MCI Paper Session 4 – Privacy, Security and Trust

After being placed into a position, it is common for new employees to be acclimated to an organization’s culture, rules, and procedures via a process called onboarding. These processes are critical to ensure that employees become valuable assets to an organization and abide by organizational rules and procedures.

In this research study, we interviewed senior undergraduate students who had recently completed internships to determine what, if any, onboarding process they completed for their placement. Applying qualitative analysis, we find that the onboarding processes for these interns varied widely, from no onboarding at all to several extensive training sessions. Similarly, some interns reported high-level technical security training, while others reported almost no restrictions while on organizational networks. We build on our findings by providing recommendations for organizational improvements for interns, and by extension, full-time employees.

We analyze the addressee detection task for complexityidentical dialog for both human conversation and devicedirected speech. Our recurrent neural model performs at least as good as humans, who have problems with this task, even native speakers, who profit from the relevant linguistic skills. We perform ablation experiments on the features used by our model and show that fundamental frequency variation is the single most relevant feature class. Therefore, we conclude that future systems can detect whether they are addressed based only on speech prosody which does not (or only to a very limited extent) reveal the content of conversations not intended for the system.

Presence detection systems are becoming more and more important and are used in smart home environments, in the Ambient Assisted Living (AAL) domain or in surveillance technology. Common systems focuses on using motion sensors or cameras, which have only a limited viewing angle and therefore monitoring gaps can easily occur within a room.

Humans produce carbon dioxide (𝐶𝑂2) through their respi- ration, which is distributed in rooms. As a result, if one (or more) persons are in a room, a significant increase in 𝐶𝑂2 concentration in the room can be noted. With this work we investigate an approach to detect the presence or absence of people indoors by monitoring the 𝐶𝑂2 concentration in the ambient air.

This paper reports on an in-depth study of electrocardiogram (ECG) biometrics in everyday life. We collected ECG data from 20 people over a week, using a non-medical chest tracker. We evaluated user identification accuracy in several scenarios and observed equal error rates of 9.15% to 21.91%, heavily depending on 1) the number of days used for training, and 2) the number of heartbeats used per identification decision. We conclude that ECG biometrics can work in the wild but are less robust than expected based on the literature, highlighting that previous lab studies obtained highly optimistic results with regard to real life deployments. We explain this with noise due to changing body postures and states as well as interrupted measures. We conclude with implications for future research and the design of ECG biometrics systems for real world deployments, including critical reflections on privacy.

Despite privacy and security concerns, personal data on smartphones could be of beneficial use to society, for example during national emergencies. User attitudes were collected through a small focus group approach to reveal what citizens’ opinions may be towards extraction of medical data in the event of a public health incident. Thematic analysis revealed four themes with an overarching theme of “my phone is part of my body”. This small-scale proof of concept study established individuals view smartphones akin to organs or limbs, where forced access, without consent, is assault. They consider the benefits to society of unconsented access to medical and other personal data on mobile devices must be overwhelming before such acquisition is considered acceptable. The analysis also points to the difficulty of gaining consent, a lack of knowledge about legal aspects, and a distrust about the state collecting data.

Der vorliegende Artikel beschäftigt sich mit Einflussfaktoren auf die Entstehung von Technologieakzeptanz am Beispiel der Sprachsteuerung von Fernsehgeräten. Im Mittelpunkt stehen hierbei drei nutzerspezifische Faktoren: Persönlichkeit (Big5), Technikaffinität und Risikobewusstsein. Die Forschungsfrage lautet, ob der Einbezug von diesen drei nutzerspezifischen Faktoren einen Mehrwert gegenüber bisher in der Theorie bestehenden Modellen liefert.

Es wurde eine repräsentative quantitative Onlinebefragung durchgeführt (N=440), um nutzerspezifische Variablen und die individuelle Akzeptanzausprägung zu erheben. Eine hierarchische Regressionsanalyse ergab zwei prägnante Ergebnisse. Zum einen wurde ersichtlich, dass die Akzeptanz der Sprachsteuerung von Fernsehgeräten von der Ausprägung des Risikobewusstseins sowie der Offenheit für neue Erfahrung und der Gewissenhaftigkeit der Befragten abhängt. Technikaffinität hingegen hatte keinen signifikanten Einfluss auf die Akzeptanz. Zum anderen erklärten die nutzerspezifischen Einflussfaktoren nur einen geringen Varianzanteil der Akzeptanz (5.9%), wohingegen die Nutzenerwartung einen deutlich höheren Varianzanteil erklären konnte (54.9%). Dies steht im Einklang mit bisherigen Forschungsergebnissen und unterstützt die Annahme, dass nutzerspezifische Einflussfaktoren einen Mehrwert in der Akzeptanzforschung liefern, dieser jedoch nicht sehr groß ist.

Online fundraising is becoming increasingly important for non-profit organizations, but the factors that convince people to make a donation online have not yet been fully investigated. In the present work, data of two studies (total N = 2525) was used to examine factors of online donation. An effect of website design perceptions (in terms of content, usability, and aesthetics) on willingness to donate was postulated. Furthermore, research questions about demographic aspects such as age and gender as well as trust in the organization were posed. In both studies, fully-functional health-related websites were evaluated (m = 30 in study 1, m = 3 in study 2). For statistical analysis, logistic regressions were performed.

The results showed different relevance of predictors of donation in different scenarios of donation. For the donation of one’s own money, the perceived content and trust in the organization showed the greatest influence. When donating other people’s money, the aesthetics of the website was the most important factor for the willingness to donate.

Thus, the design of the website, but also of trust in the organization are highly relevant. The differences between the two scenarios of online donation might be explained by the increased relevance of the decision, which results from donating one’s own money, and according cognitive processes. Nevertheless, further research should examine influences of additional variables and establishing implications for successful digital donation generation in the healthcare sector.

Organizations provide their employees with decision support systems (DSS) to facilitate successful decision making. However, the mere provision of a DSS may not be sufficient to facilitate beneficial work outcomes because employees often do not rely on a DSS. Therefore, we examined whether users’ trust in a DSS increases positive effects of DSS provision on several core work outcomes (i.e., performance, well-being, and release of cognitive capacities). Moreover, we examined whether trust effects on these work outcomes depend on specific context conditions (i.e., user accountability, distraction, and market dynamics). We tested our hypotheses in a laboratory experiment with N = 201 participants who received assistance by a DSS in a simulated sales planning scenario. In line with our assumptions, trust in the DSS was positively related to users’ performance and well-being. Moreover, the link between trust and strain as well as release of cognitive capacities were qualified by distraction, so that higher distraction diminished these links. No such moderation occurred for user accountability and market dynamics.

This study investigates the impact of avatars on interactions between customers and consultants in remote, online consultations supported by Augmented Reality (AR). Based on past research, we were interested whether the appearance of an avatar and its dynamics affect important factors for online consultations such as social presence, trust in the consultant and perceived customer satisfaction. In particular, we chose avatar size and dynamics (movement/gaze) to compare different avatars in a 2×2 experiment, in which customers wear AR head mounted devices to consult a remotely located consultant in a mock furniture consultation session. Our results show no significant differences in trust and satisfaction, but significantly different levels of perceived social presence for life-sized, dynamic avatars as well as significantly higher co-presence for all life-sized avatars. Additional data from interviews with the participants revealed a clear preference for dynamic avatars over static ones. Based on an analysis of these findings, we make design recommendations and suggest directions for future research.