Should Learning Analytics Models Include Sensitive Attributes? Explaining the Why

Deho, Oscar Blessed; Joksimovíc, Srécko; Li, Jiuyong; Zhan, Chen; Liu, Jixue; Liu, Lin

doi:10.1109/tlt.2022.3226474

Cited by 20 publications

(7 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…More work along these lines is needed. Such work should be conducted with adequate attention to student privacy concerns and thoroughly explore the impact of including sensitive attributes in LA models (Deho et al, 2022). Even still, merely specifying demographic data and being thoughtful about how data biased toward a demographic group may inadvertently perpetuate inequality does not resolve the broader question of how to anchor the field to a more cohesive epistemology that can include the values of openness, fairness, and JEDI.…”

Section: Limited Analysis Of Demographic Subgroupsmentioning

confidence: 99%

Aligning the Goals of Learning Analytics with its Research Scholarship

Ferguson,

Khosravi,

Kovanović

et al. 2023

Learning Analytics

View full text Add to dashboard Cite

show abstract

Section: Limited Analysis Of Demographic Subgroupsmentioning

confidence: 99%

Aligning the Goals of Learning Analytics with its Research Scholarship

Ferguson,

Khosravi,

Kovanović

et al. 2023

Learning Analytics

View full text Add to dashboard Cite

show abstract

“…More work along these lines is needed. Such work should be conducted with adequate attention to student privacy concerns and thoroughly exploring the impact of including sensitive attributes in LA models (Deho et al, 2022). Even still, merely specifying demographic data and being thoughtful about how data biased toward a demographic group may inadvertently perpetuate inequality does not resolve the broader question of how to anchor the field to a more cohesive epistemology that can include the values of openness and fairness and JEDI.…”

Section: Introduction and Literature Reviewmentioning

confidence: 99%

From a Nominal to Normative Commitment: Liberating Learning Analytics from Embedded Biases Through a Sociotechnical Systems Approach

Meaney

2024

Preprint

View full text Add to dashboard Cite

The emphasis on "Can we do this thing with these data?" detracts from other important goals of learning analytics (LA) beyond understanding and optimizing student learning. It also undermines LA's ability to develop a coherent epistemology and be consistent with its stated values of openness, fairness, and justice, equity, diversity, and inclusion (JEDI). Acomprehensive theoretical framework accounting for LAs' macro, meso, and micro level domains would enable more epistemologically cohesive, sociotechnical system approaches ot be taken, which could help shift nominal commitment to values to a normative one.

show abstract

“…This could lead to a system that does not adequately address or even misinterprets the needs of dyslexic learners. Research by Deho et al (2024) emphasizes the importance for AI diagnostic tools to be trained on diverse datasets to ensure accurate and unbiased assessments. Addressing the ethical challenges in AI systems extends beyond mitigating algorithmic bias; and encompasses the crucial need for transparency.…”

Section: Introductionmentioning

confidence: 99%

Advancing Cognitive Accessibility: The Role of Artificial Intelligence in Enhancing Inclusivity

2024

PSEN

View full text Add to dashboard Cite

This editorial examines the transformative role of Artificial Intelligence (AI) in enhancing cognitive accessibility for neurodiverse individuals. It explores the evolution from conventional assistive technologies to sophisticated AI-driven solutions, highlighting how these advancements are reshaping inclusivity in education and the workplace. The piece critically analyzes the benefits and challenges of AI in this context, considering ethical implications, user-centered design, and the need for equitable access. It concludes with a call to action for continued innovation and collaboration in developing AI technologies that truly cater to the diverse needs of neurodiverse individuals.

show abstract

Should Learning Analytics Models Include Sensitive Attributes? Explaining the Why

Cited by 20 publications

References 39 publications

Aligning the Goals of Learning Analytics with its Research Scholarship

Aligning the Goals of Learning Analytics with its Research Scholarship

From a Nominal to Normative Commitment: Liberating Learning Analytics from Embedded Biases Through a Sociotechnical Systems Approach

Advancing Cognitive Accessibility: The Role of Artificial Intelligence in Enhancing Inclusivity

Contact Info

Product

Resources

About