Social Interaction during Dog and Robot Group Sessions for Nursing Home Residents with Dementia: the Handler Effect

Lonneke GJA Schuurmans, Inge Noback, Jos MGA Schols, Marie-Jose Enders-Slegers

Research output: Contribution to journalArticleAcademicpeer-review


As part of an 8-week intervention study in Dutch nursing homes, we used video-analysis to observe the interaction of psychogeriatric participants with either the handler, the stimulus (dog or robot) or other clients during weekly dog, robot (RAI, robot assisted interventions) and control (human facilitator only) group sessions. Additionally, we measured the initiative of the handler to engage participants. Several baseline characteristics, including dementia severity, neuropsychiatric symptoms and medication usage, were recorded as possible confounders.Participant-handler interaction is increased in all three groups compared to a baseline of no interaction, while inter-client interaction is not. In the dog group participant-handler interaction scores are similar to participant-dog interaction scores, while in the robot group participant-handler interaction scores are significantly lower than participant-robot interaction scores. Handler initiative does not differ between the three groups.Our results suggest that a handler effect of AAI on social interaction in dementia care does exist and we hypothesize this effect is linked to the required fully embodied, mutual attunement between dog and handler and between dog-handler team and participants. This embodied interaction distinguishes AAI from RAI and when the required attunement is met, AAI can significantly increase the social interaction of people with dementia.
Original languageEnglish
JournalHuman-animal interaction bulletin
Publication statusPublished - 1 Dec 2021


  • nursing
  • dog
  • robot
  • social interaction
  • dementia


Dive into the research topics of 'Social Interaction during Dog and Robot Group Sessions for Nursing Home Residents with Dementia: the Handler Effect'. Together they form a unique fingerprint.

Cite this