Denied by an (Unexplainable) Algorithm: Teleological Explanations for Algorithmic Decisions Enhance Customer Satisfaction

Tomaino, Geoffrey; Abdulhalim, Hisham; Kireyev, Pavel; Wertenbroch, Klaus

doi:10.2139/ssrn.3683754

Cited by 11 publications

(8 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consumers tend to perceive AI technologies as opaque (Burrell, 2016), such that they avoid disclosing personal data (Puntoni et al, 2021; Thomaz et al, 2020), but a detailed explanation that reveals how AI algorithms work can decrease such concerns. Our conceptualization also builds on recent research that proposes mechanistic explanations of products (Fernbach et al, 2013) and AI algorithms (Cadario et al, 2021; Tomaino et al, 2020) as antecedents of positive attitudes and behavioral intentions. First, we propose a mechanistic explanation as a crucial inhibitor of data collection concerns, and second, we clarify the (high) level of explanatory detail needed to reduce data collection concerns.…”

Section: Discussionmentioning

confidence: 99%

“…Google already gives customers insights into why its algorithm has generated a particular outcome (Kelion, 2019), and the European Union requires companies to explain AI systems' decision‐making process as part of its General Data Protection Regulation (Skiera et al, 2022). Explaining how the AI algorithm works can increase consumers' trust in AI‐based recommendation agents (Wang & Benbasat, 2007), satisfaction with the algorithm's decisions (Tomaino et al, 2020), and the likelihood of using AI‐based services (Cadario et al, 2021). However, prior studies do not explicitly identify a mechanistic explanation of AI algorithms as a potential inhibitor of data collection concerns, nor do they specify how much detail is required in an explanation for it to be effective.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Explaining how algorithms work reduces consumers' concerns regarding the collection of personal data and promotes AI technology adoption

Querci

Barbarossa

Romani

et al. 2022

Psychology and Marketing

View full text Add to dashboard Cite

Consumers' concerns about how companies gather and use their personal data can impede the widespread adoption of artificial intelligence (AI) technologies. This study demonstrates that mechanistic explanations of AI algorithms can inhibit such data collection concerns. Four independent online experiments show a negative effect of detailed mechanistic explanations on data collection concerns (Studies 1a and 1b), as well as mediating influences of a subjective understanding of how AI algorithms work (Study 2) and increased the likelihood to adopt AI technologies after data collection concerns have been mitigated (Study 3). These findings contribute to research on consumer privacy concerns and the adoption of AI technologies, by identifying (1) a new inhibitor of data collection concerns, namely, mechanistic explanations of AI algorithms; (2) the psychological mechanisms underlying mechanist explanation effects; and (3) how diminished data collection concerns promote AI technology adoption. These insights can help companies design more effective communication strategies that reduce the perceived opacity of AI algorithms, reassure consumers, and encourage their adoption of AI technologies.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Explaining how algorithms work reduces consumers' concerns regarding the collection of personal data and promotes AI technology adoption

Querci

Barbarossa

Romani

et al. 2022

Psychology and Marketing

View full text Add to dashboard Cite

show abstract

“…Given the well known false consensus effect [69], respondents' answers should be influenced by their own preferences. 5 Nevertheless, our estimate of the link between CRT and preferences may be viewed as conservative, as it really captures the link between the respondent's CRT score and their beliefs on what explanation another consumer would prefer given the vignette.…”

Section: Methodsmentioning

confidence: 99%

Understanding Consumer Preferences for Explanations Generated by XAI Algorithms

Ramon,

Vermeire,

Toubia

et al. 2021

Preprint

View full text Add to dashboard Cite

Explaining firm decisions made by algorithms in customer-facing applications is increasingly required by regulators and expected by customers. While the emerging field of Explainable Artificial Intelligence (XAI) has mainly focused on developing algorithms that generate such explanations, there has not yet been sufficient consideration of customers' preferences for various types and formats of explanations. We discuss theoretically and study empirically people's preferences for explanations of algorithmic decisions. We focus on three main attributes that describe automatically-generated explanations from existing XAI algorithms (format, complexity, and specificity), and capture differences across contexts (online targeted advertising vs. loan applications) as well as heterogeneity in users' cognitive styles. Despite their popularity among academics, we find that counterfactual explanations are not popular among users, unless they follow a negative outcome (e.g., loan application was denied). We also find that users are willing to tolerate some complexity in explanations. Finally, our results suggest that preferences for specific (vs. more abstract) explanations are related to the level at which the decision is construed by the user, and to the deliberateness of the user's cognitive style. * Yanou Ramon and Tom Vermeire contributed equally.

show abstract

“…To address this concern, policy makers may look to behavioral sciences for appropriate interventions (Steg, Van Den Berg, and De Groot 2013; Tomaino, Teow, et al 2020). Our research suggests that such an application may not be straightforward.…”

Section: Wai Yan Leong Land Transport Authority Singaporementioning

confidence: 99%

“…The problem may be exacerbated by unexplainable algorithms, which entail statistical mechanisms so complex that they are beyond human comprehension (De Bruyn et al 2020; Gunning and Aha 2019). Faced with regulatory requirements to explain their algorithmic decision making (e.g., Articles 13–15 of the European Union's General Data Protection Regulation [GDPR]), companies that do not want to or cannot explain the mechanisms underlying their algorithmic decisions can placate consumers by explaining their objectives (Tomaino, Abdulhalim, et al 2020).…”

Section: Klaus Wertenbroch Inseadmentioning

confidence: 99%

JPP&M's Global Perspective and Impact: An Agenda for Research on Marketing and Public Policy

Shultz¹,

Hoek²,

Lee³

et al. 2021

Journal of Public Policy & Marketing

Self Cite

View full text Add to dashboard Cite

For several decades, the Journal of Public Policy & Marketing has stimulated and led debates with far-reaching implications for consumer well-being, global relationships and, ultimately, human survival. The challenges we face have not disappeared but intensified. Today, we must respond to climate change, manage a global pandemic, and address disparities and inequities that threaten our planet in ways we are yet to comprehend fully. However, the JPP&M community remains well-placed to inform responses to these crises. This article draws together perspectives on new and long-standing questions and challenges, as we highlight the increasing urgency of addressing inequities and embedding sustainability at the heart of policy-making. Yet, while progress addressing these complex questions often has been slow, we also identify compelling opportunities. Sound policies and good marketing and consumption practices in response to health crises, environmental degradation, injustice, automation, violence and war; the transformational benefits following Constructive Engagement offer hope that, even when faced with unprecedented challenges, human resilience and ingenuity can create meaningful responses. As we address chronic and novel problems, we are confident JPP&M’s community of researchers, policy makers, and advocates will continue to bring innovation, insight, and rigor, and play a leading role in discovering solutions, locally and globally.

show abstract

Denied by an (Unexplainable) Algorithm: Teleological Explanations for Algorithmic Decisions Enhance Customer Satisfaction

Cited by 11 publications

References 71 publications

Explaining how algorithms work reduces consumers' concerns regarding the collection of personal data and promotes AI technology adoption

Explaining how algorithms work reduces consumers' concerns regarding the collection of personal data and promotes AI technology adoption

Understanding Consumer Preferences for Explanations Generated by XAI Algorithms

JPP&M's Global Perspective and Impact: An Agenda for Research on Marketing and Public Policy

Contact Info

Product

Resources

About