Perspectives

Cognitive barriers in search

In general, I do not comment on articles that are only available online and have not yet been published in an issue. However an online article by Reijo Savolainen ¹ on ‘Cognitive Barriers to Information Seeking: A Conceptual Analysis’ is a very important contribution to the understanding of how we undertake the process of information seeking. The author addresses two challenging issues: the conceptualization the features of cognitive barriers to information seeking and the characterization of the impact of cognitive barriers on information seeking. This was undertaken by a very thorough review of the literature, and the bibliography extends to over 50 citations.

From the analysis, six barriers are described.

unwillingness to see needs as information needs;

From my own experience, working as an enterprise search consultant none of these barriers would be recognized by what is often a technology-led enterprise search team. The objective of the team is to develop intuitive search, often seeking to emulate Google Web search in the enterprise without understanding the futility of doing so.

Quite often there is only a single person with the responsibility for search delivery in an organization, despite the fact that search is used ubiquitously and a failure to find information can mean that a business-critical decision is made without the best available evidence base. The result is always a low level of user satisfaction and trust in search, and indeed the evidence from the 2015 Digital Workplace Trends report ² is that the levels of search satisfaction have been gradually decreasing over the last six years, probably because the size of internal digital repositories is increasing remorselessly.

The author comments in the conclusion of the article that a limitation of the study is that it approaches the cognitive barriers from the viewpoint of an individual actor seeking access to information and there is a need for expanding the research perspective by making use of ideas framed in practice theories, for example, because they would take into account how cognitive barriers are constructed socially within work teams and communities, which neatly takes me on to the next article.

Searching together

If you read and believe all the pronouncements from consultants and software vendors, then the picture of the digital workplace that emerges is of multiple instances of collaboration, usually virtually, at all levels and locations. My own experience suggests that this is a significant overstatement. Certainly, there is a substantial amount of working in teams, but, in my view, collaboration is a special case and not a generic case. I constantly refer to the work on collaboration personas that was carried out by a team at IBM, ³ which highlights the characteristics of different types of what might often be regarded as collaborative working.

Every piece of business software seems to have the support of team working as its primary objective. However, there is one enterprise activity that continues to be carried out by individuals and that is the process of search. Typically one person in a team will carry out a search to the best of their abilities and share the results with others, who make the assumption that search is a fairly straightforward exercise, and it is unlikely that the searcher missed anything important. There is no foundation for that assumption. Even if two searchers carry out an identical search at the same time on the same search application, they will take a different view of what is relevant to pass on to the team. This could have significant consequences for the work of the team. Readers based in the UK will remember that in 2000 the new Millennium Bridge across the Thames initially exhibited a wobble caused by in-phase stresses from people walking across it. This outcome had been published in a journal on earthquake research in 1993 but was not known to the design team. ⁴

In their ‘Coordinated Exploration: Organizing Joint Search by Multiple Specialists to Overcome Mutual Confusion and Joint Myopia,’ Thorbjørn Knudsen and Kannan Srikanth ⁵ explore the issue of coordinated search/exploration in some detail. I will note in passing that not many people in information retrieval research and development are likely to read Administrative Science Quarterly. The abstract to the article is so well written that I have reproduced it in its entirely below.

In this article, the authors use an agent-based simulation model to investigate how coordinated exploration by multiple specialists, as in new product development, is different from individual search. The authors find that coordinated exploration is subject to two pathologies not present in unitary search: mutual confusion and joint myopia. In joint search, feedback to one agent’s actions is confounded by the actions of the other agent. Search therefore leads to increasing mutual confusion because agents are unable to learn from feedback to correct their faulty mental models of the search space. Incorrect beliefs held by one agent lead to mistakes, and because it is unclear which agent was wrong, this confuses the other agent, either into revising (correct) beliefs or onto holding (incorrect) beliefs. Sharing knowledge aligns specialists’ mental models and counters mutual confusion by inducing coordination around particular search regions. Yet that very effort increases joint myopia, as agents prematurely reinforce each other into choosing from an increasingly narrow portion of the search space. In the extreme, high levels of shared knowledge induce agents to abandon their distinct search approach in favour of a lower common denominator. In coordinated exploration, increasing coordination efforts (such as by increasing communication) reduces mutual confusion but simultaneously increases joint myopia. Efforts to reduce joint myopia, such as by slow learning or lower levels of knowledge transfer, however, automatically increase mutual confusion. As modelled in our simulation, successful joint search needs to balance these two effects. Our results suggest that because unitary-searcher models abstract from epistemic interdependence, their predictions are potentially misleading for coordinated exploration.

In summary, this model suggests that the process of coordinated exploration is one that needs to be looked at in considerable detail to ensure that as technical solutions become available, the outcomes of real-life searches can be trusted for recall and relevance.

I have been looking into the opportunities and challenges of collaborative information seeking (CIS) for some time and have been very impressed with the work being carried out by Professor Chirag Shah at Rutgers University. He is the author of the only book on the subject and a review paper by Shah ⁶ that appeared in the Journal of Information Science in 2014. I was therefore quite surprised to find no citations to Shah’s work (which dates back several years) in this article. Perhaps a reason for this is the difference between ‘coordinated exploration’ and ‘CIS’, which illustrates rather neatly the difficulties of searching for information on novel topics. The approaches taken are complementary, and if this column does nothing else than connect two related areas of information science research, it will have been well worth the effort.

Inter-professional knowledge transfer

For the last year or so, I have been working for two global law firms on intranet-related projects. What I have found fascinating is the two-culture divide between lawyers (who generate the income) and business support (whose costs impact margins and partner income). In one firm, the business staff do not have their experience profiles on the personnel database. Knowledge management is a core business discipline in law firms, but the focus is on knowledge sharing between lawyers. I have also worked in one of the largest UK hospitals and here there is another professional divide, in this case between doctors and nurses.

This divide is examined in ‘Social Networks and Inter-professional Knowledge Transfer: The Case of Healthcare Professionals’ by Stefano Tassell, ⁷ reporting on a research project carried out in a hospital department situated in northern Italy. He collected data on the informal network of 1036 knowledge transfer ties among 118 professionals (53 doctors and 65 nurses) through a sociometric survey and qualitative data from 21 semi-structured interviews.

The author summarizes his research by reporting that found that the composition of professional cliques and the development of distinctive professional identities built upon status gaps and organizational roles tend to inhibit effective knowledge transfer between doctors and nurses: professionals are thus more likely to transfer knowledge within rather than outside their occupational group. But social structures not only constrain patterns of knowledge transfer they can also enable specific actors to facilitate the transfer of knowledge between professions. Thus inter-professional knowledge transfer is favoured by the network role of clinical directors, as subjects who act as mediators between doctors and nurses by occupying central positions in closely connected networks. Individuals who are motivated and legitimated, both intra-professionally and inter-professionally, to act as inter-professional knowledge brokers – namely, junior doctors and nurse managers – are more likely to gain access to valuable knowledge. The overall picture painted by these results is one of network structure interplaying with individual traits in shaping the dynamics of professional interactions.

It would be easy to say that this conclusion was not a surprise but that is not the point that I want to stress. What emerges from this article is the importance of knowledge boundary-spanners, in particular clinical directors, and by others in a position and with the vision to make the connections that are so important in knowledge transfer.

The article is 32 pages long and is one of the most elegant I have read on the subject of professional knowledge transfer. The bibliography is excellent and I would recommend any reader with an interest in KM to read the article both for the initial discussion of the issues and for the research methodology and outcomes.

An Indian perspective on CoPs

It has been my great good fortune to have had business experience in 39 countries, including Lithuania, Cyprus, Sri Lanka, China and Kuwait. There are many good books (notably by Richard Lewis ⁸ ) on how to manage cultural differences, but nothing quite prepares you for the reality on the ground. So much of the literature on KM, collaboration and social networking comes from US writers and US experience. Useful though this is it should not be extended to other countries, even those where there is a wide use of English in business, without considerable care to understand what is scalable and extensible.

I was delighted to see an article on ‘A Framework for Communities of Practice in Learning Organizations’ by Jyoti Jagasia et al. ⁹ Not only is the article based on research carried out in India, but the authors set out to develop a mathematical model that might predict the factors that would lead to successful Communities of Practice (CoPs) being established. The factors considered were the organizational environment (notably senior management and organizational culture), the technology environment, people-related factors (e.g., trust, participation, and personal values) and the existence of knowledge-based processes. The article begins with a concise summary of the work that has taken place to date, notably by Etienne Wenger and Richard McDermott.

When the article moves on to the research itself, it was clearly a major challenge getting a good response to the Survey Monkey questionnaire. In quite a number of cases, the data were not as variable as might be expected, leading to the supposition that the respondents had been instructed to complete the questionnaire by a senior manager. In all, some 30 factors were identified, and the table of these is a very useful contribution to the literature. Using regression analysis, the aim of the model was to see which of these independent variables had the most impact on CoPs effectiveness (CoPe). This indicates that 49 per cent of the variability in the CoPe-dependent variable can be captured by the independent variables, namely, community support, people factors, KM strategic alignment and the extent to which users help knowledge managers in identifying valuable expertise. Organizational factor did not emerge as an independent variable with influence on the dependent variable. Although this factor has emerged as important in other studies conducted in countries other than India, the authors’ research did not find this significant probably because Indian organizations are still in the early stages of CoP and KM development and hence the CoPs may not have yet matured to need a high level of governance and organizational structure.

I feel it is very important that we always take into consideration the national and organizational cultures of organizations when promulgating ‘best practice’. I continue to try to persuade multinational organizations that English is not their corporate language, just a default language that is spoken, written and understood to different levels of competence. Social applications present special problems, as people will always prefer to be social and to exchange knowledge in their primary language because they feel in command of the nuances.