What role does cognitive bias play in information retrieval?, by Maria Nagle

 Maria Nagle works as a librarian at Welsh Government Information, Library and Archive Services. She has prior experience in public, academic and special collections library sectors. This article has been written in a personal capacity and does not necessarily reflect the views or policies of the Welsh Government.

Photo by Rachel Claire from Pexels

How does our world view affect our search experience?

Every day many of us make decisions based on information we have sourced online – decisions such as where to eat, choices based on what the weather forecast looks like, how to answer an enquiry in the workplace. We each experience the world individually, through the lens of our own assimilated knowledge, previous experiences and unique interpretations - how does this affect our searches for information and the decisions they lead to?

These natural cognitive biases and their implications for navigating our information-saturated society are examined in the thought-provoking paper ‘Identifying effective cognitive biases in information retrieval’, written by Gisoo Gomroki, Hassan Behzadi, Rahmatolloah Fattahi and Javad Salehi Fadardi of Ferdowsi University of Mashhad, Iran. This paper was featured online in the Journal of Information Science in March 2021 and can be accessed by CILIP members as part of their membership via the Professional Journals area of the CILIP website.

 

Article summary

This article was a surprising read, not least as it highlighted a dearth of previous research into the effect of cognitive biases on information retrieval; two areas with much scope for overlap and investigation. The authors of this paper confronted this gap with an interesting methodological approach, consulting twenty-five information retrieval experts to determine which cognitive biases they deemed the most important, followed by monitoring the search activities of thirty graduate students  supplemented by a semi-structured interview with the latter group.

It was fascinating to compare the top cognitive biases selected by the specialists with those appearing in the students’ searching activity and to reflect on the different priorities evident in these biases. Several of the biases highlighted in the study will be familiar to many, such as:

·       ‘Familiarity’ (prioritising familiar situations, resources or search tools)

·       ‘Rush-to-solve’ (tendency to make quick decisions when searching leading to error)

·       ‘Mere exposure effect’ (where the extent of information exposure influences users’ judgement and decision-making).

 

It was clear from the findings that the specialists’ opinions of common information searching biases differed somewhat from those evidenced in the students’ searching activity. This raises the question of just how well we, as information professionals, understand the priorities and pressures on our users as they search for information. The article’s conclusion offers another point of focus for information professionals; our responsibility to educate our service users and alert them to the risks of cognitive bias.

 

Personal and professional implications

Working in government services can be a fast-paced and sometimes risk-averse environment, and many government information professionals will not only recognise these biases in their service users, but may occasionally recognise these traits in their own search activity. This article prompted me to reflect on how our role, as information professionals, is fundamentally about staying aware of and alerting our users to the often harmful impact of these biases, putting in place checks and techniques to encourage more balanced and effective searches for information.

For example, using reflective prompts may provide a more stimulating and productive way of alerting users to bias during training sessions – open questions such as ‘how do you usually search for information? Why those particular tools?’ Likewise, when enquiry handling or literature searching, it is our responsibility to effectively communicate when the information requested is complex and multifaceted – addressing our users’ need to ‘rush to solve’ with a clear, objective guide to the information relevant to their request. Even promoting government knowledge and information management (GKIM) services can help achieve this aim; encouraging our colleagues away from poor quality search tools and demonstrating the value of our services all feeds into discouraging our users from making biased search decisions.

Understanding the discussion of these common biases through the additional lens of our duty as civil servants to uphold the values of integrity, honesty, objectivity and impartiality (i.e. the Civil Service Code) is a further reminder of the importance of advocating for sound information literacy skills and practice. All decisions made in government require a reliable and robust evidence base, free from bias. We are often asked to prove or negate certain concepts as part of our enquiry-based or literature-searching roles, a task that requires us to draw upon on our professional ethics to provide an impartial service and avoid returning a biased result. Maintaining and raising an awareness of these possible biases to our service users is a fundamental step to ensuring these values are upheld in a constantly evolving information society. Furthermore, continuing professional development and keeping up to date with developments in our field can help expand our horizons and ensure we approach our work with a clear, informed vision.

 

Lessons learned

The conclusion of the article focuses on the responsibility of librarians and information professionals to educate their users and help their understanding of cognitive biases. One of my take-home messages from this article is that, in order to be effective and impartial searchers we must constantly interrogate our own biases and maintain a self-aware approach to information retrieval, sharing this approach with our colleagues and service users.

This requires vigilance as it can be all too easy to fall back into biased patterns of thought. It can be useful to draw on the expertise of our colleagues and peers when trying to get an objective perspective. For example, sharing search strategies or asking for their interpretation of a particular research question can provide interesting insights and a stronger starting point for literature searching.

 

Points for reflection

This article provides an important reminder that we must scrutinise the decisions we make when searching for information and help our users do the same, highlighting the impact each choice will have on the quality of information returned. Sharing practice and experiences across organisations and sectors can help us broaden our understanding of the best ways to address bias – it may be useful to consider the following:

·       How do you alert your users to the risk of cognitive biases in training and enquiries activities?

·       What strategies do you employ to address your own cognitive biases when undertaking literature searching or enquiry work?

·       What resources can we develop to aid our service users and help them avoid making biased decisions when searching?

 

Maria’s article demonstrates the relevance of considering and incorporating publishing LIS research into our own practices. This is something that GIG would like to continue and expand on within this blog platform. So we will welcome your reports of recently published research that have given you cause for thought and may have gone on to impact on your professional practice. Please contact info.gig@cilip.org.uk to discuss an idea for a future posting.

A number of questions are, also, raised. How do you recognise and counter your own search biases? Are there formal mechanisms within your team to help you consider and address bias? Do you make provision to consider potential for bias when training service users? What forms of bias creep into other GKIM practice areas? Is there bias in records management, access to information, freedom of information, and what measures are in place to address them? We will be keen to hear from you either as comment to this article or to info.gig@cilip.org.uk.