Algorithms for online shopping aim not only to support the customer in finding the product he or she looks after, but also by offering more (similar) choices. Even though the algorithms are able to adjust to individual preferences, they are designed for the benefit of the seller as they invite the customer to select more products. In certain settings, however, not the selection but the de-selection of goods is of utmost importance. In our video-ethnographic research on assisted shopping in health care settings, we found that care personnel supporting people with acquired brain injury (ABI) negotiate and sometimes even reject the ‘purchasability’ of a certain product. These negotiations are based on (institutional) considerations regarding the appropriateness of a given choice and it’s consequences, e.g. if it’s use might be dangerous or the price extends the individual’s budget. Therefore, we shift the focus to the development of assistive technologies aiming for the deselection of goods and explore how assistive technologies can support the deselection of goods. This can be relevant e.g. in regard to health questions or in regard to questions of buying more sustainable products. The paper will firstly, show how we used multimodal analysis to describe negotiation of choices in assisted shopping interaction, and secondly, how these insights were used to build an embodied and conversational dialog system for an assistive shopping robot that helps people to de-select a chosen product. The paper will end by presenting first findings of the analyses of interactions with the robot prototype used to understand the interactional consequences of the assistive technology for shopping interaction.
Period24 Feb 2021
Event titleDigitalizing social practices
Event typeConference
LocationOdensen, Denmark
Degree of RecognitionInternational


  • shopping
  • rejection
  • robots
  • assistance
  • assistive technologies
  • sustainability
  • disability