Human bias shapes animal behavior

The study examined experienced dog and wolf trainers engaging in contact-based interactions with dogs and wolves that had been raised in the same way and habituated to humans. The team first assessed the trainers’ conscious (explicit) and unconscious (implicit) attitudes, and then recorded and analyzed physiological responses - such as heart rate variability - as well as the behavior of both humans and animals during standardized sociability tests.

The study was conducted at the Konrad Lorenz Institute of Ethology (KLIVV) at the Vetmeduni, in cooperation with the University of Vienna and the Core Facility Wolf Science Center (WSC).

Conscious and unconscious attitudes matter

Key findings: Explicit and implicit attitudes were not correlated and acted independently. When interacting with trainers who held generally positive conscious attitudes toward wolves, the wolves showed more affiliative behavior and less discomfort. Trainers who, in the Implicit Association Test, favored dogs over wolves exhibited lower heart rate variability when interacting with both species - an indicator of higher physiological arousal. These changes were associated with wolves’ affiliative behavior.

Wolves were more sensitive than dogs

Although the trainers had long-term, comparably close familiarity with both species, they felt more positively about dogs and behaved more openly toward them. Wolves were more sensitive than dogs to the trainers’ attitudes and physiological state: increases in heart rate and heart rate variability (more movement, more relaxation) led wolves to show greater affiliative behavior, while dogs were comparatively less affected.

“Even when procedures are standardized and researchers are blind to the hypotheses, our inner stance can ‘leak’ into tests and change animals’ behavior,” says first author Svenja Capitain of the University of Veterinary Medicine Vienna. “Animals read us remarkably well—through words, actions, and subtle physiological signals.” Senior author Friederike Range (Vetmeduni) adds: “Comparative studies often attribute differences to domestication or cognition. However, if dogs and wolves respond differently to human attitudes and body signals, presumed species differences may in fact be driven by human influences. Research needs methods that better exclude the impact of involved personnel - beyond standard blinding.”

Implications for research and practice

The results point to an additional, previously underappreciated source of bias in animal studies - beyond the well-known expectancy effect. If attitudes differ by species, population, breed, or even individual, apparent differences in animals’ human-directed behavior can be unconsciously amplified or attenuated. This is practically relevant as well: subtle differences in human state and demeanor can shape animals’ reactions in everyday contexts, such as veterinary practice, pet ownership, the training of service and assistance dogs, in zoos, or during encounters with wildlife. To obtain more reliable results, the authors recommend measuring and reporting the attitudes and - where possible - the identity of experimenters; further standardizing and documenting human behavior; reducing direct human presence or automating procedures; and replicating studies across sites.

Note: The paper presented is part of a research project funded by the Austrian Science Fund (FWF) between the Domestication research group at the University of Veterinary Medicine Vienna, led by Friederike Range, and psychologists at the University of Vienna under the leadership of Claus Lamm (Department of Cognition, Emotion, and Methods in Psychology, Faculty of Psychology), aiming to better understand the factors underlying the human–wolf relationship.

The article "Invisible attitudes have visible effects on the behaviour of animals in scientific studies" by Capitain, S., Wirobski, G., Guran, CN.A. et al.was published in Animal Cognition. 
 

Scientific article

 

Scientific contact:
Svenja Capitain MSc.
Konrad-Lorenz-Institut für Vergleichende Verhaltensforschung (KLIVV)
Veterinärmedizinische Universität Wien (Vetmeduni)
Svenja.Capitain@vetmeduni.ac.at