Ethnographic Observation of Child-Robot Interaction

• In Event: 28.026 - Designing Robots for Children: Multisensory Interactions and Multimodal Assessments

Abstract

This presentation will provide an overview of ethnographic methods of participant observation and discuss their use in studies that involve communication and collaboration between children and humanoid robots. The presentation will i) review key principles and features of ethnographic participant observation, ii) discuss key theoretical frames that can be used to interpret communicative interactions within robot-mediated communicative activities involving young children, and iii) provide analytical examples from an on-going study of how humanoid robots can facilitate cross-cultural and bi-lingual collaborative interactions between kindergarten aged-children from Spanish and English dominant families.

Participant observation is an essential approach in ethnographic field work, involving immersive participation and fine-grained observation of particular settings in order to better understand beliefs, practices, behaviors, and interactions within specific contexts (Spradley, 1980). The focus in these observations is on participant meanings and local knowledge—what ethnographers have termed emic—and “thick description” (Geertz, 1973); researchers must set aside their own beliefs and assumptions about what is “going on” in setting and avoiding foisting external meanings on the context. Ethnographic observations are largely inductive; the goal is not to test hypotheses developed a priori, but allow theories and hypotheses to emerge from the setting, which will become working hypotheses that will be developed and refined through further observations (Erickson, 1986).

Participant observation occurs along a continuum, and researchers may decide to take a more passive role in settings as they observe—acting essentially as a bystander at a scene—or a more active one, where they are seeking to understand cultural rules and norms of behavior by participating in activities of setting, (Spradley, 1980). Most classroom studies involve researchers as moderate participants, where they are peripherally involved and maintaining a balance between roles as participant and observer. Researchers may audio or video record observations—and this would be essential in studies examining in-depth communication and dialogue—along with preparing descriptive field notes from observations, which would include portraits of participants, descriptions of setting, and accounts of particular events (Emerson, Fretz, & Shaw, 2011). Initial observations are highly descriptive, whereas later observations become more focused. Researchers may also prepare reflective, methodological, and analytical notes and memos.

A useful interpretive frame for making sense of interactions among humanoid robots and children include the ethnography of communication, which views language as occurring within socially contextualized speech communities and settings (Hymes, 1974); the focus of analysis of ethnographic data is on speech situations, events, and acts, which can help researchers identify roles of interlocuters, types of and norms for communication, and how certain communicative practices function within the setting. Observations of children also need to be mindful of issues of power—among children, between children and adults, and among adults—and reflect on one’s own positionality as an “adult” among children (Corsaro, 2012; Eder & Corsaro, 1999; Thorne, 1993). Researchers must authentically look and listen to children and recognize children’s active production of knowledge and peer culture (see Corsaro’s theory of interpretive reproduction, Corsaro, 2012).

Authors

• Laura Ruth Johnson, Northern Illinois University

• Karla Vilchez Neal, Northern Illinois University