Why doesn't it work?

Same benefits, different communication patterns: Comparing Children's reading with a conversational agent vs. a human partner

“…This may be due to children's perceptions of the CAs' listening ability: Studies have suggested that people are likely to talk more clearly and slowly if they perceive their partners as needing additional support in interpreting their utterances (Rooy, 2009). This pattern was also identified in children's communication with CAs: Young children adjusted their speech style if they perceived CAs as encountering difficulties in understanding them (Beneteau et al, 2019;Cheng, Yen, Chen, Chen, & Hiniker, 2018).…”

“…However, the CAs' reliance on speech may actually be positive, since this reliance-once understood by children-encourages children to practice verbal communication that is vital for their language development Xu & Warschauer, 2020b. Indeed, two studies found that preschool-aged children made efforts to have their speech understood by CAs; they adjusted sentence structures, modified word choice, or spoke more articulately (Beneteau et al, 2019;Cheng, Yen, Y. Chen, S. Chen, & Hiniker, 2018).…”

Conversational User Interfaces on Mobile Devices

“…Text entry [9,38,67,71,82,85,93] Application control [22,90,92] Speech analysis [26,36,52,68,72] Spoken output [28,31,66,77] Conversational Agents [33,43,45,48,55,61,96] Probes [18,74] Domains Education [22,43,52,61,68,77] Health & Well-being [26,28,36,71,72,92,94] Transportation [33,38,55,67,90] Gaming [18,26,36,52,66,…”

“…This theme comprises of two papers, where a technology probe using and about CUIs was deployed. Cheng et al [18] examined conversational repair strategies to correct communication breakdowns with a voice-driven interface. In this study, they have examined the repair strategies [7] that young children face when they play a game using voice interaction, and the system deliberately fails to recognise their utterance.…”

“…These papers highlight that using a conversational agent deployed as technology or cultural probe is a highly underutilised application of CUI. While there have been some work on probes for potential CUI work, such as [62], and extensive work on cultural probes that employ digital photography or audio recording [11,34,42] the work done by Cheng et al [18] and Porcheron et al [74] can be seen as an interesting turn towards using mobile CUIs for examining both people interacting with speech-based systems, but also how people act around, and outwith the technology we tend to focus on. Adapting mobile CUIs to be proactive in their interactions with participants in combination with the awareness of the mobile devices' contexts holds a further opportunity to probe specific social and cultural practices using natural conversation, instead of relying on participants to editorialise their experiences in textual or visual form, or submit to more invasive observational methods.…”