The benefits and limitations of using key informants in library and information studies research

Introduction

This paper explores the ways that key informants are used in library and information studies research. The topic is worth exploring for several reasons: both authors have used key informants in their research and have found them valuable, for many reasons, yet key informants are not clearly identified or defined in library and research methods texts. It is likely that the contribution of key informants to knowledge creation in this discipline varies from other disciplines.

Key informants are knowledgeable individuals who contribute a perspective on a research phenomenon or situation that the researchers themselves lack. They are not usually research participants per se (that is, they are not the subjects of the research; they provide information about those subjects) but contribute to expanding a researcher’s understanding and precise insights and help to reduce potential bias. Engagement or collaboration between a researcher and key informants is not often reported in library and information studies and is more usually seen in anthropology and in ethnographic studies (Tremblay, 1982). It is rarely reflected on in the library and information studies literature and its value is underestimated; it may be treated as occurring outside the formal data collection process and not be included in reported analysis. The informality of the relationship between researcher and key informant invites neglect of acknowledgement. The variety of degrees of engagement or collaboration may make it hard to determine what should be included in reportage, and what not. Consultation with a variety of parties with a peripheral or tangential interest in a serious research topic is a common aspiration for a researcher at the conceptualisation stage of a project.

Both authors have a strong interest in the application and use of research methods and methodologies in library and information studies (e.g., Cossham, 2017a; Williamson and Johanson, 2014). The research for this paper started as an analysis of the impact of key informants on the first-named author’s PhD research (Cossham, 2017a; Johanson was one of her supervisors). The benefits to the research of engaging with key informants (termed expert practitioners in the thesis) were clear to both authors, but there was no reason to analyse their use in detail in the thesis itself, although it was explicitly but briefly described in the method and was part of the ethics approval.

This paper provides a more detailed analysis of the benefits and values of the key informants. We examined whether there were similar approaches used in library and information studies, and expanded the research to look at how key informants are used in the discipline more generally. Informal conversations were held with colleagues and Johanson’s co-researchers to identify what use they had previously made of key informants and how they had themselves served as key informants in research. Our sources of data are therefore threefold: Cossham’s use of expert practitioners in her PhD research, the library and information studies and research methodology literature, and Johanson’s conversations and emails with colleagues, co-researchers, and former co-authors.

The paper first defines the term key informant and examines the characteristics of key informants (and who are not key informants) in research. It reviews the library and information studies literature to see how key informant use is represented. It then outlines the benefits of key informants to Cossham’s research presents the findings from Johanson’s recent investigation. Finally, it sets out the values and limitations of using key informants and makes suggestions about their potential further use in library and information studies research.

We emphasise that the discussion applies to the social sciences rather than to the sciences where the situation regarding key informants may be quite different.

Who are key informants?

The term key informant seems to come originally from ethnographic research and is used widely in cultural anthropology. In fact, in its early use in 1939 it was known as ‘the anthropological technique’ (Tremblay, 1982, p. 98). It was regarded as an ideal means of purposeful entrée to a cultural group to focus on identification of its characteristics, when interviewing for research was still a novelty (Skinner, 2012, p. 26); conventionally it was the precursor to a survey and widespread interviews (Tremblay, 1982, p. 103).

More recently, it has been used in health, especially public health, social welfare, and education research but there is little documented use in library and information studies research, and little written about key informants as part of research methodology. Few library and information studies research methods texts address the concept of key informants per se, as far as we can determine; the concept is, of course, covered in ethnographic research methods texts. Library and information studies research does use ethnographic methodology more generally (see, for example, Khoo, Rozaklis and Hall’s (2012) survey of this use and Williamson’s (2006) paper on ‘Research in constructivist frameworks using ethnographic techniques’).

Characteristics of key informants

Most definitions of key informants focus on expertise and a willingness to share information. Marshall (1996) emphasises that they are knowledgeable, willing to participate, communicative, impartial and have a role in the community or understanding of the phenomenon that gives them information that the researcher is seeking: ‘an expert source of information’ who can, ‘as a result of their personal skills, or position within a society, … provide more information and deeper insight into what is going on around them’ (Marshall, 1996, p. 92). They have insider information or knowledge about a concept, situation, group, culture or subject that they are willing to share with a researcher (Bernard, 2018, p. 153; Marshall, 1996; O’Leary, 2014, p. 191; Schutt, 2004 as cited in Pollak, 2015), and they provide a perspective that the researcher (as outsider) could not otherwise obtain. Their status within an examined community of interest can yield invaluable observations (Adekannbi and Adeniran, 2017; Adekannbi and Dada, 2017; Innes, Jackson, Plummer and Elliott, 2017; Katiambo, 2015; Marshall, 1996; Tremblay, 1982). Debono (2002) defines key informants as ‘those who can describe with expertise and credibility specific social impacts delivered by a library’ (p. 86); the suggestion that key informants have the ability to reflect and to share advantages of their perspicacity is found widely thorough the literature. Key informants are likely to have expertise in a particular field or about a particular topic or organization and may be termed experts (e.g., Bogner, Littig, Menz, 2009) or specialised informants (e.g., Bernard, 2018) in some contexts.

Key informants may have secondary expertise, that is, expertise or knowledge about contexts where the participants may have a limited perspective (such as children’s points of view, e.g. Walter, 2003) or lack deep knowledge of a broader situation (such as patients with a particular medical condition), or may be less able to evaluate before and after situations (such as, what was taught vs what the students demonstrated they learned, or, how a new system was supposed to be used vs how it was actually used; e.g., Garanzini, 2003; Katiambo, 2015). For example, in a project that investigates what primary school children like to read, a researcher may count parents, teachers and librarians as key informants while the children themselves will be the participants (e.g., Pickard, 2002). Commonly key informants are the providers of a service and the research project evaluates the impact and use of that service on participants. For example, Nyareza and Dick (2012) had four key informants from community radio and twenty-five participants from the farming community who listened to that radio. Public medical and social services are also common examples here.

O’Leary (2014, p. 192) lists different types of individuals who might function as key informants: experts (at the top of their field), insiders (in an organization, culture or community), the highly experienced, a leader, the observant, and those with secondary experience. Lincoln and Guba (1985) suggest that naturalistic inquiry research can use (key) informants to get around the problems caused by gatekeepers who not only have the necessary information but also control the kind and level of access that a researcher can have to a community or group. They emphasise that ‘the term is “informant”, not “informer”; informers should be devoutly avoided because they will almost certainly have an axe to grind that does not mesh with the naturalist’s purposes and/or that falls outside the pale of ethical inquiry’ (p. 258; emphasis in the original). In a similar way, O’Leary lists as possible archetypal key informants, the gossip, stool pigeon, and ‘the Ex’, but it seems to us that including these types as key informants does depend very much on the research in question and the kinds of information that the researcher is after. Each of these types have their own agenda and bias; in our view they are better characterised as informers than informants.

Key informants can be used to generate primary data especially, perhaps, to enhance the meaning of quantitative or non-personal data such as organizational documentation.

Who are not key informants?

We note that it can be hard to distinguish who the key informants are in a research project, and one way of refining the meaning of key informants is to describe what lies outside their purview, as we attempt to do here.

Some people involved in research are more clearly not key informants, although they may be significant to the research. Research supervisors are a required part of student research. They may have expertise about the research topic, but they are required ethically to encourage the originality and initiative of the candidate. Supervisors are, therefore, unlikely to be key informants.

The researcher may read the writings of other researchers with expertise in the area; however, this is part of the literature review process. Significant authors may be key informants if the researcher happens to have access to them, but like research supervisors, they are usually an integral part of the research process as published evidence in preliminary stages of a research event.

A discussion with a colleague about an aspect of research may be informationally no different to a planned consultation with a well-known specialist or expert. Every researcher, whether academic or practitioner, is part of informal networks that provide vital information and suggestive reflections (de Alwis, Mujid and Chaudhry, 2006, p. 364) yet these networks do not constitute key informants as such since they are coincidental to the project-in-hand and exist independent of the research project.

Individuals piloting the data collection instruments serve a different purpose and are also not key informants (although key informants may also be involved with this stage) because their inclusion is primarily to test the instruments rather than to provide expert content on the topic of the research.

Participants in general are not key informants either. Although some library and information studies research seems to suggest that any participant is a key informant, this departs from most usage of the term in the discipline, misrepresenting key features of the role. The most frequent use of key informant in the sense of participant was found in library and information studies research from Africa; it is not clear whether this is coincidental or whether there are cultural influences around the African use. This use was noticeable in both Google Scholar and EBSCOhost results (see discussion of literature below).

Key informants are generally not the main focus of a research project or the primary subjects of study (in that case, they can be considered participants). Using experts (such as in the Delphi method) does not mean those experts are key informants since the focus is on those experts and their knowledge. Elite interviewing seems to focus on the role of the interviewee as an elite within their organization or group, including having seniority, high status, industry experience, and long tenure in the organization (while the organization itself may have high status and recognition); (Welch, Marschan-Piekkari, Penttinen and Tahvanainen, 2012, unpaged). The term seems to be used more in political science and business; Kezar (2003, p. 397) notes that ‘elite interviews are more prevalent in journalism than academic research’ where elites are ‘the primary subject of study’, as they are in Delphi research.

There are similarities and overlaps between these groups of experts (key informants, Delphi participants, elites), notably in the way that their conception of what is relevant is brought to the researcher’s project. Key informants are likely to be experts; they are not necessarily, if at all, elites in an organization or group. We note that while elite interviewing has been used in library and information studies, it is also less visible as an approach.

How are key informants presented in the literature?

Our literature search focused primarily on library and information studies and included: discipline-specific article databases accessed through EBSCOhost and Emerald, Google Scholar, and open web searches. Google Scholar revealed few articles: less than 80 across all the variants of library and information studies/science/LIS, library/archive/records compared with about 19,000 for ethnography along with key informants. While more broadly scoped, Khoo, Rozaklis and Hall (2012) identified only eighty-one studies of libraries and library users where ethnographic methods were used.

Key informants are often one of several sources of data, especially in mixed methods or case study research, and may support a quantitative or broad survey by providing rich descriptions and qualitative data (e.g., Blanco and Pozo, 2014; Cherinet, 2018; Gracy and Croft, 2006; Grefsheim, Rankin, Perry, and McGibbon, 2008; MacDonald, 2007; O’Leary, 2014; Skot-Hansen, Rasmussen and Jochumsen, 2013). They support methods where the researcher’s interpretation of data may need to be confirmed or corrected, such as in participant observation (Innes et al., 2017; Pashia and Critten, 2015, p. 85), or where a literature review constitutes the main source of data collection (Morris, Boruff and Gore, 2016, p. 351).

It is quite common for a researcher to use key informants in important aspects of the conceptualisation of research, most often near the beginning and prior to data collection, but also at times when a change of direction might be required. Conceptualisation is the least systematised and most neglected stage in research instruction in the social sciences (Wolf, 2004, pp. 74–76). O’Leary (2014, p. 191) notes

The insights you can gather from one key informant can be instrumental not only to the data you collect, but to how you process that data, and how you might make sense of your own experiences as well as the experiences of others.

Key informants can guide the development of the research, scoping, providing background, defining concepts, or heavily in a pilot stage (e.g., Cossham, 2017b; Oppenheim, 1966 as cited in Wilson 1985). In her chapter on questionnaires, Williamson (2014) wrote about starting

with unstructured interviews with key informants, as a way of understanding the topic, particularly if my partners and I have insufficient knowledge of the topic involved. For example, before we undertook a study with the Australian Plants Society Victoria, about data and information management and sharing in environmental voluntary groups (Kennan, Williamson & Johanson, 2012), we conducted two very long interviews with key members of the organization to gain understanding of the types of questions and prompts we should use with other participants. (p. 392)

Curiously, the article Williamson cites does not mention these interviews, although Johanson confirms they were essential to the research. He explained that the key informants provided ongoing advice throughout the two-year research project. This raises the question as to why key informants are not considered a formal data source and whether they are widely used but are under-reported in library and information studies research generally.

Key informants may function as the representative(s) of a particular group (Hanell, 2016) where any member could have been selected, or where the researcher becomes unable to access a particular situation during the research project: ‘After access to the committees was denied, several people acted as key informants (Marshall, 1996) keeping me up to date with developing events’ (Bathurst, 2006, pp. 234–235); this supported and triangulated Bathurst’s other data collection of ‘observations, interviews and concert attendance’. Key informants may also be hugely important in indigenous research, where the researcher is from outside the community or needs to learn a different epistemological approach to research within a minority community (Johanson, Williamson, and Schauder, 2007; Laycock, Walker, Harrison, and Brands, 2011; Passingan, 2013). Passingan describes how key informants in remote communities in Papua and New Guinea can be used to triangulate or confirm the accuracy of research data collected in other ways.

Extra support for interpretivist research can enhance the findings and ensure that they are considered more trustworthy. Guba (1981) suggests four criteria for evaluating the trustworthiness of research: credibility, transferability, dependability, and confirmability (pp. 79-81; see also Williamson, 2014). Incorporating expert perspectives into a project can improve the research performance based on these criteria (Cossham, 2017b). Key informants can also be used to gather information that forms the basis of a much more extensive data collection (such as a large survey that will collect quantitative data) and to help select participants in focus groups whose data can validate or contradict the survey (or other) data (Tremblay, 1982). It may be that a researcher only needs to make a nodding acquaintance with a marginal topic, but must do so, nevertheless. For this passing purpose, consultation with key informants can be quick and reliable.

How did the first author use key informants?

In her PhD Cossham (2017b) used key informants for core purposes. She has a background in cataloguing and teaching library and information studies, along with a strong inclination to categorise and classify, and a tendency to understand the world through analogies and mental images. This means she has a particular way of thinking about the bibliographic universe and certain assumptions about the mental models that may be held. She had been a lecturer for eleven years at the time of starting the research (although she had been a practitioner for fourteen years before that), meaning she did not have current experience of how library catalogue users thought about the catalogue or the bibliographic universe that it represents.

The research was interpretivist and within that, social constructionist, focusing on what individuals understood of a social object: the bibliographic universe. It took a phenomenographic approach. Phenomenographic research elicits ‘the qualitative variation in people’s experience of the world around them’ (Bruce, 1999, p. 32). It does not focus on a single way of experiencing the world, but on the variety of ways (Cossham, 2017b).

To ensure her expertise about the catalogue and the way it should ideally be used did not limit her understanding of what the main participants (library catalogue users) might think, and to help conceptualise the research as a whole, she conducted interviews first with expert library practitioners (key informants) who could provide additional perspectives on library catalogue users’ possible mental models of the bibliographic universe. That is, they were not asked about their own mental models of the bibliographic universe, but about what they thought library catalogue users’ mental models were.

There were five practitioners from different libraries in New Zealand and Australia: two from academic libraries (one cataloguing librarian, one reference librarian), one from a specialist academic library, one from a public library (heritage librarian), and one from a special library. These individuals all had postgraduate qualifications and had been librarians for twenty years or more; some were in senior management roles; most had experience in several library sectors and nationally (Cossham, 2017b, p. 71) and can be considered experts in the discipline. All the key informants were known to the researcher before she began the research and were selected to obtain a representative sample of libraries and roles.

Interviews were semi-structured to provide a way to compare data across the key informants but each expanded in different directions depending on the key informants’ knowledge, role in their library, type of library, and type of users. Questions were used as a starting point to enable the researcher to understand both known unknowns as well as unknown unknowns. Input from this group also influenced the wording of questions that were asked in the main data collection phase interviews and the cards used in the accompanying card sorting activity. It also suggested explanations of user choices or likely choices when using or not using a library catalogue. Two of the expert practitioners remained in contact with the researcher through the process; however, this was more for their personal interest in the research than because the researcher sought to maintain active dialogue with them.

The experts helped Cossham to bracket and minimise her assumptions of how library catalogue users thought, and to expand her understanding of the range of ways of experiencing the phenomenon (the bibliographic universe) by drawing on a variety of expertise and knowledge of library catalogue users (Cossham, 2017b, p. 160). They ensured that she approached the interviews with a broader understanding of how users currently think about library catalogues, and with a clear picture of popular conceptions and even misrepresentations that library catalogue users had of the bibliographic universe.

This phase was framed as a data collection phase in its own right and included in the research proposal and ethics applications, and helped support the credibility, dependability and confirmability of the research. However, it contributed primarily towards the researcher’s understanding and the development of the main data collection phase, and little was written about it in the thesis other than as part of the research method and some brief reflections in the concluding chapter. This reflects the confined presentation of the contributions of key informants in much of the library and information studies literature: key informants may be used but are rarely acknowledged or written about to any great extent. The tendency for academics in library and information studies to share collaboratively with other academics, and for practitioners to share with each other separately may account for the gaps in the literature where special assistance by individuals in either category is not recognised (White and Cossham, 2017), to the detriment of the profession and the progress of disciplinary research.

Information about key informants from key informants

If we move the focus from a specific use of key informants in a single research thesis, to find out how veteran researchers in Australia have been key informants and used key informants themselves, and what their reflections show, we can span broader horizons of experience. There is strong consensus with the first author’s analysis of her individual use.

The researchers approached individuals already known to them from their own professional networks. All of the five chosen individuals had managed libraries, had undertaken much research, had received many grants, had investigated issues overseas, had employed researchers on-staff and as consultants, and had written extensively about their work in mainstream literature and published it in their own organizations. They were more experienced than the thesis key informants, with thirty plus years’ experience.

The co-authors of this article asked them a series of questions by email in August 2018, and received prompt, thorough replies. Where any doubt about the emailed data existed, phone conversations clarified uncertainty. It can be safely asserted that their total of at least 88 consultations with key informants demonstrated their bona fides. Further questions were asked about the reasons for using key informants, or not using them along with questions about the wide-ranging values of use, and measures for evaluating the effectiveness of key informants. A final question asked about how key informants themselves benefited from acting as key informants; we assumed that the advantages of participating in information-sharing might be mutual.

It is worth highlighting several observations from the five recent participants that were not mentioned anywhere in the library and information studies literature, as far as we can ascertain from our searches. One interesting comment was that sometimes a researcher can benefit from communicating with a key informant over the life of a research project, as directions or plans change, rather than once (or twice) only at any given stage. This notion of a long-term relationship is more common to anthropological research. Another alerting comment was that in a consultancy requiring research, the key informant is also the client. In this situation the researcher is under an obligation to find out all that is possible from the client (as client and key informant) as well as allowing the client to tell them what they thought was most relevant. We were reminded that key informants might share war stories with a researcher and suggest traps to avoid. None of this practical advice could be found in the literature, either in reported research or in research method instruction.

Another useful observation related to the benefit of the interactions to the key informants themselves, rather than to the researcher in isolation. We could find no allusion to this knowledge transfer in the literature. While key informants need to be disinterested, impartial and balanced, there is no reason why they should not also benefit from participating in the research. In fact, mutually-beneficial knowledge was shared. One of our five key informants told us:

A sense of contributing to worthwhile outcomes … of the value of the informant’s input (given either privately or publicly) are proper benefits for the informant.

Others noted that a key informant could benefit if they were forced to update their own knowledge or research as a pre-condition of their participation, or by unburdening themselves of a chronic concern.

Two participants mentioned the necessity for trust between the researcher and informant, and vice versa, if the shared knowledge was to be relied and acted on. Transparent motives for investigative queries and corresponding responses were noted as essential: both a researcher and a key informant have ethical responsibilities (cf. O’Leary, 2014).

The value of key informants

We conclude from our three data sources (the library and information studies literature, use of key informants in a PhD thesis, and five participants who had used (and been) key informants themselves) that there are a great many advantages to using key informants other than to expand a researcher’s understanding and highlight any unusual assumptions or biases prior to data collection and analysis.

In a practical sense, key informants can assist with the creation of the research project. They may:

• help a researcher frame preliminary understanding (e.g., refine the language of questions prior to interviews, provide background detail for a case, provide detailed information about a wide range of aspects of a phenomenon);
• identify the most pertinent questions or approaches to the research phenomenon;
• help a researcher select participants from a particular group or community, identify the principles for selection, or select other key informants in a snowball sample (or when sampling fails) (Pollak, 2016); and
• provide qualitative and rich detail about a phenomenon, community or case.

Key informants can help to give prompt answers to these types of fundamental questions:

• Can my research question be answered? Can my problem be solved?
• Is there a means to answer my research question?
• What will be the effects of answering my question?
• Would an answer be of practical value?
• Will the research answers remain significant for long?

Tremblay (1982) suggests that key informants can provide different kinds of data, mentioning ‘definitional, objective and judgmental’ (p. 99), each of which contributed to the research objectives. According to Marshall (1996, p. 93; possibly paraphrasing Lincoln and Guba, 1985), ‘The principal advantages of the key informant technique relate to the quality of data that can be obtained in a relatively short period of time’. This includes using key informants instead of large data collection approaches to make the research faster (Marshall, 1996; Pauwels and Hardyns, 2009). Pauwels and Hardyns’ findings suggest that expert key informants can provide ‘reliable and valid measures of social cohesion and disorder on two rather small units of analysis’ (Abstract); a large sample is not required, in their opinion. Bognor, Littig and Menz (2009, p. 2) support these points in their discussion of expert interviews, noting ‘Conducting expert interviews can serve to shorten time-consuming data gathering processes, particularly if the experts are seen as “crystallization points” for practical insider knowledge and are interviewed as surrogates for a wider circle of players’.

Key informants can provide information about participants or situations where the participants are less able to provide the perspective themselves or where the researcher cannot themselves know the research population in detail. Similarly, they may have knowledge of a phenomenon, case, group or topic, especially knowledge beyond that which the researcher could be expected to have, or insider knowledge where the researcher is not an insider, especially where the research project is consultancy for a client. They may provide nuances that cannot be obtained elsewhere. They can provide information about best practices or how something is supposed to work (e.g., IT or recordkeeping professionals on how the systems are designed to be used; medical professionals on the significance of particular services or procedures).

Key informants enable an emic (insider) perspective, not just an etic (outsider) perspective in ethnographic approaches, according to Pickard (2013), who notes that ‘an emic perspective compels the recognition and acceptance of multiple realities’ and ‘documenting multiple perspectives of reality is crucial to understanding of why people think and act in the different ways that they do’ (Fetterman, 1989 as cited in Pickard, 2013, p. 139). One of the five current contributors noted the dual perspectives: as a researcher in her own right she used key informants to monitor her projects ‘to make sure I’m on track’ and up-to-date on a topic or field, and for another specific project she valued ‘deep insights’ from other key informants. Key informants can also correct, expand or refine data from participant observation. Several of our five contributors (wearing their researcher hats) valued the sophisticated, frank and balanced data provided to them by their own key informants.

Incorporating key informants in a research project may be a means to reduce the tension between theory (abstractions) and practice (reality) and is also an implicit acceptance that each researcher has their own practical experience and theoretical understanding that may not be universal or discipline-wide. Working with such experts can enable a researcher to focus very effectively on how consumers experience or understand the subject of the research, since the experts’ understandings will encompass a wider range of end uses than the researcher might be familiar with. As noted, one use of key informants identified by Cossham was to balance the researcher’s knowledge and expertise and help identify assumptions and reduce bias.

Additionally, they can be valuable where the research is trans- or inter-disciplinary, or trans-organizational, as well as across different sectors within (in this case) the library and information studies disciplines (libraries, archives, recordkeeping, information management, cultural heritage institutions) or the sub-sectors within any one of those sectors (e.g., school, public, academic, research, special, state, national libraries, and so on). Education and knowledge provision have always been closely linked, but the added dimension of ubiquitous networks requires integration and collaboration even further; key informants can help a researcher to achieve this.

For the academic researcher, key informants provide an opportunity to incorporate a practitioner perspective drawn from different areas (e.g., from special collections, archives, academic or public libraries, recordkeeping, digitisation projects). Likewise, they can help practitioners understand other areas of their profession as well as the more theoretical perspectives that may be brought by academics. In both instances, and especially when researching a topic or situation that is common to many information organizations (such as use of cataloguing or classification, user education, or digitisation of cultural heritage), it also helps to ensure that the researcher’s identified problem is relevant to those working with the topic or dilemma and is accurately stated.

Research has shown that academics and practitioners frequently have quite different perspectives and there is a notable gap between research and practice in library and information studies (see for example, Booth, 2003; Chang, 2016; Haddow and Klobas, 2004; Joint, 2005; White and Cossham, 2017). Key informants can foster engagement and collaboration between the researcher and the group or culture that they are researching, as well as more generally between practice and theory, practitioners and theoreticians.

Limitations of using key informants

While there is a lot of value to a research project of using key informants, there are some significant limitations around using them. Many of these are common to any research participant but we focus on those which have greater impact when applied to key informants, given the influence on the course of research that key informants have. They are also mentioned in older ethnographic research methods texts (Marshall, 1996).

The most notable limitation is that of selecting the right key informants. Without some insider knowledge, the researcher may not know who the most appropriate individuals are to be involved, yet their knowledge is likely to be limited which is why they would seek to use key informants in the first place. A researcher is unlikely to begin a research project with no relevant knowledge and may take a snowball approach to locating key informants, refining this as they progress. Alternatively, they may have specific needs that only a set group of individuals can address (e.g., leaders of a particular organization). That is, there is a tension between knowing enough to select key informants and selecting key informants in order to know enough.

Unless the key informants are already known to the researcher, it may be difficult to determine their expertise, both in extent and appropriateness to the research. Some may be more obvious choices due to their role in an organization, while others, especially those who self-select, may have hidden agendas and want to twist the research in a particular direction. McKechnie, Julien, Pecoskie and Dixon (2006, unpaged) identify the power that an informant has:

The term informant (for example, in the case of key informants in naturalistic observation and ethnography) carries with it the sense of the individual determining just what will be shared: it is the informant who is in charge and who informs the researcher.

Again, a researcher may refine their understanding of key informants’ expertise over time as they continue to work with them, or as they make new discoveries about the research topic through other forms of data collection. Bernard (2018, p. 156) emphasises that a researcher should not choose key informants too quickly in ethnographic research. Additionally, he notes that he has found ‘the best informants to be people who are cynical about their own culture … . They are always observant, reflective, and articulate’.

Power imbalance needs to be avoided. The informants may have considerable bias or potential bias without even being aware of it. They may want to present their perspective as absolutely correct, or advocate for a particular approach to the research based on their own disciplinary background or research strengths (see, for example, McKenna, Iwasaki, Stewart and Main, 2011). A key informant may also have a narrow experience that is not immediately obvious, while a self-described opinion leader may actually be an isolated advocate with an agenda atypical of a researched group or community. It may be hard to determine this initially: Taras et al. (2015, p. 2760) recommend holding a limited number of meetings with members of the community to contextualise what community leaders have said. The other side of this is that in some instances, using a particular individual as a key informant may discourage others from talking to the researcher (Bernard, 2018, p. 156).

As with any participants, there may be cultural or social implications framing how the key informants interact with the researcher. While sharing of knowledge is generally considered to be positive in Western research contexts, giving away or sharing the knowledge of a particular group or culture may be prohibited by that group or culture. Some people are more easily able to express their understanding and have greater awareness, breadth of perspective, and cultural sensitivity (see Fetterman, 1998, p. 47) than others. These considerations are common to any individual who might participate in a research project and not limited to key informants but still worth remembering as a limitation.

Some key informants do not want to share their own intellectual property, fearful of how their knowledge might be used or the consequences of that use. They may have limited ability to be open or may withhold certain information out of fear, privacy considerations, or because they think it is not relevant to the research. These may be valid concerns but need to be explicated for the researcher to be able to judge their relevance. They also make an individual less likely to be a key informant: as we noted earlier, key informants must be willing to share their knowledge trustingly and transparently. How the researcher determines that they are being given honest and accurate information is also a challenge (Bernard, 2018, p. 156; O’Leary, 2014, p. 192-193), especially if there is a culture of story-telling, of saving or preserving ‘face, or simply an attitude that it might be fun to trick outsiders.

There are ethical dilemmas around the confidentiality of the information shared, which potentially can put either the key informant or the researcher in an unethical position. Publicising confidential information is an obvious error. There are instances in anthropology, for example, where leaking has proved fatal to the life of a research subject (Griffiths, 2018, pp. 100-103).

As a final limitation, some key informants may not want to be mentioned in published research. This potentially undermines the credibility of the research because the researcher may not be able to prove how they know what they know.

Conclusion and recommendations

The use of three bundles of data in this research has resulted in a more coherent account of the role of key informants in library and information studies than has previously been published. Our pooled sources were the PhD research of the first author, the available literature (primarily journal articles), and a discussion with five experienced researchers who had used key informants in their own research as well as acted as key informants themselves. The field of information and library studies is unusual in that acknowledgement of the contribution of key informants by researchers is quite casual, often missing altogether, despite indications that key informants are used. In contrast, the disciplines of ethnography and anthropology are systematically committed to grateful approbation.

It was determined that key informants had common features. They were impartial experts willing to share their knowledge in the pursuit of further knowledge. Often, they were leaders in their respective communities, enabling entrée to a group where cumulative knowledge was exclusive, and provided in-depth information, frequently about the quality of an information service. Although we hesitate to label them gatekeepers, they usually did control access to vital knowledge but were more concerned with sharing than restricting. From a pragmatic point of view, we excluded academic supervisors of research theses, published authors, and experts used for one-off pilot-testing of research devices, because their functions were divergent within a research project.

The literature indicated that the functions of key informants may be characterised as assisting with conceptualising, scoping, backgrounding, defining, triangulating, and more recently, as assisting with embracing epistemological differences. Overall it was felt that the many benefits of the use of key informants outweigh any limitations. To the impressive list of functions, we added further values and benefits. When we reviewed all our data sources together, it was clear that key informants provided unbiased data quickly, helping to frame research problems in a solvable fashion. Their advice was practical, nuanced, corrective, and detailed. Using key informants helps the researcher to not only access invaluable knowledge, but also to assist in creating a community of willing participants. Notably key informants reduced tensions between professional practitioners and researchers, and between inter-disciplinary or trans-organizational groups.

Cossham recounted how she used expert librarians as key informants in her PhD research to provide practical and theoretical knowledge of how users of library catalogues thought about the bigger concept of a bibliographical universe. Her data analysis integrated their important contributions (from experienced senior librarians) into her thesis discussion, as well as explaining their contribution to her thesis conclusions.

The five recently surveyed researchers shared thoughts about use of key informants that added to the knowledge garnered from our other data sources. They pointed out that a key informant can act as a good companion throughout a research project, from beginning to end, and not only once or twice. Where research is consultation in fact, then the researcher is obliged to think of the client as the key informant, who may enjoy participating in a personal and professional journey, as well as benefiting from it from a management perspective. Finally, the necessity for ongoing trust between the key informant and researcher was nominated as significant.

The limitations we uncovered in relation to the use of key informants were to be found in any research project which deployed them, rather than just to research in library and information studies alone. It is important to keep limitations in mind. Difficulty in identifying appropriate and reliable key informants required special attention, related to which was the inevitability of a power imbalance in knowledge control. Negotiating the control involved many dimensions, including proprietary issues, privacy, confidentiality, and fundamental fears about how information might be used outside an organization and in future.

We have argued that significant key informants improve the quality of research projects, particularly by way of validity and reliability. Reputational benefits can accrue from use of key informants; both the reputation of the researcher and the reputation of the key informant may be enhanced. The explicit inclusion of key informants’ help in library and information studies in published articles, papers, and theses will work to the advantage of all parties. We suggest that library and information studies research will benefit from better acknowledgement of the (often neglected) role and use of key informants in projects.

Acknowledgements

Amanda Cossham acknowledges the support of the Australian Government Research Training Scheme (now Research Training Programme Scholarship), and the guidance and support of her supervisors: Steve Wright, Kerry Tanner, and Graeme Johanson (the paper’s co-author). We We acknowledge participants and thank them for their expertise and generous contribution to our research.

This paper was presented at RAILS: Research Applications Information and Library Studies conference in November 2018. The authors acknowledge the comments and feedback from the audience and have addressed many of the points they raised. We also acknowledge the extensive and thoughtful feedback from one of the reviewers, especially around the notion of elite interviews. We appreciate their interest in the research and have made additions and clarifications in the paper in the light of their comments. However, we respectfully disagree with the reviewer that we are discussing elite interviews or that a new term is needed and have kept the original term. The authors thank the conference chair for reviewing their revision of the paper.

About the authors

Amanda Cossham is a Principal lecturer in Information and Library Studies at the Open Polytechnic of New Zealand, where she teaches cataloguing and classification, social informatics and information literacy. Her research interests include knowledge organizƒation, cataloguing and classification, and continuing professional development of information professionals. She can be contacted at amanda.cossham@openpolytechnic.ac.nz
Graeme Johanson is a lifelong researcher, and keen advocate of thorough research methods. He belongs to the Centre for Organizational and Social Informatics, Monash University https://www.monash.edu/it/our-research/research-centres-and-labs/cosi Although he has 'retired' from active service (from libraries and university research) he continues several projects where research throws light on both theory and practice in professionally- and socially-beneficial ways. As an Associate Professor, he still supervises postgraduate research and welcomes opportunities to explore new themes. He can be contacted at Graeme.johanson@monash.edu

How to cite this paper