How AI-Based Systems Can Induce Reflections: The Case of AI-Augmented Diagnostic Work

2023 | Management Information Systems Quarterly | Citations: 0

Authors: Abdel-Karim, Benjamin; Pfeuffer, Nicolas; Carl, K. Valerie; Hinz, Oliver

Abstract: This paper addresses a thus-far neglected dimension in human-artificial intelli ... Expand

Abstract: This paper addresses a thus-far neglected dimension in human-artificial intelligence (AI) augmentation: machine-induced reflections. By establishing a grounded theoretical-informed model of machine-induced reflection, we contribute to the ongoing discussion in information systems (IS) regarding AI and research on reflection theories. In our multistage study, physicians used a machine learning-based (ML) clinical decision support system (CDSS) to see if and how this interaction can stimulate reflective practice in the context of an X-ray diagnosis task. By analyzing verbal protocols, performance metrics, and survey data, we developed an integrative theoretical foundation to explain how ML-based systems can help stimulate reflective practice. Individuals engage in more critical or shallower modes depending on whether they perceive a conflict or agreement with these CDSS systems, which in turn leads to different levels of reflection depth. By uncovering the process of machine-induced reflections, we offer IS research a different perspective on how such AI-based systems can help individuals become more reflective, and consequently more effective, professionals. This perspective stands in stark contrast to the traditional, efficiency-focused view of MLbased decision support systems and also enriches theories on human-AI augmentation. Collapse

Topics: artificial intelligence decision making decision support system algorithmic transparency recommender system

Methods: cognitive walkthrough survey theory development qualitative interview Student's t-test

Theories: learning theory information systems theory theory of abduction

Algorithmic Transparency: Concepts, Antecedents, and Consequences – A Review and Research Framework

2023 | Communications of the Association for Information Systems | Citations: 0

Authors: Bitzer, Tobias; Wiener, Martin; Cram, W. Alec

Abstract: The widespread and growing use of algorithm-enabled technologies across many as ... Expand

Abstract: The widespread and growing use of algorithm-enabled technologies across many aspects of public and private life is increasingly sparking concerns about the lack of transparency regarding the inner workings of algorithms. This has led to calls for (more) algorithmic transparency (AT), which refers to the disclosure of information about algorithms to enable understanding, critical review, and adjustment. To set the stage for future research on AT, our study draws on previous work to provide a more nuanced conceptualization of AT, including the explicit distinction between AT as action and AT as perception. On this conceptual basis, we set forth to conduct a comprehensive and systematic review of the literature on AT antecedents and consequences. Subsequently, we develop an integrative framework to organize the existing literature and guide future work. Our framework consists of seven central relationships: (1) AT as action versus AT as perception; factors (2) triggering and (3) shaping AT as action; (4) factors shaping AT as perception; as well as AT as perception leading to (5) rational-cognitive and (6) affective-emotional responses, and to (7) (un-)intended behavioral effects. Building on the review insights, we identify and discuss notable research gaps and inconsistencies, along with resulting opportunities for future research. review of the literature on AT. Not only has such an endeavor, to our knowledge, not yet been undertaken; but more importantly, we expect that such a review will reveal critical research gaps and inconsistencies and inform future research in the area. In particular, we aim to extend past work and contribute to the extant body of knowledge by (1) strengthening the conceptual foundations of AT; (2) reviewing the AT literature in a systematic manner; and (3) deriving a research framework. In doing so, our study offers a more nuanced conceptualization of AT, including the explicit distinction between AT as action and AT as perception, and delineates AT from related concepts. Moreover, following established guidelines (e.g., vom Brocke et al., 2015;Okoli & Schabram, 2010;Paré et al., 2015;Webster & Watson, 2002) and drawing on a review sample of 50 studies (published between 2002 and 2021), our study offers a comprehensive review and synthesis of prior research on the antecedents and consequences of AT. Finally, we propose an integrative framework to organize existing and guide future work. Collapse

Topics: database system algorithmic transparency artificial intelligence website digital service

Methods: computational algorithm literature study synthesis mixed method archival research

Fit into work! From formalizing governance of gig platform ecosystems to helping gig workers craft their platform work

2023 | Decision Support Systems | Citations: 0

Authors: Chen, Jianyu; Luo, Jinlian; Hu, Wenan; Ma, Jun

Abstract: The current governance model of gig platform ecosystems benefits gig platforms a ... Expand

Abstract: The current governance model of gig platform ecosystems benefits gig platforms and customers, but puts gig workers in the weakest position of the triangular relationship. Thus, gig workers experience difficulty in addressing hardships related to gig work. To this end, we must determine how gig platforms build a governance environment that matches gig workers to help them craft gig work. However, extant studies have remained in the qualitative research surface, lacked a certain empirical insight, and have not clarified the direction to creating an effective platform governance model. Thus, based on such literature, we first conceptualize platform formal­ ization governance. Drawing on person-environment fit theory, the PLS-SEM analysis of 290 data from gig workers finds that platform formalization governance directly promotes gig workers' job crafting and indirectly propels it by improving performance appraisal systems' transparency. However, platform work insecurity climate weakens such an direct/indirect positive relationship. These findings contribute to strategic values and effec­ tiveness of the governance model of gig platform ecosystems, thus balancing interests between gig platforms, gig workers, and customers. Gig platforms should consider appropriately formalizing the current governance model and improving work insecurity climate, which helps gig workers more easily address hardships associated with platform work through job crafting. Collapse

Topics: platform worker platform work digital ecosystem platform governance algorithmic management

Methods: survey partial least squares path modeling data transformation PLS tool double-barreled question

Robo-advisors and investment intention: A perspective of value-based adoption

2023 | Information & Management | Citations: 0

Authors: Hong, Xianpei; Pan, Liwei; Gong, Yeming; Chen, Qian

Abstract: Financial robo-advisors have been widely used to assist individuals in their inv ... Expand

Abstract: Financial robo-advisors have been widely used to assist individuals in their investment decisions, making it important to reduce uncertainties in the assistance process. Existing empirical studies rarely explore uncertainty reduction strategies and their implications on users’ investment intentions in the context of financial roboadvisors; our study attempts to address this gap. We construct a model to explain how uncertainty reduction strategies affect users’ investment intention in using financial robo-advisors. By collecting and analyzing a sample of 307 financial robo-advisor users, we find that algorithmic interpretability, structural assurance, and interactivity as uncertainty reduction strategies are positively related to users’ investment intention through the value-based adoption mechanism. Our research extends the value-based adoption model and uncertainty reduction theory in the financial robo-advisor context. We provide insights to financial robo-advisor service providers about focusing on improving algorithmic transparency, third-party assurance, and interactivity of financial robo-advisors to enhance perceived value and investment intention. Collapse

Topics: robo advisor privacy artificial intelligence usage intention organizational productivity

Methods: survey computational algorithm factor analysis statistical hypothesis test structural equation modeling

Theories: uncertainty reduction theory

Algorithmic Accountability

2023 | Business & Information Systems Engineering | Citations: 0

Authors: Horneber, David; Laumer, Sven

Abstract: ... Expand

Abstract: Collapse

Topics: algorithmic transparency

Algorithm Sensemaking: How Platform Workers Make Sense of Algorithmic Management

2023 | Journal of the Association for Information Systems | Citations: 3

Authors: Möhlmannn, Mareike; Salge, Carolina Alves de Lima; Marabelli, Marco

Abstract: Algorithmic management can create work environment tensions that are detrimental ... Expand

Abstract: Algorithmic management can create work environment tensions that are detrimental to workplace well-being and productivity. One specific type of tension originates from the fact that algorithms often exhibit limited transparency and are perceived as highly opaque, which impedes workers’ understanding of their inner workings. While algorithmic transparency can facilitate sensemaking, the algorithm’s opaqueness may aggravate sensemaking. By conducting an empirical case study in the context of the Uber platform, we explore how platform workers make sense of the algorithms managing them. Drawing on Weick’s enactment theory, we theorize a new form of sensemaking—algorithm sensemaking—and unpack its three sub-elements: (1) focused enactment, (2) selection modes, and (3) retention sources. The sophisticated, multistep process of algorithm sensemaking allows platform workers to keep up with algorithmic instructions systematically. We add to previous literature by theorizing algorithm sensemaking as a mediator linking workers’ perceptions about tensions in their work environment and their behavioral responses. Collapse

Topics: sensemaking Uber platform worker algorithmic transparency algorithmic management

Methods: computational algorithm theory development qualitative interview case study theoretical contribution

Theories: enactment theory

ACCOUNTABILITY INCONGRUENCE AND ITS EFFECTS ON AI DEVELOPERS’ JOB SATISFACTION

2023 | European Conference On Information Systems | Citations: 0

Authors: Schmidt, Jan-Hendrik; Bartsch, Sebastian; Adam, Martin; Benlian, Alexander

Abstract: Developers of Artificial Intelligence (AI)-based systems are increasingly urged ... Expand

Abstract: Developers of Artificial Intelligence (AI)-based systems are increasingly urged to assume accountability for their development decisions, referring to the degree to which they must justify underlying algorithms and their outcomes on demand. Thereby, AI developers often have to juxtapose how much accountability they self-attribute to them and how much accountability they perceive others attribute to them, creating intrapersonal perceptual accountability (in)congruence with unknown consequences for their job satisfaction. Building on perceptual congruence research and algorithmic accountability literature, we conducted an online survey of 87 AI developers about their experiences in AI-based systems development projects. Our results show that the lower the incongruence between self-attributed and others-attributed accountability, the higher the job satisfaction of AI developers. Moreover, we find that AI developers' role ambiguity mediates this effect. Our study contributes to a more nuanced understanding of AI developers' perceived accountability, with essential insights for defining job roles and understanding AI developers. Collapse

Topics: artificial intelligence job satisfaction IT development project algorithmic transparency IT manager

Methods: survey linear regression analysis nonparametric test Wilcoxon signed-rank test statistical power analysis

Mitigating Bias in Organizational Development and Use of Artificial Intelligence

2023 | International Conference on Information Systems | Citations: 0

Authors: Tanriverdi, Hüseyin; Akinyemi, John-Patrick; Neumann, Terrence

Abstract: We theorize why some artificial intelligence (AI) algorithms unexpectedly treat ... Expand

Abstract: We theorize why some artificial intelligence (AI) algorithms unexpectedly treat protected classes unfairly. We hypothesize that mechanisms by which AI assumes agencies, rights, and responsibilities of its stakeholders can affect AI bias by increasing complexity and irreducible uncertainties: e.g., AI's learning method, anthropomorphism level, stakeholder utility optimization approach, and acquisition mode (make, buy, collaborate). In a sample of 726 agentic AI, we find that unsupervised and hybrid learning methods increase the likelihood of AI bias, whereas "strict" supervised learning reduces it. Highly anthropomorphic AI increases the likelihood of AI bias. Using AI to optimize one stakeholder's utility increases AI bias risk, whereas jointly optimizing the utilities of multiple stakeholders reduces it. User organizations that co-create AI with developer organizations instead of developing it in-house or acquiring it off-the-shelf reduce AI bias risk. The proposed theory and the findings advance our understanding of responsible development and use of agentic AI. Collapse

Topics: artificial intelligence algorithmic bias Google+ database system algorithmic transparency

Methods: computational algorithm supervised learning survey design optimization model theory development

The Effect of Online Review Portal Design: The Moderating Role of Explanations for Review Filtering

2023 | International Conference on Information Systems | Citations: 0

Authors: WANG, Guangxu; Xu, David (Jingjun); LIAO, Stephen Shaoyi

Abstract: The flood of non-constructive and fake online consumer reviews erects a conside ... Expand

Abstract: The flood of non-constructive and fake online consumer reviews erects a considerable barrier to consumers making efficient decisions. Various review filtering algorithms have been developed to address this challenge, but the design of post-development review portals continues to lack a consensus. In review portals, disclosing more transparent reviews is efficient for enhancing users' trust. However, it will cause users' diminished focus on recommended reviews, leading to sub-optimal decisions. A research model is then developed to investigate users' cognitive processes in their responses to three review exhibition designs (i.e., informed silent display design, filtered review display design, and composite display design) regarding trust in the review portal and perceived decision quality. We also suggest that explanations for review filtering play a moderating role in users' perceptions, which appears to be a viable resolution to this dilemma. This paper provides significant theoretical and practical insights for the review portal design and implementation. Collapse

Topics: algorithmic transparency decision quality electronic commerce

Methods: experiment experimental task survey statistical power analysis experimental group

Algorithm-Augmented Work and Domain Experience: The Countervailing Forces of Ability and Aversion

2022 | Organization Science | Citations: 0

Authors: Allen, Ryan; Choudhury, Prithwiraj (Raj)

Abstract: Past research offers mixed perspectives on whether domain experience helps or hu ... Expand

Abstract: Past research offers mixed perspectives on whether domain experience helps or hurts algorithm-augmented worker performance. Reconciling these perspectives, we theorize that intermediate levels of domain experience are optimal for algorithm-augmented performance, due to the interplay between two countervailing forces—ability and aversion. Although domain experience can increase performance via increased ability to complement algorithmic advice (e.g., identifying inaccurate predictions), it can also decrease performance via increased aversion to accurate algorithmic advice. Because ability developed through learning by doing increases at a decreasing rate, and algorithmic aversion is more prevalent among experts, we theorize that algorithm-augmented performance will first rise with increasing domain experience, then fall. We test this by exploiting a within-subjects experiment in which corporate information technology support workers were assigned to resolve problems both manually and using an algorithmic tool. We confirm that the difference between performance with the algorithmic tool versus without the tool was characterized by an inverted U-shape over the range of domain experience. Only workers with moderate domain experience did significantly better using the algorithm than resolving tickets manually. These findings highlight that, even if greater domain experience increases workers’ ability to complement algorithms, domain experience can also trigger other mechanisms that overcome the positive ability effect and inhibit performance. Additional analyses and participant interviews suggest that, even though the highest experience workers had the greatest ability to complement the algorithmic tool, they rejected its advice because they felt greater accountability for possible unintended consequences of accepting algorithmic advice. Collapse

Topics: human resource management IT workforce unintended consequence technological change algorithmic transparency

Methods: computational algorithm experiment qualitative interview theory development within subject experiment