Zum Inhalt springen
unterHUNDs Science & Practice Initiative für gewaltfreies Hundetraining

Research

Dogs in the Classroom: What Studies Show About School Dogs

Michael Sauerwein

A golden retriever lies relaxed on the carpet of a classroom next to three children sitting on the floor reading books, while two children write at their desks in the background.

School dogs and reading dogs are popular, and the claims made for them are broad: calmer classes, less stress, more motivation, better reading. Individual studies do report lower cortisol in children, better adherence to instructions in preschool tasks and more positive attitudes toward school (Beetz et al., 2012; Gee et al., 2009; Beetz, 2013; Meints et al., 2022). At the same time, the most recent systematic review, which covered animal-assisted services with different species and settings rather than school dog programs alone, rated 17 of 21 randomized trials (81%) as being at high risk of bias, and a randomized study with 201 children found no demonstrated additional benefit of a dog-assisted program over the control group on the primary measures (Moe et al., 2025; Nagy et al., 2026).

This article reviews what studies show about stress, behavior and classroom climate, social-emotional outcomes, attention and executive tasks, and reading, and what reviews and meta-analyses conclude. It explains the methodological weaknesses that limit these findings, summarizes what is known about safety, hygiene and allergens, and points to the dog's side of school work.

1. Why the Question Matters

1.1 School Dogs, Reading Dogs and Visiting Teams

"School dog" covers very different arrangements. In one model, studied for example in a German elementary school, the teacher's own dog accompanies its owner to class on fixed days and becomes part of everyday school life (Beetz, 2013). In other programs, volunteer handlers visit a school with their dog for short, structured sessions with individual children or small groups (Meints et al., 2022). Reading-to-dog programs are a third format: a child reads aloud to a dog, with the handler nearby, for a limited number of sessions.

These formats differ in dose, in who is present and in how familiar the dog is to the children. A finding from a twenty-minute visiting session cannot simply be transferred to a dog that spends a morning with 25 children, or the reverse. The format of each study is therefore stated below.

1.2 What Kind of Evidence Exists

The research is young and methodologically uneven. It includes laboratory stress tests, short preschool experiments, comparisons of one dog class with one control class, interview studies, surveys and a small number of randomized controlled trials (RCTs) in schools, plus several systematic reviews and two meta-analyses (Brelsford et al., 2017; Reilly et al., 2020; Dimolareva & Dunn, 2021; Meixner & Kotrschal, 2022; Wintermantel et al., 2024; Moe et al., 2025).

Two questions need to be kept apart. The first is whether children benefit at all, and on which outcomes. The second is whether the benefit is due to the dog, or whether an additional friendly adult, a break from lessons, a relaxation exercise or the excitement of something new would do the same. Many studies answer the first question tentatively and the second not at all.

1.3 How Effects Are Thought to Work

A review of human-animal interaction research in schools proposed that animals affect learning mainly indirectly, through motivation, engagement, self-regulation and social interaction, and that benefits are possible "under specific conditions, with proper safeguards" (Gee et al., 2017). Competing theories range from social facilitation and social support to biophilia and supernormal stimuli, but few studies were designed to test one against another (Serpell et al., 2017). The same authors pointed to practical challenges in design, in the heterogeneity of children and animals, and in animal welfare. Kazdin (2017) recommended more well-designed studies, explicit "small theories" of mechanism, laboratory studies of processes and a strategic research plan. In schools, social buffering through touch, stress reduction by a calm animal and the motivating effect of novelty all remain plausible and have rarely been separated (what social buffering means and how it is studied).

2. Stress and Cortisol

2.1 A Laboratory Stress Test with Insecurely Attached Boys

The most frequently cited stress studies did not take place in classrooms. In an exploratory study, 31 boys aged 7 to 12 with insecure or disorganized attachment completed the Trier Social Stress Test for Children, a standardized public speaking and arithmetic task, supported by a friendly dog (11 boys), a friendly human (11) or a toy dog (9) (Beetz, Kotrschal, et al., 2011). Salivary cortisol was measured five times. Self-reported stress did not differ between groups, but the cortisol response was significantly lower with the real dog, and the more a boy stroked the dog, the smaller his stress response (rs = −0.818).

A follow-up with the same paradigm included 47 boys aged 7 to 11 with insecure-avoidant or disorganized attachment: 24 with a therapy dog, 10 with a friendly human and 13 with a toy dog (Beetz et al., 2012). Cortisol was again significantly lower with the dog and correlated negatively with the amount of physical contact. The authors argued that this supports animal-assisted education "since stress interferes with learning and performance".

Both studies tested a laboratory stressor, not a school day, included only boys with a specific attachment profile and had small, unequal control cells. The papers do not state whether the second sample includes the first, so they should not be read as independent replications. The stroking correlation does not show direction: less stressed boys may simply have touched the dog more (what research shows about touch and petting).

2.2 A School-Based Randomized Trial

The most rigorous school test comes from a research program in Lincolnshire and Gloucestershire, England (Meints et al., 2022). Children in four mainstream and seven special-needs schools were randomly allocated, stratified by dog ownership, socioeconomic status and ability, to a dog-assisted intervention, a relaxation intervention or a no-treatment control. Sessions ran twice a week for 20 minutes over four weeks, individually or in small groups. Baseline cortisol was collected at the start and end of a term of about six weeks, and in mainstream children also before and after sessions.

The mainstream study included 105 children aged 8 to 9, 90 with complete baseline data; the special educational needs (SEN) study had baseline data for 44 children, most of them boys. Diagnoses in the SEN group included autism, ADHD, both together, Down syndrome and other learning difficulties. Twenty-three dogs and 21 volunteer handlers, mostly from a national therapy-dog charity, took part. Children met the dogs for about 30 minutes in small groups in the week before, explicitly to reduce novelty effects. Contact during sessions was child-led, and no child was made to touch a dog.

In mainstream schools, baseline cortisol rose significantly over the term in the control group and also rose in the relaxation group, but not in the dog groups, and cortisol fell acutely after dog sessions. In SEN schools, cortisol decreased after dog group sessions, with no change in the relaxation or control groups. The authors called this "crucial evidence that dog interventions can successfully attenuate stress levels in school children".

The design is the strongest in the field, but the SEN cohort was smaller than planned, the authors note that effect sizes were small in some conditions, SEN subgroups could not be analyzed separately, and cortisol was the only outcome in this paper, with no behavioral or academic measure. Nobody was blind to condition, and the dog arm added a novel adult, the handler. The lead author was funded for several years by the WALTHAM Centre for Pet Nutrition, which according to the paper had no role in the study.

2.3 Brief Interactions in High School

After a ten-minute interaction with a certified therapy dog, high school students reported lower perceived stress and higher self-efficacy, and their cortisol was lower (Melzer et al., 2026). There was no control condition. The authors described their results as "preliminary associations rather than evidence of a causal effect" and called for controlled studies separating the dog from "social interaction or time away from academic demands".

2.4 How Cortisol Is Measured in Schools

A systematic review of 25 articles found cortisol measurement in classrooms feasible, recommended passive drool or cotton swabs, and noted discrepancies in collection protocols and in reporting how many samples succeeded (Dimolareva et al., 2018). Cortisol also varies with time of day, food intake and physical activity, so a shifted sampling time or a missed sample can matter. Timing and the share of usable samples belong to any evaluation of cortisol results, in children as in dogs (how stress is measured in dogs).

2.5 Reading the Stress Studies Together

The stress studies point the same way: with a dog present, children's cortisol response to a stressor or over a term was lower than in some comparison conditions (Beetz, Kotrschal, et al., 2011; Beetz et al., 2012; Meints et al., 2022). This is the best-supported outcome in the reviews (Meixner & Kotrschal, 2022). But self-reported stress did not differ in the laboratory, and none of these studies showed that lower cortisol translated into better learning or behavior.

3. Behavior and Classroom Climate

3.1 A Viennese First-Grade Class

An early study followed one elementary class in Vienna with 24 children, mean age 6.7 years, most from first-generation immigrant families (Kotrschal & Ortbauer, 2003). After a one-month control period, one of three dogs was present during a period of similar length, and two hours of open teaching per week were video-coded for the frequency and duration of all observable behaviors. Children differed widely in their interest in the dog. With a dog present, the group became socially more homogeneous, with fewer behavioral extremes such as aggressiveness and hyperactivity; withdrawn children were better integrated, effects were stronger in boys, and children paid more attention to the teacher. The authors concluded that a dog could "positively stimulate social cohesion" and was a "relatively cheap and easy means of improving teaching conditions".

The study offers detailed behavioral observation, but it rests on one class without a parallel control class. The dog period always came second, so children also had more time to get used to each other and to school, and observers could see the dog on the videos and therefore knew the condition.

3.2 Two First-Grade Classes in Comparison

A second publication from the same Vienna school project compared two first-grade classes with 46 children, 43 with an immigrant background (Hergovich et al., 2002). A dog was present in one class for three months. Compared with the control class, the dog class improved significantly in field independence, a perceptual-cognitive measure, and in empathy with animals, and teachers reported higher social integration and fewer aggressive children. With one class per condition, the effective number of compared units is two classes, not 46 children, and the rating teachers knew which class had the dog. The 24 children of the dog class are the same as in Kotrschal and Ortbauer (2003), so the two papers are not independent replications.

3.3 A German Third-Grade Class Over a School Year

A German study compared one third-grade class with a school dog present one day a week over a school year (25 children) with one control class (21 children) (Beetz, 2013). Children completed questionnaires on depressive symptoms, emotion regulation and socio-emotional school experiences in the third week of the school year and two weeks before its end. The dog class was chosen partly because none of its children had allergies. There was no effect on depressive symptoms. The classes differed significantly in how positive attitude toward school (partial η² = 0.216) and positive emotions related to learning developed.

The direction matters. Positive attitude toward school rose slightly in the dog class, from 29.1 to 30.1 points, but fell in the control class from 28.6 to 23.7; positive emotions related to learning rose from 26.7 to 29.3 and fell from 28.0 to 24.2. The effect was largely a decline in the control class, not a large rise in the dog class, and the classes already differed in class climate at the start. The author noted that the effect cannot be separated from the specific teacher and dog team and that generalization is "as limited as in most studies on this topic". Any difference in teacher, class composition or timetable is confounded with the dog.

3.4 Interviews in Secondary Schools

A German qualitative study interviewed 30 students in grades 5 to 8, from four classes in two secondary schools and one grammar school, before and after a term in which a school dog was present in selected subjects (Mombeck & Albers, 2024). Students described an improved social climate, more integration, changed social roles and a friendlier perception of each other and the teacher. Such interviews show how students experience a school dog, which is valuable for program design, but without a comparison group they cannot show that the dog caused the changes.

3.5 Teacher and Handler Ratings

An exploratory mixed-methods study with students aged 8 to 13 reported higher self-rated self-confidence and peer relationships and higher teacher-rated work attitude, emotional stability and social behavior after the intervention (Verhoeven et al., 2023). The study was non-randomized (quasi-experimental) and analyzed 22 children in the dog-assisted program and 23 comparison children; the reported changes cannot be attributed specifically to the dog because no otherwise equivalent dog-free program was tested. Teacher and handler reports from 58 schools in the United Kingdom using a dog mentoring program for autistic children described a calmer environment, more engagement and better emotion regulation (Dimolareva et al., 2026). These 58 voluntary responses came from a program with over 600 registered schools, and because the initial questionnaire offered only positive outcome options, selection and response bias are possible. The findings reflect the perceptions of those running the program, not independently measured effects, and the authors called for quantitative data.

How much ratings depend on the rater is shown by a pilot in a class with bullying episodes (Correale et al., 2017). After five encounters with pet dogs, the teacher's ratings on a standard behavior questionnaire changed in a way the authors interpreted as the teacher having become more aware of the students' difficulties. The observer changed, not necessarily the children.

In practice: Treat a school dog as a planned classroom tool rather than as a mascot. Decide in advance in which lessons the dog is present, what its role is and how it leaves the room again, instead of bringing the dog along and seeing what happens. This is a practice recommendation, not a study result.

4. Social-Emotional Outcomes

4.1 Emotion Comprehension and Facial Expressions

In an Italian study of 104 children aged 6 to 7 from two schools, 63 took part in five dog-assisted sessions and 41 formed the comparison group (Scandurra et al., 2021). Emotion comprehension improved in the dog group up to a three-month follow-up, but not in the comparison group. Allocation was non-random and based on availability, and comparison teachers were also asked to address emotion comprehension in usual classroom activities. Differences therefore cannot be attributed unambiguously to the dog.

In France, 10 adolescents with cognitive function disorders in a program with a service dog improved over a year in recognizing dog and human, but not cat, facial expressions, while 13 neurotypical controls without a dog did not change (Toutain et al., 2025). The groups differed in diagnosis as well as dog contact, so this is a first indication. Learning to read dogs is nonetheless a relevant outcome for children who spend time with one (how dogs communicate through body language).

4.2 The Largest Randomized Study: No Demonstrated Additional Benefit

The largest study in this field is also one of the newest. Nagy et al. (2026) examined a dog-assisted program for preschool and primary school children with and without special educational needs in a randomized controlled study with 201 children, 114 in the intervention and 87 in the control group. Parents rated behavior on the Conners scales and the Strengths and Difficulties Questionnaire. For the primary outcomes, there was no significant interaction between group and time. The intervention group improved, but the control group improved similarly on some outcomes, and the authors concluded that the intervention was not better than the control condition. This means no demonstrated additional benefit on the primary measures; it is not proof that the program is ineffective.

Looking only at the intervention group, the program would have seemed successful. Only the control group shows that children improved anyway over the same period, through maturation, the school year or regression toward the mean, which uncontrolled studies cannot detect.

5. Attention, Motivation and Executive Tasks

5.1 A Series of Preschool Experiments

A research group in the United States ran six small within-subject experiments in which preschool children completed tasks with and without a dog, sometimes also with a friendly human or a stuffed dog. With a dog, 14 children aged 4 to 6 completed gross motor tasks faster, with accuracy rising in some tasks and falling in others (Gee et al., 2007). Eleven children followed instructions better with the dog in imitation tasks, but in tandem tasks they did best with a human, then the stuffed dog, and worst with the real dog (Gee et al., 2009).

In a memory task with 12 children, recognition was at ceiling, but children needed fewer instructional prompts with the dog (Gee, Crist, & Carr, 2010). Twelve children made fewer irrelevant choices in object categorization with a real dog than with a stuffed dog or a human (Gee, Church, & Altobelli, 2010). Categorization of animate objects improved only with the real dog in 17 children (Gee, Gould, et al., 2012), and 20 children recognized objects faster and more accurately with a dog than with a human (Gee, Belcher, et al., 2012).

The series suggests that a calm dog can help preschool children stay on task in short structured tasks. Its limits are equally consistent: 11 to 20 children per study, one laboratory, often the same small dogs, probably unblinded coding and brief tasks instead of classroom learning. The tandem result shows that the effect depends on the task.

5.2 Attention in Older Children

In a randomized crossover study, 24 children aged 10 to 14 spent 15 minutes with a therapy dog or with a robotic dog before attention and memory tests (Hediger & Turner, 2014). Learning effects in a memory test and a cancellation task were significantly larger after the dog, but two other attention tests showed no effect. A frontal brain activity measure (hemoencephalography) declined over time with the robot in one test but not with the dog. With four tests and several measures, some significant results are expected by chance, and a robot controls for novelty, not for a friendly person.

A pilot study compared 8 adolescents with atypical development in a specialized education program, who had daily contact with an animal-assisted education dog, with 13 typically developing adolescents in mainstream schooling without a dog (Toutain et al., 2026). They completed a visual search task with dog faces, human faces and cars before the dog arrived and about one month later. The atypical-development group's performance stayed stable and did not vary by stimulus type, so no short-term gain in selective attention was detectable. The sample is very small, and without a comparison group of atypically developing adolescents without a dog, no causal conclusion is possible.

5.3 Activity Without Reported Motivation

In a randomized crossover study outside school, 12 children aged 8 to 12 in an obesity program were more physically active with a dog than with a human confederate, yet rated their motivation, wellbeing and satisfaction no differently (Wohlfarth et al., 2013). The authors proposed implicit motives. For school research this means that motivation questionnaires may miss behavioral changes.

5.4 Spatial Ability: Short-Term Gains, No Lasting Effect

The Lincolnshire program also tested spatial ability in 105 mainstream and 64 SEN children before and after the intervention and after six weeks, six months and one year (Brelsford et al., 2022). The children overlap with those of Meints et al. (2022), so the papers are not independent evidence. In mainstream schools, the dog group scored higher immediately and after six weeks, the relaxation group also improved in the short term, and controls did not improve significantly; there were no long-term effects. In SEN schools, spatial ability improved only after relaxation and did not change after the dog interventions. Careful designs often show exactly this: a short-term benefit, partly matched by an active control, without lasting difference.

6. Reading to Dogs

6.1 A Systematic Review of the Evidence

A systematic review graded 48 reports on children reading to dogs, including grey literature, by level of evidence (Hall et al., 2016). Twenty-seven were at the lowest level (Level 5), 13 at Level 4, seven at Level 2c and only one at Level 2b, the highest level reached. The authors concluded that reading to a dog may benefit behavioral processes around reading and thereby performance, but that "the evidence base … is of low quality" and the "mechanisms for any effect remain a matter of conjecture". The review is sometimes called a meta-analysis; it did not pool effect sizes.

6.2 A Randomized Trial with Active Controls

One randomized study included both an adult and a toy control (le Roux et al., 2014). According to the published summary, 102 third-graders with poor reading skills were assigned to read to a dog, to an adult or to a teddy bear, or to usual school activity, and reading comprehension improved more in the dog group than in the three others. The design addresses the question most reading studies cannot: whether a dog adds anything beyond an attentive adult.

6.3 Small Trials With Mixed Results

In a randomized crossover study, 12 second-graders read with a dog present or with a friendly female student, and did better with the dog on three of four parameters, without differences in reading time (Wohlfarth et al., 2014); with only 12 children, this is a very small sample. In Austria, 36 third-graders had two tutor sessions, one with a dog and handler, in counterbalanced order (Schretzmayer et al., 2017). Standardized reading scores were unaffected. In the first session, children improved on repeated reading only with the dog and showed fewer nervous movements, but heart rate measures did not differ, and in the second session cortisol was higher with the dog. The authors' title, "minor immediate effects", is accurate.

Several small trials found no gain in reading skill. In a five-week pilot with 18 children entering second grade, of whom 15 completed the study, reading ability did not change significantly in the 9 children reading to dogs, while the control group's academic reading attitude declined, with differential dropout weakening the comparison (Lenihan et al., 2016). In a randomized study with 28 second-graders (14 per group), reading skill did not change significantly in either group over a six-week program (Linder et al., 2018). Academic reading attitude improved within the dog group (p = 0.002), while the within-group change in the control group was not statistically significant (p = 0.06); a direct between-group comparison is not reported in the accessible material, so this does not establish an additional dog effect. In nine children with autism, those reading to a dog attended more sessions and parents reported more motivation, but reading and cognitive tests did not differ (Uccheddu et al., 2019).

In a Canadian crossover study with 24 children aged 7 to 8 needing reading support, reading improved particularly after reading to a dog, while improvements in social competence and behavior depended on the order of conditions and on one social competence measure the adult condition did better (Syrnyk et al., 2023). Teachers and parents believed that both forms of support had helped, and the pandemic interrupted the final round of testing. Two tiny studies point to engagement: three third-graders showed signs of relaxation and higher fluency after reading to a dog, without a control (Fung, 2019), and on-task behavior increased in three students with emotional and behavioral disabilities who read to a therapy dog (Bassette & Taber-Doughty, 2013).

6.4 Motivation Versus Skill

Across these studies, the most consistent observations concern behavior during reading: attendance, on-task behavior, nervousness. Effects on standardized reading skills are inconsistent and mostly absent in the small randomized trials (Schretzmayer et al., 2017; Linder et al., 2018; Uccheddu et al., 2019). A conceptual paper argued from motivation theory for extending reading-to-dog programs to children with special educational needs and combining them with dialogic reading (Fung, 2017), a coherent rationale but not evidence. A later review found the research "underdeveloped" and practice lacking "a substantial evidence base", while allowing "cautious optimism" (Steel, 2024). A reading dog can reasonably be offered to make practice more attractive for reluctant readers, but not as a substitute for reading instruction.

In practice: Keep reading-dog sessions short, hold them in a quiet corner rather than in front of the class, and build them around the child's reading task, with the dog as calm company rather than the center of attention. Let children choose how close they sit to the dog. This is a practice recommendation, not a tested protocol.

7. What Reviews and Meta-Analyses Conclude

7.1 Systematic Reviews

A review of animal-assisted interventions in the classroom included 25 papers from 841 records (Brelsford et al., 2017). Most reported significant benefits, but methods, interventions, measures, sample sizes and exposure time varied greatly, and the authors noted "a worrying lack of reference to risk assessment and animal welfare".

A review of dog-assisted interventions in special education identified 18 controlled studies (Meixner & Kotrschal, 2022). Heterogeneity precluded a meta-analysis, and no study reported negative effects. The strongest evidence concerned lower cortisol under challenge and motivation or adherence to instructions; academic and social outcomes were inconclusive, and there were no long-term data. The authors named small samples, non-random allocation, inadequate controls and researcher expectancy as problems and wrote that "serious publication bias in this field of research can be assumed".

A review of social and emotional therapy-dog interventions in mainstream schools found only seven studies with comparison groups among 405 records and stressed the "need for more rigorous and current research" (Wintermantel et al., 2024).

7.2 The Most Rigorous Review: High Risk of Bias

The most recent and strictest review searched up to August 2023 for controlled studies of animal-assisted services for school-aged children outside healthcare and included 30 studies, 21 of them RCTs (Moe et al., 2025). It covers different animal species and settings, not only school dog programs. Of the 21 RCTs, 17, or 81%, had a high overall risk of bias, mainly because outcome assessors knew the allocation (13 trials) or because of missing data (5 trials). Of the nine non-randomized studies, five had a high and four a moderate risk. The authors found positive impacts on socio-emotional functioning and behavior but stated that "all studies had study design problems, which caused high risk of bias". The review was funded by Skandia's "Idéer för livet" foundation and the Agria/SKK Research Fund, and one author is affiliated with a therapy-dog school; the authors declared no competing interests.

7.3 Meta-Analyses

A meta-analysis of elementary to high school students pooled 20 studies, 66 effect sizes and, as reported by the authors, 1,716 participants (Reilly et al., 2020). Average effects were statistically significant for reading (Hedges' g = 0.31), social outcomes (g = 0.71) and emotional/behavioral outcomes (g = 0.38), but not for cognition (g = 0.03) or motor outcomes (g = 0.12). Heterogeneity was high, especially for social outcomes (I² = 93.6%), where journal articles (g = 1.33) differed markedly from dissertations (g = −0.11). A trim-and-fill adjustment for possible publication bias reduced the estimates for reading to g = 0.20 and for emotional/behavioral outcomes to g = 0.24. These findings do not establish a reliable benefit of a school dog in every setting. A meta-analysis of animal-assisted interventions for school-aged children with autism, 16 studies with 489 participants, found small effects on social interaction, communication and symptoms, little evidence of a dose-response relationship, and effects that "may be comparable to activities used in active control conditions" (Dimolareva & Dunn, 2021).

7.4 What the Reviews Agree On

Most studies report positive effects; most are small, unblinded and at high risk of bias; long-term data are almost absent; and active controls often do about as well on some outcomes (Brelsford et al., 2017; Dimolareva & Dunn, 2021; Meixner & Kotrschal, 2022; Moe et al., 2025). Dogs in schools are promising for short-term outcomes such as stress responses and engagement, and unproven for achievement and lasting behavioral change.

8. Methodological Weaknesses

8.1 No Blinding, Involved Raters

Everyone involved knows whether a dog is present, so blinding is "rarely possible" (Meixner & Kotrschal, 2022). Unblinded outcome assessors were the main reason for high risk of bias in the strictest review (Moe et al., 2025), and even careful studies coded videos in which the dog was visible (Schretzmayer et al., 2017). Many school studies rely on teachers rating their own class or parents their own child (Hergovich et al., 2002; Verhoeven et al., 2023; Dimolareva et al., 2026) (how reliable ratings by caregivers are), and ratings can shift because the rater changes (Correale et al., 2017). Notably, the largest randomized study, also based on parent ratings, found no significant group-by-time interaction on the primary measures (Nagy et al., 2026) (how observer bias and blinding affect behavioral measurement).

8.2 Novelty

Most designs compare first contact with a dog against a familiar routine, so effects may reflect novelty rather than the dog (Kotrschal & Ortbauer, 2003; Mombeck & Albers, 2024; Melzer et al., 2026). Among the studies reviewed here, only the Lincolnshire program reports a deliberate familiarization phase (Meints et al., 2022), and in the Austrian reading study, behavioral differences appeared mainly in the first session (Schretzmayer et al., 2017). Whether effects persist once a permanent school dog has become ordinary is largely unknown.

8.3 Small Samples and Clustering

Many studies include 3 to 25 children (Bassette & Taber-Doughty, 2013; Uccheddu et al., 2019; Fung, 2019; Toutain et al., 2025). When one class receives the dog and another serves as control, the analysis often treats each child as independent, although children in a class share a teacher and a climate (Hergovich et al., 2002; Beetz, 2013). With one class per condition, dog and teacher effects cannot be separated.

8.4 Missing Active Controls

A no-treatment control shows that something happened, not that the dog caused it. Where active controls exist, they sometimes match or exceed the dog on some outcomes: relaxation in SEN schools (Brelsford et al., 2022), a human partner in tandem tasks (Gee et al., 2009), an adult reader for social competence (Syrnyk et al., 2023) and control activities in autism research (Dimolareva & Dunn, 2021).

8.5 Publication Bias, Expectations and Reporting

Research on animal-assisted interventions faces "unusual pressure from the public and media to report and publish positive findings" (Serpell et al., 2017). For pet ownership in general, Herzog (2011) argued that a generalized "pet effect" is "not a fact but an unsubstantiated hypothesis". Studies also vary in children, dogs, handlers, dose and measures and rarely quantify what happened between child and dog (Rodriguez et al., 2021). In 2026, extensions of the SPIRIT and CONSORT guidelines for dog-assisted trials were published, with 69 items covering methodological, ethical and welfare aspects, because reporting "remains poor" (Shoesmith et al., 2026).

9. Safety, Hygiene and Allergens

9.1 Bites and Child Safety

Children are the group most at risk of dog bites and zoonoses, according to a scoping review of 393 documents, which concluded that the benefits of child-dog interactions outweigh the risks for children, but not for dogs (Giraudet et al., 2022). No incidence data exist for bites by school or reading dogs. In 111 cases of dogs referred to a university behavior clinic after biting children, most dogs had no previous history of biting children (Reisner et al., 2007). Clinic dogs are not comparable to screened school dogs, but a first bite without known history is a reminder that screening does not replace supervision (what research shows about dog bites and children).

Education can improve what children know. In 96 children aged 3 to 6, a computer-based bite prevention program improved the selection of safe responses in simulated situations, retained after two weeks; real-world interactions with dogs and bite incidence were not assessed (Meints & de Keuster, 2009). After a primary-school safety program, a retrospective study recorded 296 dog-related injuries in children, including 212 dog bites. Severe injuries accounted for 8% of the bite cases (17 of 212), compared with 26% in an earlier period, and inpatient treatment fell from 27.5% to 9.0%, although the number of injuries did not decrease (Kienesberger et al., 2022). This historical comparison cannot establish a causal effect of the prevention program. A school dog program is an opportunity to teach such rules, including how to recognize when a dog wants to be left alone (what is known about calming signals).

In practice: Clear, simple rules work better than general appeals to be kind to the dog: approach only when the teacher allows it, let the dog come to you, never disturb it on its mat, no hugging, no food. Practice the rules with the class before the dog's first visit and repeat them regularly, rather than explaining them once with the dog already in the room.

9.2 Allergens

In a Swedish study, dog allergen was found in all 57 samples from clothing and classrooms of four school classes without a school dog, with significantly more on the clothing of children with dogs at home (Berge et al., 1998). In seven day-care centers, dog allergen was present in all sections, floor levels correlated with the number of people with pets at home, and the cleaning method had no measurable influence (Munir et al., 1995). Cross-sectional measurements of dust and air in 14 classrooms from six schools where no animals were kept also detected the dog allergen Can f 1, carried in from homes; one classroom showed a moderate level, and one of 22 school dust samples exceeded 2 µg/g (Niesler et al., 2016). Without a before-and-after measurement, this describes background exposure in schools without a dog, not whether a school dog would raise levels. Dog allergen thus reaches classrooms through clothing anyway. How much a school dog adds is unknown: none of the studies reviewed here measured levels before and after a dog's introduction or symptoms in allergic classmates. Allergies remain an individual question to clarify with families beforehand.

9.3 Hygiene Policies and Implementation

No school-specific infection data were found in the literature reviewed here. A survey of 45 hospitals, 45 eldercare facilities and 27 therapy-animal organizations in the United States found widely varying hygiene policies, and 70% of the organizations allowed animals fed raw meat to visit (Linder et al., 2017). There are no unified standardized guidelines or risk assessment tools for animal-assisted interventions, which "take place in an unregulated manner"; this refers to the lack of unified AAI standards and risk-assessment tools, not to the legal situation in every country. The authors of the Lincolnshire trials published their own risk assessment and welfare tools (Brelsford et al., 2020). School staff and therapy-dog coordinators named sanitation, allergies and safety as risks, whole-school support, communication and dog welfare as facilitators, and workload and missing guidelines as barriers (Grové et al., 2021). In a self-selected UK online survey, 318 mostly female caregivers rated hypothetical scenarios; reading to a dog and one-to-one contact received descriptively higher acceptability ratings than a classroom dog aimed at behavior (Fynn & Runacres, 2022). The reported statistic is internally inconsistent (a chi-square of 3.06 with two degrees of freedom corresponds to p ≈ 0.22, not p < 0.001), so the difference should not be treated as statistically established, and no real programs or child outcomes were observed.

In practice: A fixed, written hygiene routine is easy to establish and to explain to parents: hand washing after contact, no dog where food is prepared or eaten, a washable mat as the dog's place, regular veterinary checks and parasite prevention, and documentation. Schools with such a routine tend to have fewer discussions with concerned parents.

10. The Dog's Side

Child outcomes are only half of the picture. Most classroom studies did not mention risk assessment or animal welfare (Brelsford et al., 2017), the main reading review did not address dog welfare (Hall et al., 2016), and the welfare of dogs working with children is underexplored (Giraudet et al., 2022). A teacher and a handler reporting on more than ten years of a school literacy program described practical challenges: selecting teams, preparing teachers and students, adapting to abilities, and dealing with loss when a dog leaves or retires (Kirnan & Thompson, 2023).

What studies of therapy, visiting and school dogs show about cortisol, heart rate and behavior during sessions is reviewed in a separate article (what animal-assisted work means for the dog). In schools, noise, many children at once, sudden movements and long days add to the load (how stressors accumulate), and the ability to withdraw matters (why control and predictability matter for dogs).

In practice: A school dog needs a quiet zone that children do not enter, such as a mat or crate away from the door, and the right to go there at any time. Make retreat a routine part of the dog's school day, and plan a shorter presence, such as selected lessons, rather than a full school day. None of the studies reviewed here determined how long a school dog can be present without strain.

11. Summary at a Glance

Lower cortisol responses with a dog are the best-supported finding, from laboratory tests and one school RCT, without evidence that they improve learning (Beetz et al., 2012; Meints et al., 2022). Classroom climate studies rest on single classes and teacher ratings, and the largest randomized study found no demonstrated additional benefit on the primary measures (Beetz, 2013; Nagy et al., 2026). Preschool task effects depend on the task, and a spatial-ability benefit did not last (Gee et al., 2009; Brelsford et al., 2022). Reading-to-dog evidence is of low quality, with engagement effects more consistent than skill effects (Hall et al., 2016; Steel, 2024). Most randomized trials carry a high risk of bias, outcome assessors are rarely blinded, novelty is seldom controlled, and publication bias is likely (Meixner & Kotrschal, 2022; Moe et al., 2025).

Safety data are thin: dog allergen is present in classrooms even without a school dog, a computer-based bite prevention program improved children's choices in simulated situations, and the studies reviewed here provide no incidence data for school dogs (Berge et al., 1998; Meints & de Keuster, 2009). Dog welfare in school settings is rarely addressed in the child-outcome literature (Brelsford et al., 2017).

12. Limitations and Research Gaps

12.1 Short Time Frames and Permanent School Dogs

Almost none of the studies reviewed here follow children beyond weeks; the one-year follow-up found no lasting effect (Brelsford et al., 2022), and none links a school dog to grades. Randomized trials mostly use visiting teams (le Roux et al., 2014; Meints et al., 2022), while teacher-and-dog teams that accompany a class for months were studied only in single-class or interview designs (Kotrschal & Ortbauer, 2003; Hergovich et al., 2002; Beetz, 2013; Mombeck & Albers, 2024).

12.2 Mechanism and Target Groups

Whether effects come from the dog, an extra friendly adult, novelty or a break in routine is unresolved; active controls sometimes match the dog, and the authors of a recent study named social interaction and time away from academic demands as alternative explanations (Syrnyk et al., 2023; Melzer et al., 2026). The laboratory studies with insecurely attached boys have not been tested in classrooms (Beetz et al., 2012). Results for children with special educational needs are inconsistent: lower cortisol after group sessions (Meints et al., 2022), no dog effect on spatial ability (Brelsford et al., 2022), no demonstrated additional benefit on the primary behavioral measures (Nagy et al., 2026), small effects in autism (Dimolareva & Dunn, 2021).

12.3 Safety and Design

Among the studies reviewed here, there are no bite incidence data for school dogs and no school-specific infection data; hygiene evidence comes from hospital and eldercare policies (Linder et al., 2017). Existing allergen studies measured background levels in schools and day-care centers without a dog (Berge et al., 1998; Munir et al., 1995); none measured levels before and after a dog's introduction or symptoms in allergic children. Dog-side data from school and reading settings rest on very few animals (Giraudet et al., 2022). Most randomized trials are at high risk of bias, and reporting standards have only just been published (Moe et al., 2025; Shoesmith et al., 2026). Kazdin (2017) argued that progress requires explicit theories of mechanism, not only more studies.

13. Conclusion

The research on dogs in schools is more encouraging than skeptics assume and much weaker than advocates claim. Consistent findings concern short-term changes: lower cortisol responses, better adherence to instructions in preschool tasks, more engagement while reading (Beetz et al., 2012; Meints et al., 2022; Gee et al., 2009; Hall et al., 2016). Evidence for lasting effects on behavior or academic skills is thin, and recent well-designed studies found no demonstrated additional benefit of the dog on several outcomes (Brelsford et al., 2022; Nagy et al., 2026).

A well-prepared dog in a well-planned program, with a handler who watches the dog and clear rules for children, can make school more pleasant for some children, and the controlled studies in special education reported no negative effects for children (Meixner & Kotrschal, 2022). It is not a proven educational intervention, and its value depends on planning, safety and the welfare of the dog as much as on the dog's presence.

Key Insights (Takeaways)

  • Lower cortisol responses with a dog present are the best-supported finding, from laboratory tests with insecurely attached boys and a school-based RCT (Beetz et al., 2012; Meints et al., 2022).

  • Classroom studies reporting better climate and behavior mostly compare one dog class with one control class and rely on teacher ratings (Hergovich et al., 2002; Beetz, 2013).

  • The largest randomized study, with 201 children, found no demonstrated additional benefit of a dog-assisted program over the control group on the primary measures, which is not proof of ineffectiveness (Nagy et al., 2026).

  • Preschool experiments suggest better adherence to instructions with a dog in short tasks, but the effect depends on the task (Gee et al., 2009).

  • Reading to dogs is supported by low-quality evidence; effects on engagement are more consistent than effects on reading skill (Hall et al., 2016; Steel, 2024).

  • In the most rigorous review, which spanned different animal species and settings, 81% of randomized trials (17 of 21) had a high risk of bias, mainly because outcome assessors were not blinded (Moe et al., 2025).

  • Dog allergen is present in classrooms even without a school dog, but none of the studies reviewed here measured what a school dog adds (Berge et al., 1998).

References

Bassette, L. A., & Taber-Doughty, T. (2013). The effects of a dog reading visitation program on academic engagement behavior in three elementary students with emotional and behavioral disabilities: A single case design. Child & Youth Care Forum, 42(3), 239–256. https://doi.org/10.1007/s10566-013-9197-y

Beetz, A. (2013). Socio-emotional correlates of a schooldog-teacher-team in the classroom. Frontiers in Psychology, 4, 886. https://doi.org/10.3389/fpsyg.2013.00886

Beetz, A., Julius, H., Turner, D., & Kotrschal, K. (2012). Effects of social support by a dog on stress modulation in male children with insecure attachment. Frontiers in Psychology, 3, 352. https://doi.org/10.3389/fpsyg.2012.00352

Beetz, A., Kotrschal, K., Turner, D. C., Hediger, K., Uvnäs-Moberg, K., & Julius, H. (2011). The effect of a real dog, toy dog and friendly person on insecurely attached children during a stressful task: An exploratory study. Anthrozoös, 24(4), 349–368. https://doi.org/10.2752/175303711X13159027359746

Berge, M., Munir, A. K., & Dreborg, S. (1998). Concentrations of cat (Fel d1), dog (Can f1) and mite (Der f1 and Der p1) allergens in the clothing and school environment of Swedish schoolchildren with and without pets at home. Pediatric Allergy and Immunology, 9(1), 25–30. https://doi.org/10.1111/j.1399-3038.1998.tb00296.x

Brelsford, V. L., Dimolareva, M., Gee, N. R., & Meints, K. (2020). Best practice standards in animal-assisted interventions: How the LEAD risk assessment tool can help. Animals, 10(6), 974. https://doi.org/10.3390/ani10060974

Brelsford, V. L., Dimolareva, M., Rowan, E., Gee, N. R., & Meints, K. (2022). Can dog-assisted and relaxation interventions boost spatial ability in children with and without special educational needs? A longitudinal, randomized controlled trial. Frontiers in Pediatrics, 10, 886324. https://doi.org/10.3389/fped.2022.886324

Brelsford, V. L., Meints, K., Gee, N. R., & Pfeffer, K. (2017). Animal-assisted interventions in the classroom – A systematic review. International Journal of Environmental Research and Public Health, 14(7), 669. https://doi.org/10.3390/ijerph14070669

Correale, C., Crescimbene, L., Borgi, M., & Cirulli, F. (2017). Development of a dog-assisted activity program in an elementary classroom. Veterinary Sciences, 4(4), 62. https://doi.org/10.3390/vetsci4040062

Dimolareva, M., & Dunn, T. J. (2021). Animal-assisted interventions for school-aged children with autism spectrum disorder: A meta-analysis. Journal of Autism and Developmental Disorders, 51(7), 2436–2449. https://doi.org/10.1007/s10803-020-04715-w

Dimolareva, M., Doolan-Dransfield, E., Duckworth, J., Brelsford, V. L., Meints, K., & Gee, N. R. (2026). Pawsitive impact: Measuring the Dog Mentor's effect in neurodivergent students. Behavioral Sciences, 16(3), 323. https://doi.org/10.3390/bs16030323

Dimolareva, M., Gee, N. R., Pfeffer, K., Maréchal, L., Pennington, K., & Meints, K. (2018). Measuring cortisol in the classroom with school-aged children – A systematic review and recommendations. International Journal of Environmental Research and Public Health, 15(5), 1025. https://doi.org/10.3390/ijerph15051025

Fung, S. (2017). Canine-assisted reading programs for children with special educational needs: Rationale and recommendations for the use of dogs in assisting learning. Educational Review, 69(4), 435–450. https://doi.org/10.1080/00131911.2016.1228611

Fung, S. (2019). Effect of a canine-assisted read aloud intervention on reading ability and physiological response: A pilot study. Animals, 9(8), 474. https://doi.org/10.3390/ani9080474

Fynn, W. I., & Runacres, J. (2022). Dogs at school: A quantitative analysis of parental perceptions of canine-assisted activities in schools mediated by child anxiety. International Journal of Child Care and Education Policy, 16(1), 4. https://doi.org/10.1186/s40723-022-00097-x

Gee, N. R., Belcher, J. M., Grabski, J. L., DeJesus, M., & Riley, W. (2012). The presence of a therapy dog results in improved object recognition performance in preschool children. Anthrozoös, 25(3), 289–300. https://doi.org/10.2752/175303712X13403555186172

Gee, N. R., Church, M. T., & Altobelli, C. L. (2010). Preschoolers make fewer errors on an object categorization task in the presence of a dog. Anthrozoös, 23(3), 223–230. https://doi.org/10.2752/175303710X12750451258896

Gee, N. R., Crist, E. N., & Carr, D. N. (2010). Preschool children require fewer instructional prompts to perform a memory task in the presence of a dog. Anthrozoös, 23(2), 173–184. https://doi.org/10.2752/175303710X12682332910051

Gee, N. R., Gould, J. K., Swanson, C. C., & Wagner, A. K. (2012). Preschoolers categorize animate objects better in the presence of a dog. Anthrozoös, 25(2), 187–198. https://doi.org/10.2752/175303712X13316289505387

Gee, N. R., Griffin, J. A., & McCardle, P. (2017). Human–animal interaction research in school settings: Current knowledge and future directions. AERA Open, 3(3). https://doi.org/10.1177/2332858417724346

Gee, N. R., Harris, S. L., & Johnson, K. L. (2007). The role of therapy dogs in speed and accuracy to complete motor skills tasks for preschool children. Anthrozoös, 20(4), 375–386. https://doi.org/10.2752/089279307X245509

Gee, N. R., Sherlock, T. R., Bennett, E. A., & Harris, S. L. (2009). Preschoolers' adherence to instructions as a function of presence of a dog and motor skills task. Anthrozoös, 22(3), 267–276. https://doi.org/10.2752/175303709X457603

Giraudet, C. S. E., Liu, K., McElligott, A. G., & Cobb, M. (2022). Are children and dogs best friends? A scoping review to explore the positive and negative effects of child-dog interactions. PeerJ, 10, e14532. https://doi.org/10.7717/peerj.14532

Grové, C., Henderson, L., Lee, F., & Wardlaw, P. (2021). Therapy dogs in educational settings: Guidelines and recommendations for implementation. Frontiers in Veterinary Science, 8, 655104. https://doi.org/10.3389/fvets.2021.655104

Hall, S. S., Gee, N. R., & Mills, D. S. (2016). Children reading to dogs: A systematic review of the literature. PLOS ONE, 11(2), e0149759. https://doi.org/10.1371/journal.pone.0149759

Hediger, K., & Turner, D. C. (2014). Can dogs increase children's attention and concentration performance? A randomised controlled trial. Human-Animal Interaction Bulletin. https://doi.org/10.1079/hai.2014.0010

Hergovich, A., Monshi, B., Semmler, G., & Zieglmayer, V. (2002). The effects of the presence of a dog in the classroom. Anthrozoös, 15(1), 37–50. https://doi.org/10.2752/089279302786992775

Herzog, H. (2011). The impact of pets on human health and psychological well-being. Current Directions in Psychological Science, 20(4), 236–239. https://doi.org/10.1177/0963721411415220

Kazdin, A. E. (2017). Strategies to improve the evidence base of animal-assisted interventions. Applied Developmental Science, 21(2), 150–164. https://doi.org/10.1080/10888691.2016.1191952

Kienesberger, B., Arneitz, C., Wolfschluckner, V., Flucher, C., Spitzer, P., Singer, G., Castellani, C., Till, H., & Schalamon, J. (2022). Child safety programs for primary school children decrease the injury severity of dog bites. European Journal of Pediatrics, 181(2), 709–714. https://doi.org/10.1007/s00431-021-04256-z

Kirnan, J., & Thompson, A. (2023). When reading gets ruff: Welfare challenges and strategies in an elementary school dog-assisted literacy program. Animal Behaviour and Welfare Cases. https://doi.org/10.1079/abwcases.2023.0020

Kotrschal, K., & Ortbauer, B. (2003). Behavioral effects of the presence of a dog in a classroom. Anthrozoös, 16(2), 147–159. https://doi.org/10.2752/089279303786992170

Lenihan, D., McCobb, E., Diurba, A., Linder, D. E., & Freeman, L. M. (2016). Measuring the effects of reading assistance dogs on reading ability and attitudes in elementary schoolchildren. Journal of Research in Childhood Education, 30(2), 252–259. https://doi.org/10.1080/02568543.2016.1143896

le Roux, M. C., Swartz, L., & Swart, E. (2014). The effect of an animal-assisted reading program on the reading rate, accuracy and comprehension of grade 3 students: A randomized control study. Child & Youth Care Forum, 43(6), 655–673. https://doi.org/10.1007/s10566-014-9262-1

Linder, D. E., Mueller, M. K., Gibbs, D. M., Alper, J. A., & Freeman, L. M. (2018). Effects of an animal-assisted intervention on reading skills and attitudes in second grade students. Early Childhood Education Journal, 46(3), 323–329. https://doi.org/10.1007/s10643-017-0862-x

Linder, D. E., Siebens, H. C., Mueller, M. K., Gibbs, D. M., & Freeman, L. M. (2017). Animal-assisted interventions: A national survey of health and safety policies in hospitals, eldercare facilities, and therapy animal organizations. American Journal of Infection Control, 45(8), 883–887. https://doi.org/10.1016/j.ajic.2017.04.287

Meints, K., Brelsford, V. L., Dimolareva, M., Maréchal, L., Pennington, K., Rowan, E., & Gee, N. R. (2022). Can dogs reduce stress levels in school children? Effects of dog-assisted interventions on salivary cortisol in children with and without special educational needs using randomized controlled trials. PLOS ONE, 17(6), e0269333. https://doi.org/10.1371/journal.pone.0269333

Meints, K., & de Keuster, T. (2009). Brief report: Don't kiss a sleeping dog: The first assessment of "The Blue Dog" bite prevention program. Journal of Pediatric Psychology, 34(10), 1084–1090. https://doi.org/10.1093/jpepsy/jsp053

Meixner, J., & Kotrschal, K. (2022). Animal-assisted interventions with dogs in special education – A systematic review. Frontiers in Psychology, 13, 876290. https://doi.org/10.3389/fpsyg.2022.876290

Melzer, D. K., Yeater, D. M., DeRubis, M., Kohlroser, C., & Pierce, B. J. (2026). Brief therapy dog interactions may increase self-efficacy and lower perceived stress and cortisol in high school students. Frontiers in Psychology, 17, 1876920. https://doi.org/10.3389/fpsyg.2026.1876920

Moe, I., Ho, P. J., Andersson, M., Karlberg, S., Lidfors, L., Sampaio, F., & Feldman, I. (2025). Effectiveness of animal-assisted services for school-aged children: A systematic review. European Child & Adolescent Psychiatry, 34(10), 3017–3033. https://doi.org/10.1007/s00787-025-02740-7

Mombeck, M. M., & Albers, T. (2024). The impact of a school dog on children's social inclusion and social climate in a school class. European Journal of Investigation in Health, Psychology and Education, 14(1), 1–17. https://doi.org/10.3390/ejihpe14010001

Munir, A. K. M., Einarsson, R., & Dreborg, S. K. G. (1995). Mite (Der p I, Der f I), cat (Fel d I) and dog (Can f I) allergens in dust from Swedish day-care centres. Clinical & Experimental Allergy, 25(2), 119–126. https://doi.org/10.1111/j.1365-2222.1995.tb01016.x

Nagy, B. E., Balogh, É. Z., & Kovács, K. E. (2026). Effects of a dog-assisted animal-assisted therapy on behavioural functioning in preschool and primary school children: A randomised controlled study. Children, 13(7), 924. https://doi.org/10.3390/children13070924

Niesler, A., Ścigała, G., & Łudzeń-Izbińska, B. (2016). Cat (Fel d 1) and dog (Can f 1) allergen levels in cars, dwellings and schools. Aerobiologia, 32(3), 571–580. https://doi.org/10.1007/s10453-016-9433-7

Reilly, K. M., Adesope, O. O., & Erdman, P. (2020). The effects of dogs on learning: A meta-analysis. Anthrozoös, 33(3), 339–360. https://doi.org/10.1080/08927936.2020.1746523

Reisner, I. R., Shofer, F. S., & Nance, M. L. (2007). Behavioral assessment of child-directed canine aggression. Injury Prevention, 13(5), 348–351. https://doi.org/10.1136/ip.2007.015396

Rodriguez, K. E., Herzog, H., & Gee, N. R. (2021). Variability in human-animal interaction research. Frontiers in Veterinary Science, 7, 619600. https://doi.org/10.3389/fvets.2020.619600

Scandurra, C., Santaniello, A., Cristiano, S., Mezza, F., Garzillo, S., Pizzo, R., Menna, L. F., & Bochicchio, V. (2021). An animal-assisted education intervention with dogs to promote emotion comprehension in primary school children – The Federico II Model of Healthcare Zooanthropology. Animals, 11(6), 1504. https://doi.org/10.3390/ani11061504

Schretzmayer, L., Kotrschal, K., & Beetz, A. (2017). Minor immediate effects of a dog on children's reading performance and physiology. Frontiers in Veterinary Science, 4, 90. https://doi.org/10.3389/fvets.2017.00090

Serpell, J., McCune, S., Gee, N., & Griffin, J. A. (2017). Current challenges to research on animal-assisted interventions. Applied Developmental Science, 21(3), 223–233. https://doi.org/10.1080/10888691.2016.1262775

Shoesmith, E., Ratschen, E., Riga, E., Mills, D. S., Gibsone, S., McMillan, D., Wu, Q., Clarke, C., Sprange, K., Meints, K., Gee, N. R., Nieforth, L. O., Dollion, N., Fine, A., & Hall, S. S. (2026). Guidance for protocol content and reporting of dog-assisted interventions in randomised controlled trials: Explanation and elaboration of the SPIRIT 2025 and CONSORT 2025 extensions. BMC Medical Research Methodology, 26(1), 114. https://doi.org/10.1186/s12874-025-02663-6

Steel, J. (2024). Reading to dogs as a form of animal-assisted education: Are positive outcomes supported by quality research? Literacy, 58(1), 102–119. https://doi.org/10.1111/lit.12345

Syrnyk, C., McArthur, A., Zwack, A., & Handelsman, A. (2023). Supporting young readers: A mixed-methods study of their literacy, behaviour, and perceptions when reading aloud to dogs. Early Childhood Education Journal, 51(6), 1091–1105. https://doi.org/10.1007/s10643-022-01392-5

Toutain, M., Dollion, N., Henry, L., & Grandgeorge, M. (2025). Does a dog at school help identify human and animal facial expressions? A preliminary longitudinal study. European Journal of Investigation in Health, Psychology and Education, 15(2), 13. https://doi.org/10.3390/ejihpe15020013

Toutain, M., Dollion, N., Roger, A., Beylie, A., Pignard, E., Henry, L., & Grandgeorge, M. (2026). Short paper: Pilot study of the effect of the presence of an animal-assisted education dog in a school setting on the selective attention capacities of adolescents with atypical development. Human-Animal Interactions, Article 0041. https://doi.org/10.1079/hai.2026.0041

Uccheddu, S., Albertini, M., Pierantoni, L., Fantino, S., & Pirrone, F. (2019). The impacts of a reading-to-dog programme on attending and reading of nine children with autism spectrum disorders. Animals, 9(8), 491. https://doi.org/10.3390/ani9080491

Verhoeven, R., Butter, R., Martens, R., & Enders-Slegers, M.-J. (2023). Animal-assisted education: Exploratory research on the positive impact of dogs on behavioral and emotional outcomes of elementary school students. Children, 10(8), 1316. https://doi.org/10.3390/children10081316

Wintermantel, L., Grové, C., Henderson, L., & Laletas, S. (2024). Social and emotional therapy dog-assisted interventions in mainstream school settings: A systematic review. The Educational and Developmental Psychologist, 41(1), 74–90. https://doi.org/10.1080/20590776.2023.2256444

Wohlfarth, R., Mutschler, B., Beetz, A., Kreuser, F., & Korsten-Reck, U. (2013). Dogs motivate obese children for physical activity: Key elements of a motivational theory of animal-assisted interventions. Frontiers in Psychology, 4, 796. https://doi.org/10.3389/fpsyg.2013.00796

Wohlfarth, R., Mutschler, B., Beetz, A., & Schleider, K. (2014). An investigation into the efficacy of therapy dogs on reading performance in 6–7 year old children. Human-Animal Interaction Bulletin. https://doi.org/10.1079/hai.2014.0013