Working with AI: How Attitudes Shape Human-AI Collaboration

2023 | International Conference on Information Systems | Citations: 0

Authors: Glienke, Maximilian; Hartwich, Nicole; Antons, David

Abstract: Recent research shows that people tend to avoid relying on artificial intellige ... Expand

Abstract: Recent research shows that people tend to avoid relying on artificial intelligence (AI) when making decisions at work. Although previous studies have investigated the causes and effects of this phenomenon, the role of attitudes, which are considered as a crucial factor in biased decision-making, remains unexplored. To investigate whether and how attitudes toward AI influence decision-making in work-related situations, we developed and conducted a lab experiment in which 156 participants received AI advice on process improvement while performing a manufacturing task. Our results suggest that negative attitudes toward AI are associated with greater reluctance to accept AI advice, resulting in lower performance in work-related situations. Furthermore, we show that human involvement in the process of providing AI advice cannot mitigate the negative effects of strong negative attitudes. Collapse

Topics: artificial intelligence decision making business process improvement human AI collaboration task productivity

Methods: experiment experimental group computational algorithm laboratory experiment regression analysis method

Theories: attitude theory

On Intelligence Augmentation and Visual Analytics to Enhance Clinical Decision Support Systems

2022 | HICSS | Citations: 0

Authors: Heart, Tsipi; Padman, Rema; Ben-Assuli, Ofir; Gefen, David; Klempfner, Robert

Abstract: Human-in-the-loop intelligence augmentation (IA) methods combined with visual an ... Expand

Abstract: Human-in-the-loop intelligence augmentation (IA) methods combined with visual analytics (VA) have the potential to provide additional functional capability and cognitively driven interpretability to Decision Support Systems (DSS) for health risk assessment and patient-clinician shared decision making. This paper presents some key ideas underlying the synthesis of IA with VA (IA/VA) and the challenges in the design, implementation, and use of IA/VA-enabled clinical decision support systems (CDSS) in the practice of medicine through data driven analytical models. An illustrative IA/VA solution provides a visualization of the distribution of health risk, and the impact of various parameters on the assessment, at the population and individual levels. It also allows the clinician to ask “what-if” questions using interactive visualizations that change actionable risk factors of the patient and visually assess their impact. This approach holds promise in enhancing decision support systems design, deployment and use outside the medical sphere as well. Collapse

Topics: decision support system missing data artificial intelligence decision making user interface

Methods: design artifact descriptive statistic data imputation qualitative interview design principle

Phishing, Data-Disclosure and The Cognitive Reflection Test

2022 | HICSS | Citations: 0

Authors: Tjostheim, Ingvar

Abstract: Phishing is a form of online identity theft that aims to steal sensitive informa ... Expand

Abstract: Phishing is a form of online identity theft that aims to steal sensitive information such as online banking passwords and credit card information from users. Data is key for the digital economy, but disclosing personal data online increases vulnerabilities and the likelihood of experiencing negative consequences from disclosure. In this paper, we analyze willingness to share personal data, a preference for an intuitive decision style and susceptibility to phishes. We report the results of three large-scale national studies in Norway that included the cognitive reflection test (CRT) and a choice experiment on willingness to share personal data. With a binary logistic regression method, we analyzed the relationship between the CRT, willingness to share data and demographical variables with susceptibility to phishes as the outcome variable. Our main finding is that the willingness to share personal data and an intuitive thinking style significantly predict the probability of falling for phish. These results are based on three largescale studies with national populations, in contrast to earlier studies that in most cases relying on student populations, giving them greater validity. Collapse

Topics: phishing logistics management electronic mail decision making Facebook

Methods: archival research survey experiment digital trace data logistic regression

The Case of Human-Machine Trading as Bilateral Organizational Learning

2021 | International Conference on Information Systems | Citations: 0

Authors: Sturm, Timo; Koppe, Timo; Scholz, Yven; Buxmann, Peter

Abstract: In today's organizations, both humans and machine learning (ML) systems jointly ... Expand

Abstract: In today's organizations, both humans and machine learning (ML) systems jointly form routines. Yet, we do not know much about the underlying reciprocal interplay between them, which complicates their effective coordination. Taking an organizational learning perspective, we study the dynamics of human learning and ML to understand how organizations can benefit from their respective idiosyncrasies when enabling bilateral learning. Drawing on a case of human traders and a reinforcement ML system trading productively at Allianz Global Investors, we apply human-machine pattern recognition on digital trace data to explore their (interconnected) dynamics. We find that bilateral learning can increase trading performance, which appears to result from an emerging virtuous cycle between humans and the ML system. Our explorative case study offers insights into how organizational learning depends on the coordination of both human learning and ML, which can help manage the collaboration between human and artificial intelligence within organizational routines. Collapse

Topics: organization learning artificial intelligence human AI collaboration human computer interaction organizational context

Methods: digital trace data computational algorithm case study descriptive statistic theory development

Theories: organizational learning theory

The effect of interactive analytical dashboard features on situation awareness and task performance

2020 | Decision Support Systems | Citations: 6

Authors: Nadj, Mario; Maedche, Alexander; Schieder, Christian

Abstract: In recent years, new types of interactive analytical dashboard features have eme ... Expand

Abstract: In recent years, new types of interactive analytical dashboard features have emerged for operational decision support systems (DSS). Analytical components of such features solve optimization problems hidden from the human eye, whereas interactive components involve the individual in the optimization process via graphical user interfaces (GUIs). Despite their expected value for organizations, little is known about the effectiveness of interactive analytical dashboards in operational DSS or their influences on human cognitive abilities. This paper contributes to the closing of this gap by exploring and empirically testing the effects of interactive analytical dashboard features on situation awareness (SA) and task performance in operational DSS. Using the theoretical lens of SA, we develop hypotheses about the effects of a what-if analysis as an interactive analytical dashboard feature on operational decision-makers' SA and task performance. The resulting research model is studied with a laboratory experiment, including eye-tracking data of 83 participants. Our findings show that although a what-if analysis leads to higher task performance, it may also reduce SA, nourishing a potential out-of-the-loop problem. Thus, designers and users of interactive analytical dashboards have to carefully mitigate these effects in the implementation and application of operational DSS. In this article, we translate our findings into implications for designing dashboards within operational DSS to help practitioners in their efforts to address the danger of the out-of-the-loop syndrome. Collapse

Topics: digital dashboard task productivity decision support system situation awareness decision making

Methods: oculometry descriptive statistic correlation analysis laboratory experiment Wilcoxon signed-rank test

Predicting user behavior in electronic markets based on personality-mining in large online social networks

2017 | Electronic Markets | Citations: 0

Authors: Buettner, Ricardo

Abstract: Determining a user’s preferences is an important condition for effectively opera ... Expand

Abstract: Determining a user’s preferences is an important condition for effectively operating automatic recommendation systems. Since personality theory claims that a user’s personality substantially influences preference, I propose a personality-based product recommender (PBPR) framework to analyze social media data in order to predict a user’s personality and to subsequently derive its personality-based product preferences. The PBRS framework will be evaluated as an IT-artefact with a unique online social network XING dataset and a unique coffeemaker preference dataset. My evaluation results show (a) the possibility of predicting a user’s personality from social media data, as I reached a predictive gain between 23.2 and 41.8 percent and (b) the possibility of recommending products based on a user’s personality, as I reached a predictive gain of 45.1 percent. Collapse

Topics: personality social network social media recommender system electronic market

Methods: experimental group computational algorithm design artifact machine learning experiment

Theories: big five model personality theory

The Impact of Computerized Agents on Immediate Emotions, Overall Arousal and Bidding Behavior in Electronic Auctions

2015 | Journal of the Association for Information Systems | Citations: 47

Authors: Teubner, Timm; Adam, Marc; Riordan, Ryan

Abstract: The presence of computerized agents has become pervasive in everyday live. In th ... Expand

Abstract: The presence of computerized agents has become pervasive in everyday live. In this paper, we examine the impact of agency on human bidders' affective processes and bidding behavior in an electronic auction environment. In particular, we use skin conductance response and heart rate measurements as proxies for the immediate emotions and overall arousal of human bidders in a lab experiment with human and computerized counterparts. Our results show that computerized agents mitigated 1) the intensity of bidders' immediate emotions in response to discrete auction events, such as submitting a bid and winning or losing an auction, and 2) the bidders' overall arousal levels during the auction. Moreover, agency affected bidding behavior and its relation to overall arousal: whereas overall arousal and bids were negatively correlated when competing against human bidders, we did not observe this relationship for computerized agents. In other words, lower levels of agency yield less emotional behavior. The results of our study have implications for the design of electronic auction platforms and markets that include both human and computerized actors. Collapse

Topics: online auction decision making website electronic trading anonymity

Methods: experiment electrodermal activity electrocardiography laboratory experiment Wald test

Theories: social comparison theory auction theory

The benefits and dangers of enjoyment with social networking websites

2012 | European Journal of Information Systems | Citations: 265

Authors: Turel, Ofir; Serenko, Alexander

Abstract: Information Systems enjoyment has been identified as a desirable phenomenon, bec ... Expand

Abstract: Information Systems enjoyment has been identified as a desirable phenomenon, because it can drive various aspects of system use. In this study, we argue that it can also be a key ingredient in the formation of adverse outcomes, such as technology-related addictions, through the positive reinforcement it generates. We rely on several theoretical mechanisms and, consistent with previous studies, suggest that enjoyment can lead to presumably positive outcomes, such as high engagement. Nevertheless, it can also facilitate the development of a strong habit and reinforce it until it becomes a ‘bad habit’, that can help forming a strong pathological and maladaptive psychological dependency on the use of the IT artifact (i.e., technology addiction). We test and validate this dual effect of enjoyment, with a data set of 194 social networking website users analyzed with SEM techniques. The potential duality of MIS constructs and other implications for research and practice are discussed. Collapse

Topics: information system use information technology addiction website social network internet technology

Methods: survey descriptive statistic longitudinal research structural equation modeling data transformation

Cognitive Feedback in GDSS: Improving Control and Convergence

1993 | Management Information Systems Quarterly | Citations: 30

Authors: Sengupta, Kishore; Te'eni, Dov

Abstract: Cognitive feedback in group decision making is in formation that provides decisi ... Expand

Abstract: Cognitive feedback in group decision making is in formation that provides decision makers with a better understanding of their own decision processes and that of the other group members. It appears to be an effective aid in group decision making. Although it has been suggested as a potential feature of group decision support systems (GDSS), little research has examined its use arid impact. This article investigates the effect of computer generated cognitive feedback in computer-supported group decision processes. It views group decision making as a combination of individual and collective activity. The article tests whether cognitive feedback can enhance control over the individual and collective decision making processes and can facilitate the process of convergence among group members. In a laboratory experiment with groups of three decision makers, 15 groups received online cognitive feedback and 15 groups did not. Users receiving cognitive feedback maintained a higher level of control over the decision-making process as their decision strategies converged. This research indicates that (1) developers should include cognitive feedback as an integral part of the GDSS at every level, and (2) they should design the human-computer interaction so there is an intuitive and effective transition across the components of feedback at all levels- Researchers should extend the concepts explored here to other models of conflict that deal with ill-structured decisions, as well as study the impact of cognitive feedback over time. Finally, researchers trying to enhance the capabilities of GOSS should continue examining how to take advantage of the differences between individual, interpersonal, and collective decision making. Collapse

Topics: decision making group decision support system usability human computer interaction

Methods: experiment self reported survey multivariate analysis of variance descriptive statistic multiple regression

Intelligent Interface Design: An Empirical Assessment of Knowledge Presentation in Expert Systems

1990 | Management Information Systems Quarterly | Citations: 29

Authors: Lamberti, Donna; Wallace, William

Abstract: This research evaluates intelligent interface requirements for knowledge presen ... Expand

Abstract: This research evaluates intelligent interface requirements for knowledge presentation in an expert system used for diagnostic problem solving. In a field study, interactions between employee expertise, knowledge presentation format (procedural vs. declarative), question type (requiring abstract vs. concrete knowledge organization), and task uncertainty are examined for employee problem-solving and decision-making performance (speed and accuracy). Also evaluated are confidence in system recommendations and lines-of-reasoning, as weft as user satisfaction with the system interface. The study provides findings that are discussed within the context of intelligent interface requirements for organizational information systems. The results show that highskill users perform significantly faster and more accurately when solving the problems and have self-reported confidence ratings that are higher than those of low-skill users. The expert system, This research is based on the first author's Ph.D. thesis, which was sponsored by an IBM Fellowship in Information Systems Research.however, has a greater impact on improving performance for low-skill users than for high-skill users. A relationship is found between skill level and task uncertainty indicating that different skilllevel users require different presentation formats paralleling their conceptual representations of the problem. The interaction between skill level and knowledge organization is confirmed with results showing that low-skill users perform faster than high.skill users on questions requiring concrete knowledge organization; whereas high-skill users perform better when presented with questions requiring abstract knowledge organization. Collapse

Topics: software developer problem solving task uncertainty expert system user satisfaction

Methods: experimental group parametric test survey ANOVA descriptive statistic