Animal-assisted interventions in mainstream schools: What works?

The inclusion of animals within the educational environment has increased in popularity in recent years, with many positive aspects being credited to the practice (Gee, Fine, McCardle, 2017; Fine, 2015). Whilst the value of animal-assisted interventions (AAI) has been reported for both psychological and physiological health of humans (for overviews and systematic reviews, e.g. Brelsford, Meints, Gee & Pfeffer, 2017; Gee, Fine & McCardle, 2017; Hall, Gee & Mills, 2016; Gee, Fine & Schuck 2015; Meaujean, Pepping & Kendall. 2015; Kamioka et al., 2014; O’Haire, 2013a), little is known about what effect AAI is having on school children as the practice is not routinely, nor systematically, studied or subjected to on-going external scrutiny (Gee, Fine & Schuck, 2017). Importantly, there is also a lack of rigorous research carried out into the effects of animal-assisted interventions which specifically focus on children’s learning, development and personal wellbeing over time (Brelsford, et al., 2017; O’Haire, 2013b). In order to achieve optimal outcomes for children, schools need to ensure interventions are effective. Additionally, such interventions require the attendance of live animals within educational settings, whose welfare needs must equally be upheld; it is therefore vital that animal-assisted interventions are assessed rigorously within the educational environment to ensure best outcomes for all (Brelsford, et al., 2017). To address this issue, this research project comprises a longitudinal, randomised-controlled study carried out within mainstream schools in Lincolnshire, UK. Children were assigned to either a dog intervention, relaxation intervention or no treatment control condition. Intervention sessions were carried out for 20-minutes, twice per week, over 4- consecutive weeks. In order to assess cost efficiencies for schools, and the potential for less contact time for dogs, intervention-sessions were carried out as either one-to-one or group scenarios. Baseline measures of children’s cognitive, socio-emotional, physiological and behavioural functioning were completed through standardised tests and experimental tasks before and after interventions. These were repeated after 4 weeks of intervention, and again after 6-weeks, 6-months and 1-year. Standardised measures collected included socioemotional measures of empathising/systemising, self-esteem and anxiety, cognitive measures of language (sentence comprehension & syntactic formulation) and cognition (non-verbal reasoning & spatial ability). Experimental tasks of categorisation, maths and a Stroop task were also employed. Physiological measures of baseline salivary cortisol were collected before and after intervention, and acute cortisol was measured by collecting cortisol before and after the first, the fourth and the last intervention session. Questionnaire data on children’s behaviour at home, their sleep efficiency, family information, and pet-ownership details were collected through parent-questionnaires. Teachers also completed a classroom behaviour questionnaire for each child before and after intervention. The project adhered to strict protocols for the application of dog-assisted interventions. Risk assessments were carried out for all school settings and care plans completed for all dogs taking part in the research. Dogs were recruited through Pets as Therapy (PAT) and underwent further behavioural assessment to ensure suitability to work with children; handlers, children and teachers also received safety training. Repeated measures analysis of variance (ANOVA) was carried out to assess children’s scores over time, with intervention condition, gender and dog ownership status factored into the analysis. Highly significant effects of learning were found in all groups over time, and no differences were found in scores between the schools taking part in the project. With respect to the intervention type, analyses revealed significant immediate effects of the individual dog intervention on cognitive functioning with improved spatial ability, reduced interference in the Fruit Stroop task and improved sentence comprehension. Highly significant effects were also revealed in relation to children’s physiological measures of baseline salivary cortisol with children in the dog intervention showing no increase in cortisol before and after intervention, while both the relaxation group and the control group do show significant increases in stress levels over the school term. Acute cortisol levels significantly decreased after all dog interventions, too. This clearly highlights the successful stress-moderating effects of the dog intervention compared to other conditions. Effects on behaviour were found with dog ownership a significant factor in children’s behaviour at home and within the classroom, as rated by teachers and parents. No significant effects were found for socio-emotional measures of empathy-systemising, self-esteem or anxiety, nor were any effects found for sleep efficiency. Results are discussed relative to the hypotheses of the project with animal assisted interventions within the classroom environment showing some immediate positive effects on certain areas of learning but not all. No longitudinal benefits of dog-assisted interventions were evident across the study for any of the measures of cognition, language, socio-emotional or physiological functioning. Overall the results of the study support the Biopsychosocial model whereby multiple factors of biological, psychological and social functioning of the child are affected by interaction with a dog. Conclusions also discuss the results in light of theoretical underpinnings of the wider field of human-animal interaction research and recommendations are made for future study