Beyond Fair Pay: Ethical Implications of NLP Crowdsourcing

By Boaz Shmueli, Jan Fell, Soumya Ray, Lun-Wei Ku. 2021

97 citations recorded by the source corpus.

The use of crowdworkers in NLP research is growing rapidly, in tandem with the exponential increase in research production in machine learning and AI. Ethical discussion regarding the use of crowdworkers within the NLP research community is typically confined in scope to issues related to labor conditions such as fair pay. We draw attention to the lack of ethical considerations related to the various tasks performed by workers, including labeling, evaluation, and production. We find that the Final Rule, the common ethical framework used by researchers, did not anticipate the use of online crowdsourcing platforms for data collection, resulting in gaps between the spirit and practice of human-subjects ethics in NLP research. We enumerate common scenarios where crowdworkers performing NLP tasks are at risk of harm. We thus recommend that researchers evaluate these risks by considering the three ethical principles set up by the Belmont Report. We also clarify some common misconceptions regarding the Institutional Review Board (IRB) application. We hope this paper will serve to reopen the discussion within our community regarding the ethical use of crowdworkers.

Subfields: Computer Science Applications, Safety Research, Sociology and Political Science, Natural Language Processing, Safety, Robustness & Interpretability

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

In Natural Language Processing

In Safety, Robustness & Interpretability

Citations in the corpus

Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text (2022, background): “Shmueli et al. (2021) cover a broader set of ethical considerations for crowdsourcing work, including potential psychological harms, exposing sensitive information about workers, and breaching workers’ anonymity.”
CrowdWorkSheets: Accounting for Individual and Collective Identities Underlying Crowdsourced Dataset Annotation (2022, background): “[54] identified risks of harm to crowdworkers engaged in NLP tasks.”
A Review of Human Evaluation for Style Transfer (2021, background): “…is also key to assessing the alignment of the evaluation with the models and task proposed (H¨am¨al¨ainen and Alnajjar, 2021), and to understand potential biases and ethical issues that might arise from, e.g., compensation mechanisms (Vaughan, 2018; Schoch et al., 2020; Shmueli et al., 2021).”
Just What do You Think You're Doing, Dave?' A Checklist for Responsible Data Use in NLP (2021, unknown): “necessary for the following reasons: (1) There is a lot of variation in the researcher perception of when ethics review is needed (Shmueli et al., 2021).”
ConvAbuse: Data, Analysis, and Benchmarks for Nuanced Abuse Detection in Conversational AI (2021, background): “Moreover, crowdsourcing is not without its own ethical issues (Shmueli et al., 2021).”
Bugs in the Data: How ImageNet Misrepresents Biodiversity (2022, methodology): “While the initial ImageNet dataset was crowdsourced using the Amazon Mechanical Turk platform, there has since been extensive work that has questioned the ethics of anonymous crowdsourcing platforms (Irani and Silberman 2016; Shmueli et al. 2021).”
Metadata Archaeology: Unearthing Data Subsets by Leveraging Training Dynamics (2022, methodology): “We first evaluate convergence dynamics of different probe suites (Section 3.1), validating the approach of MAP-D .”