About

DOI: https://doi.org/10.5281/zenodo.17078016

Overview of Concepts¶

Positionality Silences Accessibility Representation Collaboration Durability FAIR Ownership Recruitment Unintended Use Transparency Availability Expertise Multivocality Impact Harmful Language Accuracy Historicity Methodology Privacy Provenance Reproducibility Documentation

Why a Bias Vocabulary?¶

There is a crisis of shared understanding around bias. The term carries different meanings across disciplines: archivists focus on categorisation issues, historians examine power structures, digital humanists address representation, and computer scientists consider training data biases. This vocabulary addresses the fragmentation by establishing a common language for discussing bias across disciplines, transforming abstract concerns into specific, identifiable manifestations that researchers can systematically address.

Forms of Bias¶

We have conceptualized bias as emerging in two overarching forms:

Discrimination: Biases that perpetuate unfair treatment or representation
Opacity: Biases that obscure processes, decisions, or perspectives

All expressions of bias documented here, along with their mitigation strategies, are connected to and rooted in reducing the impact of these forms of bias.

By centering expressions of bias rather than the lifecycle stages, this document allows researchers to focus on specific bias types regardless of where they occur in the research process.

Good-Better-Best Practices¶

Each expression of bias includes mitigation strategies organized in a good-better-best format:

Good: Basic approaches to addressing the bias
Better: More comprehensive strategies (automatically includes points from “Good”)
Best: Most thorough approaches (automatically includes points from both “Good” and “Better”)

This progressive structure allows researchers to implement strategies based on their available resources and project requirements. The structure has been humbly taken over from Chilcott (2019). We recommend you create your own Good-Better-Best schema (see template) that aligns with your research project and include it in your project documentation. The process of filling out this template will allow for your team to identify and discuss biases with regards to your project collectively.

Limitations¶

We acknowledge that we will inadvertently have missed certain considerations of bias, as well as forms of bias as they appear in other fields, such as cognitive biases. This overview is not intended to be - nor should it be interpreted as – conclusive or comprehensive. For any queries, comments or feedback, please feel free to contact us.