Abstract
Medical detection dogs (MDD) are utilised to aid human health; Bio-detection dogs are trained to identify ex-situ conditions' odours, and Alert Assistance dogs to alert crises in people with chronic illnesses (Rooney et al., 2013). Individual variation is associated with detection dog performance (Lazarowsky et al., 2020). However, it is unexplored in MDD. Identifying how MDD characteristics relate to task success will aid the selection of optimal dogs.This project aims to identify traits relevant to MDD and explore how these vary across tasks and training stages. It seeks to develop a test battery to investigate these traits in an MDD sample.
A practitioner survey revealed traits important for MDD and those that differ most from ideal. Some characteristics varied significantly across tasks. In total, 27 relevant attributes were derived.
Trainers from the charity Medical Detection Dogs® rated these 27 traits and overall ability of 58 MDD at multiple time-points. Ratings showed low consistency over training for most traits. Some characteristics were significantly associated with derived success measures. Success measures were explored and derived for subsequent studies: training outcome, composite total ability score, scent sensitivity, and specificity.
The dogs were tested with a test battery designed to quantify the most relevant MDD traits. The variables measured were clustered into 11 components. Some were associated with dogs' demography, some with impulsivity scores from the Dog impulsivity assessment scale, and some with success measures.
A cognitive bias test assessed the dogs' tendency to make decisions over ambiguity. The dogs' latency to approach ambiguous locations was significantly linked with some test battery components and success measures.
Results indicate that certain traits in MDD are associated with different performance levels and vary across tasks and training stages. The test battery may be useful to predict MDD ability. Further exploration is necessary.
| Date of Award | 24 Jan 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Nicola J Rooney (Supervisor) & Michael T Mendl (Supervisor) |