Assessment, Trust, Transparency and Chatbots
In this presentation we discuss a work in progress, the use of analytics produced by a chatbot in a university setting (Victoria University, Australia) to explore how students understand trust and transparency. How do they discuss these closely linked concepts and what do students want to know in order to form an opinion about trustworthiness? Explanations need to be tailored to tailored to learner’s needs, priorities, previous knowledge (Miller et al., 2017). And important to those who are developing technology that allows a series of questions, "being an explanation" is not a property of a statement, it’s about the interaction (Hoffman et al. 2018). According to our participants, assessment-driven anxiety is in the foreground of the university experience. The pandemic has amplified the pressure on students. Our next step is to undertake quantitative ethnography (coding transcripts) and to create network diagrams to show how automated systems can be made 'human-aware' (Kambhampati 2020).
Speakers:
Natasha Dwyer (Speaker) Victoria University, Senior Lecturer
Hector Miller-Bakewell (Speaker)
Stephen Marsh (Speaker)
Tessa Darbyshire (Speaker)
