic achievement, but moderate or low friend support does not [~W] — learners who differ on it differ in outcomes
Observer-rated physical neighborhood disorder is negatively associated with GPA, with each standard deviation of disorder linked to a 20.7% of a standard deviation decrease [+W] — learners who differ on it differ in outcomes
Perceived neighborhood safety is not significantly associated with standardized test scores or GPA once confounding variables are controlled [-W] — learners who differ on it differ in outcomes
Perceived neighborhood safety is positively associated with school attachment among 15-year-olds after controlling for individual, family, school, and external neighborhood factors [+W] — learners who differ on it differ in outcomes
Perceived neighborhood safety is protective for school attachment among low-income youth in high-disorder neighborhoods, and mediates the affluence-school attachment link only among non-low-income youth [~W] — learners who differ on it differ in outcomes
The association between perceived neighborhood safety and GPA is moderated by observer-rated neighborhood disorder and by neighborhood affluence [~W] — learners who differ on it differ in outcomes
High-tech VAD equipment made some participants feel under observation and stressed, and viewing recorded mistakes before teammates was seen as critical in both VAD forms, a potential threat to psychological safety [+M] — instruction changes it
Qualitative research shows resource fit—safety and financial resources—shapes marginalized students' sense of belonging [+M] — learners who differ on it differ in outcomes
School racial climate subscales showed positive relationships with motivational constructs including interest, belonging, and utility value [+M] — learners who differ on it differ in outcomes
Corrective feedback plus either GLA significantly improved students' safety behavior in VR, with no difference between re-experiencing and self-explaining [+M] — instruction changes it
The Pro-Bully Paradox: victimization relates to weaker anti-bullying attitudes, an effect largely specific to physical victimization [~M] — learners who differ on it differ in outcomes
Well-being is more strongly related to verbal and particularly relational victimization than to physical victimization [+M] — learners who differ on it differ in outcomes
Students' implementation of self-assessment is influenced by perceived usefulness, affective attitude, self-efficacy, important others, and psychological safety [+M] — learners who differ on it differ in outcomes
Cooperative learning is associated with greater peer acceptance and cross-group friendships and has been shown to reduce bullying and perceived stress among marginalized students [+M] — instruction changes it
Identity- or bias-based harassment shows negative social-emotional and academic consequences over and above general victimization [+M] — learners who differ on it differ in outcomes
Peer victimization is associated with more school absences and lower grades and achievement among school-aged students [+M] — learners who differ on it differ in outcomes
Low environment-change beliefs lead students to allocate cognitive resources to monitoring for breakdowns, creating extraneous load [+W] — learners who differ on it differ in outcomes
Higher levels of and stronger increases in perceived negative relationships with teachers are associated with lower levels of and stronger decreases in autonomous academic motivation across the secondary school transition [+M] — learners who differ on it differ in outcomes
Peer victimization shows no statistically significant association with the level or change of either controlled or autonomous academic motivation across the secondary school transition [-M] — learners who differ on it differ in outcomes
Cooperative learning reduced emotional problems and increased peer relatedness for all students, not only marginalized students [+M] — instruction changes it
Cooperative learning showed no significant main effect on bullying across the full sample [-M] — instruction changes it
Cooperative learning significantly reduced bullying among marginalized (low-engagement) middle school students in a cluster randomized trial [+M] — instruction changes it
Cooperative learning significantly reduced perceived stress among marginalized middle school students [+M] — instruction changes it
Cooperative learning significantly reduced victimization among marginalized middle school students [+M] — instruction changes it

Measures¶

Electronic Bullying Questionnaire (Kowalski & Limber, 2007): self-report of how often a person has been cyberbullied or has cyberbullied others.
Alcohol Use Disorders Identification Test (World Health Organization): self-report screening of drinking behaviour and alcohol-related problems.
Peer Victimization Scale (Austin & Joseph, 1996): assesses being bullied at school for children aged 8 to 11.
Gatehouse Bullying Scale (Bond et al., 2007): assesses bullying victimisation at school for ages 10 to 15.
Neighborhood Environment Walkability Scale (NEWS) (Saelens et al., 2003): residents' perceptions of neighbourhood features including safety, traffic hazards and crime.
Neighborhood Questionnaire (Greenberg & Lengua, 1995): a parent's view of neighbourhood safety, public services and violent crime.

Adverse Experiences: feeling unsafe can be traumatic, and trauma lowers felt safety.
Affect Regulation: bullying and victimisation bring more negative emotion and mental health strain.
Belonging: a stronger sense of belonging goes with less risk-taking.
Socioeconomic Status: low-income households face higher rates of violence.
Access: learners with hearing or vision loss can meet safety risks in getting around.
Physical Well-Being: unsafe neighbourhoods limit outdoor play and exercise, and bullying goes with poorer health.
Self-Regulation: impulsivity underlies some adolescent risk-taking.
Social Supports: support networks protect against bullying, though for some adults study can strain them.
Sleep: unsafe living conditions go with more disturbed sleep.
Attention: intoxication, one form of risk-taking, narrows what is attended to.
Social Awareness and Relationship Skills: peers, including online, can draw young people into risk-taking.
Executive Function: heavy alcohol use is associated with weaker inhibition.

Examples¶

Key Sources¶

Digital Promise. Learner Variability Navigator: Safety (Adult Learner, Building Literacy 4-6, Building Literacy 7-12, Building Literacy PK-3, Building Math 3-6, Building Math 7-10, Building Math PK-2, Portrait of a Learner 4-8, Portrait of a Learner 9-12, Portrait of a Learner PK-3). lvp.digitalpromiseglobal.org. Licensed CC BY-NC 4.0.

Page metadata

Field	Value
`type`	`learner-variable`
`id`	`safety`
`title`	`Safety`
`description`	`Being, and feeling, physically and psychologically secure at home, at school or work, and in the neighbourhood and wider community.`
`status`	`review`
`canonical`	`True`
`generated`	by `claude/unspecified` on `2026-10-16`
`trust`	unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.