Background
Affiliations may take many forms, including institutional, geographical, professional, political, ideological, commercial, reputational, or other social and organisational associations. Affiliation bias most commonly occurs during the evaluation and dissemination of evidence. It has been studied particularly within peer-review processes. In these settings, the factors listed above may influence perceptions of credibility, expertise, relevance, or scientific quality.
The bias arises because affiliations can function as heuristics, allowing evaluators to make rapid judgments. These factors may be used as proxies for research quality, leading to evidence being evaluated differently despite comparable methodological standards. Such judgements may advantage or disadvantage research independently of its scientific merits.
Affiliation bias overlaps with several related concepts, including prestige bias, authority bias, allegiance bias, and sponsorship bias. However, affiliation bias concerns the influence of an individual’s or organisation’s associations on the evaluation of evidence. By contrast, allegiance bias concerns commitment to a particular theory, intervention, position, or outcome influencing their judgement, while sponsorship bias primarily concerns distortions in the design, conduct, analysis, reporting, or publication of research that favour sponsor interests.
Research affiliated with prestigious institutions may, on average, have greater access to funding, infrastructure, training, and collaborative networks resulting in higher-quality research. However, affiliation bias refers specifically to situations in which affiliations influence evaluations beyond any differences attributable to research quality itself.
Example
Geographical example:
Fox et al. (2023) found that in the single-anonymous condition, where author identities were known to reviewers, authors from wealthier nations were 28% more likely to be invited to revise and resubmit their papers than authors from lower-income nations. In the double-anonymous condition, where author identities were concealed from reviewers, the rates of invitations to revise and resubmit papers fell to just 4% for wealthier nations, compared with authors from lower-income nations. Because author identities were experimentally concealed in one review condition but not the other, the findings suggest that geographical affiliation influenced reviewer and editorial assessments independently of manuscript quality.
Institutional example:
Tomkins et al. (2017) conducted a controlled experiment during the peer-review process for the 2017 ACM International Conference on Web Search and Data Mining. Each submitted paper was reviewed under both single-anonymous and double-anonymous conditions, allowing direct comparison of reviewer behaviour when author identities and affiliations were visible versus concealed. Reviewers who knew author identities were significantly more likely to recommend acceptance of papers from famous authors and leading companies than reviewers who were blinded to this information. Because the same manuscripts were being evaluated under different review conditions, the findings suggest that institutional affiliation and author prestige influenced assessments independently of the scientific content of the work.
Impact
Observational differences in acceptance or citation rates could reflect genuine quality differences. Experimental studies of peer review have demonstrated that revealing author identities and affiliations can influence reviewer recommendations and editorial decisions. In controlled comparisons of single-anonymous and double-anonymous reviews, manuscripts associated with prestigious institutions, well-known authors, or researchers from higher-income countries have received more favourable evaluations than comparable manuscripts assessed under anonymised conditions. These findings suggest that affiliation bias may influence which studies progress through peer review and, ultimately, enter the published literature.
Preventive steps
Several approaches may reduce affiliation bias during the evaluation and dissemination of research.
Blinding of author identities and affiliations: Double-anonymous peer review, in which reviewers do not know authors’ identities or affiliations, may reduce the influence of institutional prestige, geographical origin, professional status, or organisational reputation on manuscript evaluation. Similar anonymisation approaches may be applied to grant review, conference abstract assessment, and other evaluative processes where feasible.
Reviewer and editor training: Training programmes that increase awareness of cognitive biases and the potential influence of affiliations may help reviewers and editors recognise and mitigate inappropriate reliance on affiliation-based heuristics.
Transparency and oversight: Clear conflict-of-interest policies, monitoring of editorial decisions, and periodic audits of acceptance rates by institution, geography, or other affiliation characteristics may help identify systematic disparities.
Diversification of reviewers and decision-makers: Drawing reviewers, editors, and grant assessors from diverse institutional, geographical, and professional backgrounds may reduce the influence of shared networks or prestige hierarchies.
Open peer review and accountability measures: In some settings, increased transparency regarding review processes and editorial decisions may help reduce inappropriate reliance on affiliations, although evidence regarding effectiveness remains mixed.
