Evaluating the Fairness of Predictive Student Models Through Slicing Analysis

Gardner, Josh; Brooks, Christopher; Baker, Ryan S.

doi:10.1145/3303772.3303791

Cited by 111 publications

(97 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Generally, separation requires that an algorithm's performance, not just its predictions, be fair across groups, or stated another way, that correct and incorrect predictions are distributed equally in relation to the groups under consideration. This criterion has generated several attempts at precise formulation, from equal opportunity/equalized odds (Hardt et al, 2016) to slicing analysis (Gardner et al, 2019) to predictive parity (Chouldechova, 2017).…”

Section: Formal Fairness and Its Application (In A Messy World)mentioning

confidence: 99%

Algorithmic Bias in Education

Baker¹,

Hawn²

2021

Preprint

View full text Add to dashboard Cite

Draft Preprint. In this paper, we review algorithmic bias in education, discussing the causes of that bias and reviewing the empirical literature on the specific ways that algorithmic bias is known to have manifested in education. While other recent work has reviewed mathematical definitions of fairness and expanded algorithmic approaches to reducing bias, our review focuses instead on solidifying the current understanding of the concrete impacts of algorithmic bias in education—which groups are known to be impacted and which stages and agents in the development and deployment of educational algorithms are implicated. We discuss theoretical and formal perspectives on algorithmic bias, connect those perspectives to the machine learning pipeline, and review metrics for assessing bias. Next, we review the evidence around algorithmic bias in education, beginning with the most heavily-studied categories of race/ethnicity, gender, and nationality, and moving to the available evidence of bias for less-studied categories, such as socioeconomic status, disability, and military-connected status. Acknowledging the gaps in what has been studied, we propose a framework for moving from unknown bias to known bias and from fairness to equity. We discuss obstacles to addressing these challenges and propose four areas of effort for mitigating and resolving the problems of algorithmic bias in AIED systems and other educational technology.

show abstract

Section: Formal Fairness and Its Application (In A Messy World)mentioning

confidence: 99%

Algorithmic Bias in Education

Baker¹,

Hawn²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, the widely used demographic parity is defined such that the probability of yielding a positive prediction should be the same across protected groups (e.g., grouped with genders or races). However, demographic parity is not ideal in that (1) it only ensures aggregated fairness while ignoring individual fairness and (2) it does not reflect the true tendency of a disadvantaged group [14,17,19]. For example, female students might have the tendency to enroll in various courses to find the best fit and thus tend to drop out at the start of courses after final course schedules have been planned [21].…”

Section: Equalized Oddsmentioning

confidence: 99%

“…However, limited attention has been paid to the fairness of prediction with ML in educational settings [14]. Studies have shown that ML models can be biased by demographic factors such as genders [4,37].…”

Section: Introductionmentioning

confidence: 99%

Yet Another Predictive Model? Fair Predictions of Students’ Learning Outcomes in an Online Math Learning Platform

Xing

Leite

2021

LAK21: 11th International Learning Analytics and Knowledge Conference

View full text Add to dashboard Cite

To support online learners at a large scale, extensive studies have adopted machine learning (ML) techniques to analyze students' artifacts and predict their learning outcomes automatically. However, limited attention has been paid to the fairness of prediction with ML in educational settings. This study intends to fill the gap by introducing a generic algorithm that can orchestrate with existing ML algorithms while yielding fairer results. Specifically, we have implemented logistic regression with the Seldonian algorithm and compared the fairness-aware model with fairness-unaware ML models. The results show that the Seldonian algorithm can achieve comparable predictive performance while producing notably higher fairness. CCS CONCEPTS• Applied computing → Interactive learning environments;• Computing methodologies → Machine learning algorithms.

show abstract

“…It would probably not be an overstatement to say that fairness in AI is quickly becoming its own sub-field, with a new annual ACM conference on Fairness, Accountability, and Transparency having been inaugurated in 2018 17 and relevant research appearing at many impactful publication venues, such as Science (Caliskan et al, 2017), NIPS (Pleiss et al, 2017;Kim et al, 2018), ICML (Kearns et al, 2018), ACL (Hovy and Spruit, 2016;Sun et al, 2019;Sap et al, 2019), KDD (Speicher et al, 2018), AAAI (Zhang and Bareinboim, 2018), and others (Dwork et al, 2012;Hajian and Domingo-Ferrer, 2013). There is also recent work that examines fairness and ethical considerations when using AI in an education (Mayfield et al, 2019;Gardner et al, 2019).…”

Section: Increased Attention To Fairnessmentioning

confidence: 99%