Explanations as Mechanisms for Supporting Algorithmic Transparency

By Emilee J. Rader, Kelley Cotter, Janghee Cho. 2018

393 citations recorded by the source corpus.

Transparency can empower users to make informed choices about how they use an algorithmic decision-making system and judge its potential consequences. However, transparency is often conceptualized by the outcomes it is intended to bring about, not the specifics of mechanisms to achieve those outcomes. We conducted an online experiment focusing on how different ways of explaining Facebook's News Feed algorithm might affect participants' beliefs and judgments about the News Feed. We found that all explanations caused participants to become more aware of how the system works, and helped them to determine whether the system is biased and if they can control what they see. The explanations were less effective for helping participants evaluate the correctness of the system's output, and form opinions about how sensible and consistent its behavior is. We present implications for the design of transparency mechanisms in algorithmic decision-making systems based on these results.

Subfields: Computer Science Applications, Safety Research, Sociology and Political Science

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Science Applications

In Safety Research

In Sociology and Political Science

Citations in the corpus

Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward (2020, background): “, 2019), autonomous vehicles and traffic management (Ye, 2018), social-media news-feed (Rader et al., 2018), work scheduling and optimisation (O’Neil, 2016), and several others.”
Manipulating and Measuring Model Interpretability (2018, methodology): “We therefore take a human-centered approach, building on previous psychology and social science research on human trust in models [4, 17, 21] and similar efforts in the machine learning and human-computer interaction communities [2, 6, 8, 13, 15, 23, 24].”
Automated rationale generation: a technique for explainable AI and its effects on human perceptions (2019, methodology): “[36] use explanations manually generated from content analysis of Facebook’s News Feed to study perceptions of algorithmic transparency.”
Transparency you can trust: Transparency requirements for artificial intelligence between legal norms and contextual concerns (2019, background): “Rader et al. (2018), for example, studied the Facebook newsfeed algorithm to examine the eﬀects of what -explanations, how -explanations, why -explanations, and objective - explanations on diﬀerent outcomes.”
Epistemic values in feature importance methods: Lessons from feminist epistemology (2021, background): “In the case of ML explanations, the intertwining of epistemic and social goals is particularly striking, since the demand for an explanation is often not just a demand for knowledge of how the algorithm works, but also part of an ethical demand for greater accountability or transparency [22, 60].”
How Platform-User Power Relations Shape Algorithmic Accountability: A Case Study of Instant Loan Platforms and Financially Stressed Users in India (2022, background): “In fact, some studies with audits and explainability mechanisms have documented positive outcomes such as raising users’ critical awareness [100], increasing their desires to seek accountability from the designers of ADS [113], and influencing technology providers to make changes in ADS [46, 101].”
Expanding Explainability: Towards Social Transparency in AI systems (2021, background): “Moreover, while XAI is often claimed to be a critical step towards accountable AI, empirical studies have found little evidence that explanations improve a user’s perceived accountability or control over AI systems [81, 90].”
Algorithmic bias: review, synthesis, and future research directions (2021, background): “(2019); (Rader et al., 2018) Algorithmic transparency ×”
Towards a Science of Human-AI Decision Making: A Survey of Empirical Studies (2021, methodology): “Other metrics measures the correctness of people’s assessment of model performance [36, 55, 107, 125], detection of errors [143], identification of important features [28, 134], or whether they can provide a correct description of model behaviors [29, 77, 112].”
Guidelines for Human-AI Interaction (2019, background): “For example, a large body of work exists and continues to grow around how to increase transparency or explain the behaviors of AI systems (e.g., [14, 21, 23, 36, 38, 44], to name a few).”
Towards Transparency by Design for Artificial Intelligence (2020, background): “The overview suggests that the trust-generating capacity of transparency in AI systems depends on different factors, such as user expectations, how transparency is implemented (Rader et al. 2018) and which technologies are investigated (Kulesza et al. 2013).”
Playing the visibility game: How digital influencers and algorithms negotiate influence on Instagram (2018, background): “In general, algorithms function behind the scenes with many users unaware of their presence (Eslami et al., 2015; Rader et al., 2018).”
The fallacy of inscrutability (2018, background): “Another approach to defeating inscrutability which many scholars have proposed is the use of explainable systems, which provide an explanation of their outputs [4,5,30,49,50].”
A Multidisciplinary Survey and Framework for Design and Evaluation of Explainable AI Systems (2018, background): “Examples of AI application examples in which privacy awareness is primarily important include personalized advertisements using users’ online advertisement [44] and personalized news feed in social media [58, 170].”