Read past
the headline.
Sources support stated points. They do not endorse Shared Air or provide independent veterinary review of this publication.
Scientific Reports
The 2024 study tested eighteen dogs and odours from three donors. Its reported approach effect depended on a specific session and bowl position. It does not diagnose an individual dog's feelings.
Open the source ↗University of Bristol
A university explanation of the same 2024 study. Its broad headline is read alongside the more specific methods and results in the paper, not counted as independent replication.
Open the source ↗Royal Society Biology Letters
The 2016 study used a cross-modal looking task with seventeen dogs. Matching emotional valence across faces and voices is not direct access to a private human thought.
Open the source ↗University of Lincoln
A university explanation of the same 2016 study, with the unfamiliar faces and voices described. This is a related account, not another experiment.
Open the source ↗PLOS ONE
The 2022 study involved four trained dogs, samples from thirty-six human participants and a baseline-versus-stress odour task. Discrimination in that task is not a diagnosis by an untrained pet.
Open the source ↗Queen's University Belfast
A university explanation of the same 2022 scent study. Possible service applications are not proof that every pet already performs a trained alert.
Open the source ↗Dogs Trust
Welfare guidance on the whole body, context, individual differences and space. It is not a single-cue emotion dictionary. Local UK legal statements are not presented as worldwide rules.
Open the source ↗One study is not two
A university release and the paper it describes are related accounts, not independent replications. The 2016, 2022 and 2024 tasks are kept separate here.