Currently, more and more learning environments exist in whole or in part online. Although these learning environments can be scaled to large audiences, the lack of interaction between educators and students is difficult to replace and remains a primary challenge in the field. This is why Smart Personal Assistant Technology (SPAT) has received attention in the field. SPAT describes computer systems intended to converse with humans. These systems may alleviate the problem of individual support in online and blended learning environments by engaging in natural interaction with the students. Although SPAT seems to be promising for education, three important challenges related to SPAT design emerge and will be tackled in this dissertation. Research Challenge (RC) 1: There is limited knowledge about existing SPAT and corresponding requirements of SPAT in online and blended learning environments. RC2: There is limited knowledge of how to design SPAT to improve students learning outcomes. RC3: It is not yet clear how non-tech-savvy educators can design and use SPAT on their own to offer individual support. This dissertation tackles these RCs by following a multi-method approach that identifies research gaps and requirements of SPAT in online and blended learn-ing environments in order to propose design principles (DPs) and design knowledge regarding how educators can use the technology. The dissertation has three main core findings. First, the findings show that there is a lack of consensus on how to design and develop SPAT for online and blended learning environments. Second, based on the first finding, this study proposes five DPs on how to design SPAT in these contexts and investigate the effectiveness of these principles in a series of laboratory and field experiments across different school levels (secondary education, higher education) and school types (high school, vocational business school, university). The findings show that these DPs can have a positive influence on students learning processes resulting in increased students learning outcomes. It also addresses the third RC by developing a SPAT Platform that empowers non-tech-savvy educators to create their own SPAT in order to make this technology accessible to educators who are not technically inclined. The dissertation contributes to a theory of design and action that explains how non-tech-savvy educators can design effective SPAT in their online and blended learning environments.
20 Sep 2021