The most common type of human animal interaction (HAI) programme used in prisons involves prisoners caring for and training unwanted dogs from rescue shelters, to prepare the dogs for rehoming. Such programmes have been previously developed specifically aimed towards male young offenders, and are claimed to improve emotional, social and practical outcomes. Paws for Progress, the first prison based dog training programme in the UK, was introduced to HM YOI Polmont in 2011. By clearly communicating each step of the 5 Step approach (1. Identify the problem; 2. Review the evidence; 3. Develop a logic model; 4. Identify indicators and monitor the logic model; 5. Evaluate the logic model), it has enhanced our understanding of the development processes required for effective prison based dog training programmes. This evaluation provides the first comprehensive quantitative analysis of short, medium and long term outcomes for Scottish young offenders serving custodial sentences (N = 70) following participation. The aims of Paws for Progress are to improve behaviour, increase engagement in education, develop employability skills, and enhance well-being. Using a mixed design with two control groups and triangulating quantitative and qualitative outcomes, the evaluation assesses the efficacy of the programme in meeting these aims. Systematic analyses of semi-structured interviews pre and post participation in the programme support findings from the quantitative analyses. Analyses of institutional behaviour, measured by Disciplinary Reports, educational progress measured by written assessments and qualifications, employability skills measured by psychometric tests, and prisoner well-being all improved for participants, but such improvements were not shown by control groups. Paws for Progress positively impacts short and medium term outcomes and data on longer term outcomes also indicate the benefits are far reaching. By clearly relating programme aims to the outcomes achieved, and considering the contribution of Paws for Progress to future desistance from crime, the value and relevance of these findings are evident. The evaluation contributes to our understanding of effective methodologies in this applied context, which can be utilised to improve research practice in interventions in criminal justice and in human animal interaction.

<p> There has been substantial growth in recent decades in the variety and popularity of roles for dogs assisting humans in professional therapeutic partnerships. Simultaneously, increasingly rigorous research has repeatedly demonstrated the effects of professional human-canine partnerships in remedying important issues of public health among several at-risk populations. Yet, despite these areas of growth, mechanisms of action and predictors of efficacy in the field of human-animal interaction (HAI) remain poorly understood, and the role of human diversity has been rarely discussed. Thus, the present dissertation examines potential mechanisms and diverse predictors in two distinct samples of professional human-canine partnerships, while building the impetus to explore diversity in the HAI field as a whole. For the first three studies (Chapters 2-4), the selected samples of professional human-canine partnerships include military veterans working with psychiatric service dogs to mediate their symptoms of PTSD and healthcare professionals in pediatric hospitals working with facility dogs to benefit their patients. Following the introduction in Chapter 1, the objective of Chapters 2-3 was to examine primary human outcomes in the selected professional canine partnerships. In a crosssectional study of N=198 military veterans with PTSD, Chapter 2 compared PTSD symptom severity between n=112 veterans with service dogs and n=86 veterans on the waitlist to receive service dogs in the future. Next, in a cross-sectional study of N=130 healthcare professionals in pediatric hospitals, Chapter 3 compared job-related well-being and mental health of n=65 professionals working with facility dogs to n=65 working without. Findings suggested benefits to the mental health and well-being of both military veterans with PTSD and pediatric healthcare professionals, which were significantly associated with their professional canine partnerships. Subsequently, the objective of Chapter 4 was to explore how variances within a specific professional canine partnership may suggest predictors and potential mechanisms for the observed human outcomes. Thus, in a longitudinal study of N=82 veterans with PTSD and their service dogs, Chapter 4 explored associations of veterans’ outcomes with veteran-service dog demographics and interactions. Results suggested components of the human-canine partnership which might explain observed human outcomes, including social connections, a calming influence, and strong humananimal bonds. </p>

<p>The attached document contains the dissertation of Clare Jensen, including descriptions of all procedures used in the associated research studies.</p>

<p>Psychiatric assistance dogs for military veterans with posttraumatic stress disorder (PTSD) currently make up over 19% of assistance dog partnerships globally. We conducted a systematic review of the literature relating to these partnerships, with specific aims to (1) summarize their characteristics, (2) evaluate the quality of existing evidence, and (3) summarize outcomes. A total of 432 records were independently screened (Cohen’s kappa=0.90). Of these, 41 articles (29 peer-reviewed publications and 12 unpublished dissertations) met inclusion criteria. Data extraction was conducted to address the research aims, including a meta-analysis (quantitative outcomes) and meta-synthesis (qualitative outcomes). All peer-reviewed publications on the topic of psychiatric assistance dogs for veterans with PTSD were published within the last five years. The majority of included articles were quantitative (53%), 41% were qualitative, and 6% employed mixed methods. Mean methodological rigor scores were 80% for peer reviewed articles and 71% for dissertations, where higher scores represent more rigorous methodology. Quantitative articles reported significant improvements in the domains of PTSD severity, mental health, and social health. Impacts on physical health and global quality of life appear inconclusive. Meta-analysis (9 articles) revealed that partnership with an assistance dog had a clinically meaningful, significant, and large effect on PTSD severity scores (<em>g</em>=−1.129; <em>p</em><0.0001). Qualitative meta-synthesis identified two third order constructs: (1) Impact on the individual: mental & physical health and (2) Impact beyond the individual: building relationships & connection. This synthesis of increasingly prevalent research on assistance dogs for veterans with PTSD provides support for the impact of this complementary and integrative health intervention on PTSD symptom severity, and signs of meaningful improvements in adjacent domains including mental and social health. Gaps between quantitative and qualitative findings, along with the need to report greater demographic detail, highlight key opportunities for future research.</p>