Substantive use of artificial intelligence: The role of individual differences

Klesel, Michael; Messer, Uwe Working Paper Substantive use of artificial intelligence: The role of individual differences Working Papers, No. 32 Provided in Cooperation with: Faculty of Business and Law, Frankfurt University of Applied Sciences Suggested Citation: Klesel, Michael; Messer, Uwe (2024) : Substantive use of artificial intelligence: The role of individual differences, Working Papers, No. 32, Frankfurt University of Applied Sciences, Fachbereich 3: Wirtschaft und Recht, Frankfurt a. M., https://doi.org/10.48718/8d9d-b049 This Version is available at: https://hdl.handle.net/10419/306858 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Substantive Use of Artificial Intelligence: The Role of Individual Differences Michael Klesel Uwe Messer Working Papers Fachbereich Wirtschaft und Recht Frankfurt University of Applied Sciences www.frankfurt-university.de/fb3 September 2024 ISSN-Nr. 2702-5802 DOI:https://doi.org/10.48718/8d9d-b049 Nr. 32 Fachbereich 3 Wirtschaft und Recht | Business and Law 1 Das Urheberrecht liegt bei den Autor*innen. Working Papers des Fachbereichs Wirtschaft und Recht der Frankfurt University of Applied Sciences dienen der Verbreitung von Forschungsergebnissen aus laufenden Arbeiten im Vorfeld einer späteren Publikation. Sie sollen den Ideenaustausch und die akademische Debatte befördern. Die Zugänglichmachung von Forschungsergebnissen in einem Fachbereichs Working Paper ist nicht gleichzusetzen mit deren endgültiger Veröffentlichung und steht der Publikation an anderem Ort und in anderer Form ausdrücklich nicht entgegen. Working Papers, die vom Fachbereich Wirtschaft und Recht herausgegeben werden, geben die Ansichten des/der jeweiligen Autor*innen wieder und nicht die der gesamten Institution des Fachbereichs Wirtschaft und Recht oder der Frankfurt University of Applied Sciences. Bitte zitieren als: Klesel, Michael; Messer, Uwe (2024): Substantive Use of Artificial Intelligence: The Role of Individual Differences. Working Paper Nr. 32 des Fachbereichs 3 Wirtschaft und Recht. Frankfurt University of Applied Sciences. https://doi.org/10.48718/8d9d-b049 Frankfurt University of Applied Sciences Fachbereich 3: Wirtschaft und Recht Nibelungenplatz 1 2 Abstract English Artificial intelligence (AI) is becoming increasingly powerful, enabling users to perform tasks more efficiently and effectively. However, not all users are equally able to take advantage of its capabilities. We draw on previous literature that has introduced the concept of “substantive use” – the reflective consideration of how to use a system's features – to better understand individual differences in the context of AI. We contribute to the current literature in three ways: First, we summarize the literature on technology use and describe its relevance for AI-related research. Second, we review the literature and show that IS has already begun to investigate individual differences to understand the use of AI systems. Third, we propose a theoretical model that accounts for the direct and configurational effects of individual differences on substantive use behavior. Abstract Deutsch Künstliche Intelligenz (KI) wird immer leistungsfähiger und ermöglicht es den Nutzern, Aufgaben effizienter und effektiver auszuführen. Allerdings sind nicht alle Benutzer gleichermaßen in der Lage, diese Fähigkeiten zu nutzen. Wir stützen uns auf frühere Literatur, die das Konzept "substantive use" eingeführt hat – die Überlegung, wie die Funktionen eines Systems genutzt werden können –, um individuelle Unterschiede im Kontext von KI besser zu verstehen. Wir tragen auf drei Arten zur aktuellen Literatur bei: Erstens fassen wir die Literatur zur Nutzung von Technologie zusammen und beschreiben ihre Relevanz für die KI-bezogene Forschung. Zweitens sehen wir die bestehende Literatur durch und zeigen, dass die Forschung im Bereich Wirtschafsinformatik bereits damit begonnen hat, individuelle Unterschiede zu untersuchen, um die Nutzung von KI-Systemen zu verstehen. Drittens schlagen wir ein theoretisches Modell vor, das die direkten und konfigurativen Auswirkungen individueller Unterschiede auf „substantive use“ berücksichtigt. 3 Inhaltsverzeichnis / Table of Contents 1. Introduction ................................................................................................................ 4 2. Related Work .............................................................................................................. 4 2.1. Conceptualizations of system use construct ........................................................... 4 2.2. System Use with AI technologies ............................................................................ 5 2.3. Individual Differences ............................................................................................ 7 3. Methodology ............................................................................................................... 8 3.1. Structured Literature Review .................................................................................. 8 3.2. Results ................................................................................................................. 9 4. Discussion ................................................................................................................ 10 5. Outlook .................................................................................................................... 11 6. References ............................................................................................................... 12 Abbildungsverzeichnis Figure 1: An Example of an AI-based Fraud-Detection Dashboard ........................................... 6 Figure 2: An Integrated Model of Substantive Use and Traits .................................................. 10 Tabellenverzeichnis / List of Tables Table 1: Overview Conceptualizations of Technology Usage ................................................... 5 Table 2: Group characteristics ............................................................................................. 8 Table 3: Results of the Literature Review ............................................................................... 9 4 1. Introduction It is hard to overstate the impact of machine learning (ML) and artificial intelligence (AI) in academia and in practice. In Information Systems (IS) research, several AI-related special issues have been published (Berente et al., 2021; Benbya, Strich and Tamm, 2024), and all leading conferences, including the “International Conference on Wirtschaftsinformatik”, have addressed the impact of AI in various domains, such as its impact on human-computer-interaction. A fundamental promise of current advances of AI is the increase in performance. From a technological perspective, this progress is well documented. Modern machine learning models exceed the performance of previous generation by far. For example, the current version of GPT outperforms its predecessors in many ways. Similarly, the current literature has shown how ML and AI can be used to create business value (Shollo et al., 2022). Despite all the optimism surrounding AI, the IS literature to date has repeatedly shown that technology cannot be used equally by all individuals. In fact, most re-search suggests that there are significant individual differences when it comes to technology use. For example, studies have shown that individual traits such as being mindful with IT, are significantly related to technology adoption (Thatcher et al., 2018). However, our literature review (see Section 3) shows that there is a paucity of research that has integrated the effects of individual differences with substantive use of technology with AI. As a result, there is a lack of theoretical knowledge to inform how to promote substantive use of AI or how to support individuals who struggle to do so. We seek to address this important issue and contribute to the existing literature in three ways: First, we review the existing literature on technology use and identify the conceptualization of substantive use as the most promising for theorizing in the field of AI. Second, we consolidate the current literature on individual difference traits and identify those most relevant to AI research. Finally, we propose a conceptual frame-work that allows scholars to study substantive use behavior for AI. Integrating previous literature, our conceptual model integrates two mechanisms: 1) isolated impact of individual difference traits on substantive use, and 2) configurational impact of individual difference traits on substantive use. The paper is structured as follows: In section 2.1, we review previous literature that has conceptualized system use in order to identify the most promising conceptualization for the domain of AI. In section 2.2, we demonstrate the suitability of substantive use of technology through an example in the application of Explainable AI (XAI). Section 3.3 is devoted to previous literature on individual differences. In section 3, we will present the methodological approach used to review the current literature on the intersection of AI use and individual differences. We discuss the results of this study in section 4 and provide an outlook for future research in section 5. 2. Related Work 2.1. Conceptualizations of system use construct The “(system) usage” construct is arguably one of the most widely disseminated construct in IS research. It is the fundamental dependent variable in the technology acceptance and adoption stream (Venkatesh et al., 2003) and has been applied in various domains and with different applications (Venkatesh, Thong and Xu, 2016). It has also undergone various reconceptualizations that recognize the richness of the construct. For instance, it has been shown that system use can be measured with different degrees of richness, recognizing three domains: user, task, and system (Burton-Jones and Straub, 2006). This conceptualization has paved the way for more comprehensive conceptualizations of (system) use. Other studies have suggested that technology use is best understood when it is conceptualized as interaction behavior, which has been conceptualized as a use-related activity (Barki, Titah and Boffo, 2007). Others have emphasized how features of a 5 particular technology are used (Jasperson, Carter, and Zmud, 2005), with a particular focus on how users change their IT use after the adoption phase (Bagayogo, Lapointe and Bassellier, 2014). To better understand how and why users use specific features of an information system, the concept of adaptive use has been proposed and evaluated (Sun, 2012). More recently, it has also been argued that the use construct should be studied beyond a specific domain, such as the work domain, and instead should be conceptualized as an overarching construct that spans multiple domains. For this reason, the notion of transgressive use has been suggested (Klesel et al., 2017). An overview of conceptualizations of technology use is provided in Table 1. Concept Method Technology Substantive use behavior (Jasperson, Carter and Zmud, 2005) Conceptual No specific technology System usage (Burton-Jones and Straub, 2006) Survey Microsoft Excel IS-related activity (Barki, Titah and Boffo, 2007) Survey No specific technology Adaptive use (Sun, 2012) Survey Microsoft Office Enhanced use (Bagayogo, Lapointe and Bassellier, 2014) Grounded Theory No specific technology Transgressive use (Klesel et al., 2017) Case Study No specific technology Table 1: Overview Conceptualizations of Technology Usage Most studies that have examined the nature of system use have taken a technology-agnostic perspective or have focused on Microsoft Office products such as Microsoft Excel. While these conceptualizations arguably have a different emphasis on specific aspects, most concepts allow for a more nuanced perspective on how individuals use technology with respect to specific features. This is made very explicit in the notion of substantive use (Jasperson, Carter and Zmud, 2005), adaptive use (Sun, 2012), and the notion of enhanced use of technology (Bagayogo, Lapointe and Bassellier, 2014), where the authors examine specific features of a class of systems. While we acknowledge that there is an ongoing discourse on the conceptualization of one of the most fundamental constructs of the IS discipline, we also note that there is only little research available that has re-evaluated the suitability of current conceptualizations with modern technologies such as AI. For this reason, we will now review why AI is a class of systems that requires a contextualized form of the use construct. 2.2. System Use with AI technologies In a number of studies, authors have argued that AI systems have distinct characteristics that distinguish them from existing systems. Research has also shown that people confronted with AI are often influenced by what is known as “algorithm aversion” (Turel and Kalhan, 2023). Consequently, 6 it has been argued that new ways of managing AI is needed to realize the potential of AI technologies in organizations (Berente et al., 2021). IT management can be studied from different levels, which means that the investigation of AI can also be conducted with a focus on the entire organization or at the individual level. While there is a rich body of knowledge that investigates IT management at the firm level (Li et al., 2021), most research on adoption takes an individual perspective. In particular, research that focusing on human-computer-interaction is typically conducted at the individual level. Therefore, in the following, we also focus on the individual level. AI technologies are ubiquitous and can be found in a wide range of applications. Therefore, we use a specific human-computer interface (HCI) example which is used to demonstrate that a specific contextualization of the use construct is required. We use a simplified AI-based fraud detection dashboard that allows users to identify fraudulent documents (see Figure 1). While this example is simplified, it contains all basic and necessary components. It includes a machine learning model (i.e., for classification) and an Explainable AI (XAI) component that allows users to learn more information about the decision made by the system. In this case, words that increase the likelihood of a fraud case are highlighted (red, underlined). This type of dashboard has previously been developed and evaluated in the IS literature to study hate speech (Meske and Bunde, 2022), diabetes self-management (Van Der Waa et al., 2021), and signature forgery detection (Hamm et al., 2023). The XAI module is most commonly implemented using SHapley Additive exPlanation (SHAP values) (Lundberg and Lee, 2017; Lundberg et al., 2020). Figure 1: An Example of an AI-based Fraud-Detection Dashboard Given the prior literature on algorithmic aversion, the use of an AI-based dash-board may vary significantly from user to user. For example, a user may completely distrust the system and ignore it altogether. Alternatively, a user may be guided by a so-called automation bias and use the information provided without further elaboration. To conceptualize the use of AI, we draw on previous literature. In particular, we acknowledge the notion that system use is an interplay between a user, a task, and an (AI) system (Burton-Jones and Straub, 2006). Furthermore, we adapt the notion of substantive use (Jasperson, Carter and Zmud, 2005), which is defined as “an individual’s reflective consideration to use a single feature (or a select subset of features) available in an IT application” (Jasperson, Carter and Zmud, 2005, p. 535). In the context of our example, the deliberate use of the XAI component would be considered as an example of substantive use. There are at least two fundamental reasons why we argue that a substantive use of an AI-based system is preferable: First, the literature has shown that engaged behavior leads to positive outcomes. For example, a mindful use of technology has been shown to be positively associated with outcome variables (Thatcher et al., 2018). As a result, substantive use of an AI system is 7 preferable. Second, prior literature has shown that users who are not engaged are more likely to produce errors (Reason, 1990). This is particularly relevant for AI applications, because they are often used to support decision making in high-stakes situations, such as diagnostic decisions in medical contexts (Jussupow et al., 2021). There is preliminary evidence that the use of AI dashboards is influenced by contextual variables. In the field of XAI, it has been shown that there are differences in outcome variables depending on the type of XAI (Van Der Waa et al., 2021). Others have shown that socio-demographic variables are also relevant when it comes to how users engage with AI-based systems (Hamm et al., 2023). However, there is a significant gap in terms of a comprehensive understanding of the relative importance of contextual variables and their impact on AI use. Against this background, we proceed with a review of individual differences that allows to address this gap. 2.3. Individual Differences Individual differences are relatively stable characteristics of individuals that persist across time and context, although stability does not imply that they are unchangeable (Sackett et al., 2017). Differential psychology has traditionally studied individual differences and their assessment. Here, we focus on the most important among these traits. However, determining the number of traits that exist is not a straightforward task. There are more than several hundred traits, some of which have been labeled differently by different researchers and research traditions, even though they are essentially very similar (Cooper, 2019). Trait taxonomies have been developed to organize findings in the field. One of the most well-known models is the Big Five fac-tor model of personality. In this model, personality is divided into five higher order factors, including conscientiousness and agreeableness, which are further divided into lower order facets (Crowe, Lynam and Miller, 2017). The most commonly studied dispositional characteristics include ability, interest, and personality (Lubinsky, 2000). To identify the most important of these traits, we reviewed studies published in leading differential psychology journals, including the “Journal of Personality and Social Psychology”, “Personality and Individual Differences”, and “Psychological Bulletin”. By excluding traits unrelated to technology use (e.g., psychopathological traits, vocational interests, and psychomotor abilities), we narrowed the traits to a manageable number and arrived at a list of six traits (see Table 2). 14 Rabl, T., Petzsche, V., Baum, M., and Franzke, S. (2023), 'Can support by digital technologies stimulate intrapreneurial behaviour? The moderating role of management support for innovation and intrapreneurial self‐efficacy', Information Systems Journal, 33(3), 567-597. Reason, J. T. (1990) Human Error. Cambridge; New York: Cambridge University Press. Sackett, P. R., Lievens, F., Van Iddekinge, C. H., and Kuncel, N. R. (2017), 'Individual differences and their measurement: A review of 100 years of research', Journal of Applied Psychology, 102(3), 254. Salgado, J. F., Anderson, N., Moscoso, S., Bertua, C., and De Fruyt, F. (2003), 'International validity generalization of GMA and cognitive abilities: A European Community meta‐analysis', Personnel Psychology, 56(3), 573-605. Salgado, J. F., Anderson, N., Moscoso, S., Bertua, C., De Fruyt, F., and Rolland, J. P. (2003), 'A meta-analytic study of general mental ability validity for different occupations in the European community', Journal of applied psychology, 88(6), 1068. Schwarzer, R., and Jerusalem, M. (1995), 'Generalized self-efficacy scale', in J. Weinman, S. Wright, & M. Johnston (eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs, 35, 37. Shollo, A. et al. (2022) ‘Shifting Ml Value Creation Mechanisms: A Process Model of Ml Value Creation’, The Journal of Strategic Information Systems, 31(3), p. 101734. doi: 10.1016/j.jsis.2022.101734. Stein, J. P., Liebold, B., and Ohler, P. (2019), 'Stay back, clever thing! Linking situational control and human uniqueness concerns to the aversion against autonomous technology', Computers in Human Behavior, 95, 73-82. Sun (2012) ‘Understanding User Revisions When Using Information System Features: Adaptive System Use and Triggers’, MIS Quarterly, 36(2), p. 453. doi: 10.2307/41703463. Thatcher, J. B. et al. (2018) ‘Mindfulness in information technology use: definitions, distinctions, and a new measure’, MIS Quarterly, 42(3), pp. 831–847. doi: 10.25300/MISQ/2018/11881. Turel, O. and Kalhan, S. (2023) ‘Prejudiced Against the Machine? Implicit Associations and the Transience of Algorithm Aversion’, MIS Quarterly, 47(4), pp. 1369–1394. doi: 10.25300/MISQ/2022/17961. Van Der Waa, J. et al. (2021) ‘Evaluating XAI: A Comparison of Rule-Based and Example-Based Explanations’, Artificial Intelligence, 291, p. 103404. doi: 10.1016/j.artint.2020.103404. VandenBos, G. R. (2007), APA dictionary of psychology, American Psychological Association. Venkatesh, V. et al. (2003) ‘User Acceptance of Information Technology: Toward a Unified View’, MIS Quarterly, 27(3), pp. 425–478. doi: 10.2307/30036540. Venkatesh, V., Thong, J. and Xu, X. (2016) ‘Unified Theory of Acceptance and Use of Technology: A Synthesis and the Road Ahead’, Journal of the Association for Information Systems, 17(5), pp. 328–376. doi: 10.17705/1jais.00428. 15 Bisher erschienene Working Papers des Fachbereich 3: Nr. 31 Bülbül, Dilek (2024): International Cooperation in Research: Cooperative Doctorates in Finance. DOI: https://doi.org/10.48718/gyx4-g297 Nr. 30 Krause, Tobias; Ivanov, Igor; Sidki, Marcus (2023): Blame or Gain? Is Institutional Trust impacted by the Perception of Political In-fluence in State-owned Enterprises? DOI: https://doi.org/10.48718/p69v-1y25 Nr. 29 Anderie, Lutz (2023): NFTs (Non-Fungible Tokens) – Funktion und Potenzialanalyse. DOI: https://doi.org/10.48718/k9ak-b724 Nr. 28 Rosenbusch, Christoph et al. (2023): Building a European University Consortium: the Case of the U!REKA-Network. DOI: https://doi.org/10.48718/je0y-9q78 Nr. 27 Anderie, Lutz; Hönig, Michaela (2023): Untersuchungen zum Potenzial von Metaverse. DOI: https://doi.org/10.48718/6xxa-c637 Nr. 26 Lämmlein, Barbara; Gerdiken, Ulrike (2002): Von Bigband bis Urban Gardening. Motivationale Gründe Studierender für ein kulturelles Engagement an Hochschulen. DOI: https://doi.org/10.48718/twym-pw08 Nr. 25 Jung, Constantin (2022): Sustainable Corporate Governance in the United Kingdom. Environmental Sustainability in Directors’ Decision-Making. DOI: https://doi.org/10.48718/1k89-pj62 Nr. 24 Schlegler, Maren; Koch, Susanne (2022): Pilot study on the learning success of students in service-learning compared to other teaching and learning formats. DOI https://doi.org/10.48718/n9fy-cv17 Nr. 23 Hagen, Tobias; Hamann, Jonas; Saki, Siavash (2022): Discretization of Urban Areas using POI-based Tesselation. DOI https://doi.org/10.48718/7jjr-1c66 Nr. 22 Balioamoune, Mina; Bausony, Mohamed A.K.; Lutz, Stefan; K.A. Mohamed, Ehab (2022): International Ownership and SMEs in Middle Eastern and African Economies. DOI: https://doi.org/10.48718/c5sy-3h14 Nr. 21 Graf, Erika; Franz, Catharina; Rugbarth, Matthias; Schmidt, Leonard (2021): Wie beeinflussen die Namen von Fleischersatzprodukten die Akzeptanz der Verbraucher? DOI: https://doi.org/10.48718/smsh-p640 Nr. 20 Hagen, Tobias; Saki, Siavash; Scheel-Kopeinig, Sabine (2021): start2park – Determining, Explaining and Predicting Cruising for Parking. Nr. 19 Graf, Erika (2021): Banken auf dem Holzweg? Eine empirische Untersuchung der Bewertung von Kreditkarten aus Holz. Nr. 18 Ziegler, Yvonne; Uli, Vincenzo Uli; Kramer, Astrid; Tatari, Mahmoud Tatari (2021): Development of an innovative halal logistics concept for the air cargo supply chain. Nr. 17 Hagen, Tobias; Scheel-Kopeinig, Sabine (2020): Would Customers be willing to use an alternative (chargeable) delivery concept for the last mile? Nr. 16 Giegler, Nicolas; Schneider, Swen (2020): Leadership und Digitalisierung. Nr. 15 Graml, Regine; Hagen, Tobias; Ziegler, Yvonne; Khachatryan, Kristine; Astrida Herman, Ricky (2020): Lesbische Frauen in der Arbeitswelt – The L-Word in Business. Nr. 14 Voigt, Martina; Ruppert, Andrea (2018): Follow-up –Studie: Gendertypische Verhandlungskompetenz und ihre Auswirkungen auf Gehalts-und Aufstiegsverhandlungen. 16 Nr. 13 Celebi, Kaan; Hönig, Michaela (2018): Dynamic Macroeconomic Effects on the German Stock Market before and after the Financial Crisis. Nr. 12 Lutz, Stefan (2018): R&D, IP, and firm profits in the North American automotive supplier industry. Nr. 11 Voigt, Martina; Ruppert, Andrea (2016): Durchsetzungsorientiert, hart oder sachorientiert, kooperativ verhandeln –welche Einschätzungen und Erwartungen haben Studierende? Nr. 10 Hagen, Tobias (2016): Econometric Evaluation of a Placement Coaching Program for Recipients of Disability Insurance Benefits in Switzerland. Nr. 9 Weissenrieder, Caprice Oona; Spura, Anastassja (2015): Akzeptanz von Führungskräften‐ Analyse wahrgenommener Verhaltensweisen von Frauen und Männern in Führungspositionen. Nr. 8 Graf, Erika (2015): Raising Sustainability Awareness and Understanding in Higher Education. Nr. 7 Weissenrieder, Caprice Oona; Graml, Regine; Hagen, Tobias; Ziegler, Yvonne (2015): Explorative Untersuchung der Unternehmenskultur auf die Karrierechancen von Frauen. Nr. 6 Ruppert, Andrea; Voigt, Martina (2014): Verhandlungsstrategien und Verhandlungstaktiken in Gehaltsverhandlungen. Nr. 5 Jungmittag, Andre (2014): Combination of Forecasts across Estimation Windows: An Application to Air Travel Demand. Nr. 4 Jungmittag, Andre (2014): Der Trade-off zwischen deutschen Direktinvestitionen und Exporten: Wie wichtig sind die verschiedenen Dimensionen der Distanz? Nr. 3 Hagen, Tobias; Waldeck, Stefanie (2014): Using Panel Econometric Methods to Estimate the Effect of Milk Consumption on the Mortality Rate of Prostate and Ovarian Cancer. Nr. 2 Hagen, Tobias (2014): Impact of National Financial Regulation on Macroeco-nomic and Fiscal Performance After the 2007 Financial Shock Econometric Analyses Based on Cross-Country Data. Nr. 1 Rieck, Christian; Bendig, Helena; Hünnemeyer, Julius; Nitzsche, Lisa (2012): Diversität im Aufsichtsrat. Studie über die Zusammensetzung deutscher Aufsichtsräte. Alle Publikationen des Fachbereichs 3 Wirtschaft und Recht finden Sie unter: https://www.frankfurt-university.de/de/hochschule/fachbereich-3-wirtschaft-und-recht/forschungund-transfer/publikationen/