Research

Smartphone app designers often use behavioural design to influence users, increase sales, and boost advertising revenue. Behavioural design relies on elements ranging from app appearance to black-box algorithms and personalization. It commonly exploits behavioural biases, such as the lack of self-control. Consumers are seldom aware of such design and usually have no control over it. Aiming to protect consumers, the recently enacted European Digital Services Act requires app design to be more transparent and adjustable.

In a framed field experiment, we document that behavioural design increases app usage time, especially in the case of vulnerable users. An app version that adds transparency and offers protection features helps to overcome temptation. The higher willingness to pay for the transparent version shows that the positive effects of app transparency and increased consumer protection might not only materialize on the demand side but may also challenge current practices on the supply side.

A crucial building block of responsible artificial intelligence is responsible data collection and governance. Its importance is also underlined in the latest EU regulations. The data should be of high quality, foremost correct and representative. In addition, individuals providing the data should have autonomy over what data is collected. In this paper, we consider the setting of collecting personally measured fitness data (physical activity measurements), in which individuals may not have an incentive to measure and report accurate data. This can significantly degrade the quality of the collected data. On the other hand, high-quality collective data of this nature could be used for reliable scientific insights or to build trustworthy artificial intelligence applications. We conduct a framed field experiment (N = 691) to examine the effect of offering fixed and quality-dependent monetary incentives, on the quality of the collected data. We use a peer-based incentive-compatible mechanism for the quality-dependent incentives without spot-checking or surveilling individuals. We find that the incentive-compatible mechanism can elicit good data while providing a good experience and compensating fairly. Although, in the specific study context the data quality does not differ under the two incentive schemes. We contribute new design insights from the experiment and discuss directions that future field experiments and applications on explainable and transparent data collection may focus on.

Der Trendbegriff »Dark Patterns« hat Konjunktur. Er ist ebenso omnipräsent wie interpretationsoffen. Designmuster, die den Nutzer in eine bestimmte (verbraucherschädigende) Richtung lenken sollen, stellen Wissenschaft und Regulierungsbehörden sowie den Gesetzgeber vor besondere Herausforderungen. Um Dark Patterns in all ihren Facetten zu beleuchten und Gefahren für Verbraucher wirksam abzuwehren, bedarf es eines interdisziplinären Schulterschlusses der Verhaltensökonomie, Rechtswissenschaft und Informatik.

Der Beitrag analysiert den Forschungsstand rund um das Thema Governance von und durch Verhaltensdesign in digitalen Medien und Anwendungen, basierend auf den Erkenntnissen aus mehreren Feldstudien. Die Feldstudien, die er vorstellt, wurden interdisziplinär aufbereitet durch die fünf Linsen der Rechtswissenschaften, Ethik, Informatik, Technikfolgenabschätzung und Verhaltensökonomie. Anhand der gewonnen Einsichten werden generelle Implikationen auf die Verbraucherforschung, die Gesellschaft und der Demokratie diskutiert. “Bequeme Verbraucher” werden durch eine verantwortungsvolle Anwendung von Verhaltensdesign herausgefordert, zum “lernenden, partizipativen Verbraucher” zu avancieren.

Durch Bias im Entwicklungsprozess können Softwareanwendungen, Algorithmen, KI und automatisierte Entscheidungssysteme unbeabsichtigt das Verhalten ihrer Nutzer beeinflussen und ungewollte Effekte hervorrufen.

Reading newsfeeds gets more and more popular, although digital media contain an increasing number of hidden risks for users, such as the application of design elements that influence behaviour. In order to study the influence of such design elements, we investigate reading behaviour in the present study, which is a scoping study for a follow-up case study implementing a newsfeed reader app. Using a survey, we ask potential users of a newsfeed reader app about their reading habits and needs. Moreover, we ask participants to assess their risk perception of digital media with regard to the YouTube and Facebook apps. The overall 77 participants of the study were recruited at the Ruhr-University Bochum. We used an online survey to collect data. For the analysis of the descriptive data, we used figures and tables. The participants make up a demanding and conscious readership. Most of them read several times a day and often in between. They are conscious about the risks of digital and social media. An app design that aims to make readers reading more often has good chances to succeed in increasing reading frequency and thus also total reading time unless it contains more than one or two push notifications a day. We also recommend to build in social aspects, such as comparison of reading performance, and possibilities to discuss or rate articles. All-round but still transparent filtering options, a possibility to self-select news content, speed reading, text-to-speech, and reading aid functionalities are desirable. Distracting elements, such as gamification, ads, tabloid press and opinion bubbles should be avoided.

We study the effect on coordination in a minimum-effort game of a leader’s gender depending on whether the leader is democratically elected or is randomly-selected. Leaders use non-binding messages to try to convince followers to coordinate on the Pareto-efficient equilibrium. We find that teams with elected leaders coordinate on higher effort levels. Initially, the benefits of being elected are captured solely by male leaders. However, this gender difference disappears with repeated interaction because unsuccessful male leaders are reelected more often than unsuccessful female leaders.

In this paper, we reveal gender using avatars.

We study gender differences in the behavior and effectiveness of randomly selected leaders in a laboratory experiment using the minimum effort coordination game. Leaders can send non‐binding numeric messages to try to convince followers to coordinate on the Pareto‐efficient equilibrium. The treatment variations consist of the gender of the leader, and whether participants know or do not know the gender of the leader in their group. We find that female leaders choose more often to send a riskier high message in the beginning of the game, which hurts their effectiveness especially if gender is not revealed. However, if gender is revealed, both male and female leaders make more careful choices, and thus we do not observe any significant gender difference in leader effectiveness.

We study gender differences in the behavior, communication, and effectiveness of randomly selected leaders in a laboratory experiment using the turnaround game. Leaders can send nonbinding pre‐play text messages to try to convince followers to coordinate on the Pareto‐efficient equilibrium. The treatment variations consist of the gender of the leader, and whether the communication is one‐way (only leaders send messages) or two‐way (first followers send messages to their leader, and subsequently the leader sends messages to the group). We find that male leaders communicate more assertively. Communication with the followers induces female leaders to express significantly more often that they are part of the group, rather than standing above the group. Despite the different paths in communication, both men and women are equally likely to request the highest effort contribution. Men and women are equally effective leaders.

Whenever a common goal requires coordinated action of a team, leadership facilitates the endeavor. Up to date, women are still a minority in top-level business positions. Using variations of the minimum effort coordination game in laboratory experiments, we investigate the effectiveness of male and female leaders. The first study shows with a basic setup that men and women are equally effective leaders, no matter if gender is revealed or not. The second study considers communication styles and finds that despite the different paths in communication, men and women are equally effective leaders. The third study varies the selection process and presents a replication of the democracy effect on leader effectiveness using the turnaround game as studied earlier by Brandts et al. (2015). Moreover, we find that elected women are marginally less effective leaders than elected men, although the gender difference disappears with repeated interaction. The broad conclusion is that organizations can benefit from both employee participation in choosing group leaders and reconsidering gender imbalance in top-level positions.