Seite - 40 - in Critical Issues in Science, Technology and Society Studies - Conference Proceedings of the 17th STS Conference Graz 2018

algorithms set up by handling institutions such as app providers or traffic management units. The resulting enriched data can then be used to generate interdependent recommendations and thus individually recommend actions. We assume, that the whole big data process is heavily dependent on the trust that users place in their applications as well as the trust in data protection and privacy on part of the providers. Both providers of data gathering applications and users need mutual trust to ensure, that the data being put in has maximal quality. This affects the accuracy of the data and the willingness to comply with the suggested actions. Trust is therefore a key element for the algorithms that generate and use big data. Core model To grasp this subject theoretically, we take the Technology Acceptance Model (TAM) (Davis et al. 1989) as a starting point and establish an extended research model, including trust as a mediating variable and various external factors. Figure 3: Research model of our study (own figure) Figure 3 shows the full research model for our study. The integration of trust as a separate variable inside the core model as well as its positioning follow the work of David Gefen et al., who applied a similar core model to their research on trust in online shopping (cf. Gefen et al. 2003, p. 53). Due to the thematical similarity of both Gefen’s and our subject, we decided to adapt the established core model and add specific external factors that match our research. Concerning the term of trust itself, we work with the typology that was established by Harrison McKnight and Norman Chervany (2001). More precisely, we focus on institutional trust (McKnight and Chervany 2001, p. 33) in the shape of trust in national privacy regulations and interpersonal trust (ibid.) in the shape of the relationship between user apps and app providers (which can together be seen as the trustee). External factors The selection process of the external factors was also conducted following the maxim of thematical suitability. Following this pattern, the external factors can be divided into three groups: general personality traits that were found to be relevant such as perceived control and 40