Epsilon Archive for Student Projects

Abstract

Animal-assisted education (AAE) involving school dogs has become increasingly common in
educational settings. While most research has primarily focused on potential benefits for pupils,
less attention has been given to the welfare of the participating dogs. The aim of this study was to
assess the welfare of school dogs by analysing behavioural indicators observed during repeated
school-dog intervention meetings.
Behavioural data were collected from video recordings and handler-completed behavioural
protocols involving eight school dogs and thirteen pupils from six schools in Sweden. The
intervention lasted for 20 weeks, during which pupils met the school dogs twice per week in
approximately 30-minute meetings. Video recordings were obtained at three time points during the
intervention period (beginning, middle, and end of the intervention). The videos were analysed
using The Observer XT according to a predefined ethogram, and behaviours were coded for both
frequency and duration. Handler-completed protocols were evaluated descriptively and compared
with the video observations. Statistical analyses were performed in SPSS using Friedman tests and
Wilcoxon signed-rank tests.
Overall, the dogs displayed substantially more positive than negative behavioural indicators
during the intervention meetings. Tail wagging showed the highest duration percentage across
meetings (Beginning: 27.20 ± 7.16%; Middle: 20.92 ± 5.03%; End: 22.76 ± 8.51%). Physical
contact initiated by pupils was also frequently observed (Beginning: 15.04 ± 6.54%; Middle: 14.44
± 6.13%; End: 6.30 ± 2.61%). Behaviours commonly associated with stress or discomfort, such as
avoidance, low body posture, trembling, and tail between the legs, occurred rarely or were not
observed. Significant behavioural changes across meetings were detected for following behaviour
towards the handler (duration: p = 0.006; frequency: p = 0.015) and command execution from
handler (p = 0.034), both of which decreased across the intervention period. Eye contact with the
handler tended to increase across meetings. Comparisons between video observations and handler
protocols revealed high agreement for conspicuous behaviours such as tail wagging, reward, and
contact-seeking behaviour, whereas subtle behaviours including yawning, avoidance, and shaking
the whole body were more frequently identified through video analysis.
The findings suggest generally good welfare among the participating school dogs during the
intervention meetings. At the same time, the results highlight the complexity of interpreting canine
behavioural indicators and demonstrate the value of combining different assessment methods
when evaluating welfare in animal-assisted educational settings. Due to the limited sample size,
conclusions should be restricted to the dogs included in this project.