Using Informal Conversations in Qualitative Research

Abstract

The aim of this paper is to promote a greater use of informal conversations in qualitative research. Although not a new innovation, we posit that they are a neglected innovation and a method that should become more widely employed. We argue that these conversations create a greater ease of communication and often produce more naturalistic data. While many researchers have written about the use of informal conversations in ethnography, as part of participant observation, we are advocating that these conversations have an application beyond ethnography and can be used in more general qualitative exploration that occurs in everyday settings where talking is involved. They can be used as the main method but also to complement and add to more formal types of data created through interview. Sometimes informal conversations are not only the best way, they are the only way to generate data. We use examples to show how we have used informal conversations in our research, which we interrogate and use to raise a number of, mainly ethical and methodological, issues. We discuss the main advantages and disadvantages of using this method, including the status and validity of data produced.

Introduction

Many qualitative researchers choosing to interview people will be familiar with the tale of what happens as you end an interview by thanking the interviewee and, just after you have turned off your digital recorder (or other recording device), the person begins speaking again, telling you things that were perhaps not covered in the interview schedule, or opening up in a more relaxed way and talking in greater depth about a particular issue (see Warren et al., 2003; p. 107) who write about the “after-the-interview interactional strip”). As you begin to feel that these data may be richer and more revealing than what had been said during the more bounded space of the formal interview time, a slight panic takes hold as you realise you have various choices available of what to do next, that your decision may have important implications, and must be made immediately. You can either (a) quickly close the conversation down by signalling that it is time to leave – perhaps by standing up and/or making a show of packing your things away; or (b) you can tell the interviewee that what they are saying is really interesting and ask if they would mind if you turned on the recorder again; (c) you can again inform the person that what they are saying is fascinating and ask them if they would mind you making contemporaneous notes; or, a further alternative is (d) to let them carry on talking and try and make notes of the conversation as soon as possible after you part company, based on what you can remember.

Like many situations involved in qualitative research, each of these four scenarios has advantages and disadvantages and your (the researcher’s) decision may be based on ethical, methodological and practical decisions. One of the concerns some researchers have with option (a) is a feeling that the researcher is not doing his/her job by choosing to ignore further additions to the story, or issues they are seeking to understand; an ethical concern with the rest of the other three options is that the researcher may feel that any data ‘collected’, or, as we prefer to say with our social constructionist hats on, ‘generated’, outside of the formal, bounded, interview time is somehow not eligible or legitimate as it is not something the interviewee signed up for when signing a consent form. A further worry, specifically with option (b), is that, as you turn on the recorder again, you are signalling to the interviewee that the formal interview is restarting and that subconsciously the interviewee may revert back to more guarded responses, the more natural flow of the conversation will be lost, and data produced less compelling or meaningful. Although option (c) may seem a little more naturalistic, it is difficult to maintain eye contact, note taking can be disruptive, you may have to ask the interlocutor to pause while you frantically try and scribble down their words in a neat enough form to read afterwards, and a pause in speech often leads to a pause in thinking. Additionally, you may be worried that you cannot capture the words verbatim and use the interviewee’s actual words in your research report or paper, and as you are only able to record a remembered approximation of the interviewee’s words they cannot be regarded or categorised as ‘real’ or ‘proper’ data. This last point also refers to option (d), although this strategy may be the one that creates the richest, most expansive and most authentic data of all. Moreover, Rutakumwa et al. (2020) suggest that, in some circumstances, not recording an interview may actually be the best approach, not ‘second best’.

Most of these scenarios outlined above happened when the first author was more of a novice interviewer, and he has now learned from (slightly bitter) experience that it is a good idea to keep the recorder running, even after the interview has been formally closed. However, when he does not leave the recorder on, he now tends to use option (d) as it is a method of generating data from what can be called informal, or natural, conversations, which is the main focus of this paper. Its primary aim is to promote a greater use of informal conversations in qualitative research. While many researchers have written about the use of informal conversations in ethnography – as part of participant observation – we are advocating that these conversations have an application beyond ethnography and can be used in a variety of qualitative methodologies that occur in everyday settings where talking is involved. Few scholars have written about them, they very rarely feature in lists of key words in academic articles, and, we maintain, the majority of researchers tend to rely on data from more structured interviews, which has generally been audio (and sometimes, video) recorded.

Although, as we stress above, like any method employed in qualitative fieldwork, using informal conversations has advantages and disadvantages, nonetheless we believe that the benefits far outweigh the possible limitations. We argue that these conversations generally create a greater ease of communication (although this is dependent of course on the characters and personalities of the individuals involved) and have the potential to produce more realistic or naturalistic data with less performativity from both interviewer and interviewee. Essentially, they can replace the ‘me’ and ‘you’ with ‘we’ – befitting in-depth communication, and, although Abell et al. (2006) found that this was always the case in their analysis of power dynamics during interviews, we maintain that conversations are also less artificial where people (or participants) are not distracted by the researcher ceremoniously switching a recording device on and off to signal the beginning and end of a bounded and formal interview.

We begin the paper by briefly looking at their historical and methodological origins and previous use in social research. We then use on our own work as examples to show how we have used informal conversations in our research, drawing especially on the second author’s (2020) doctoral research, which was an ethnographic exploration of knife-carrying in the lives of young Black, Asian and minority ethnic (BAME) men on a London housing estate. Given that the young men were very wary of being recorded, using more formal or structured interviews that were recorded needed to be used sparingly. Therefore, most of the data emerged through conversations on the street, in youth clubs, and, occasionally, in the young men’s homes. We present one extract from the first author’s research and three from the second author’s, which we interrogate and use to raise a number of, mainly ethical and methodological, issues and debates: although we believe ethnographers will find this discussion useful, the themes we have chosen highlight some of the potential issues and difficulties that may arise when using this method and are particularly aimed at qualitative researchers using other approaches and methodologies. In the final section, we discuss the main advantages and disadvantages of using this method, and what we consider to be the main differences between informal and more structured recorded conversations, including the status and validity of data produced.

We posit that there are actually two distinct types of informal conversation that can be used as data: those that are overheard or observed conversations that occur during researchers’ observations during fieldwork – either structured or unstructured – (e.g. during a classroom observation, sitting in a staff room, or as part of an observation at an event such as a football tournament); and those shared or participatory conversations that involve an interactive dialogue between the researcher and another person or other people. We deliberately used the word ‘people’ here, rather than ‘participant’, and we will discuss this point further below. We also make a distinction between observed conversations that arise during fieldwork, where people are aware of the researcher’s presence (to a greater or lesser extent), against more covert approaches where the researcher is surreptitiously listening into conversations without people’s knowledge or awareness. Although we do not wish to diminish the role of covert research (and there are plenty of good examples where this methodology is justified), we do not work in this way. We feel that there are ethical problems with ‘listened-into’ conversations, and we are not advocating this approach in this paper.

The Origin and (Under) Use of Informal Conversations in Social Research

Initiating a conversation with a person or group of people has been recognised as an integral element of qualitative research for well over 100 years. Conversations were regarded as a method in their own right, and a central element of social inquiry in several of the early ‘methods’ textbooks of the 1920s and 1930s in the UK (Webb & Webb, 1932) and US (Palmer, 1928). Informal, or unstructured, conversations formed the basis of many early ‘classic’ ethnographies from anthropologists such as Margaret Mead (1928) and Bronislaw Malinowsky (1922), and they became a constituent source of data production in numerous studies by sociologists from the Chicago School in the early 20th century, with ethnographies of settings such as cities and communities (Park et al, 1925), slums (Wirth, 1927), homeless people (Anderson, 1923), life histories (Thomas & Znaniecki, 1918), institutions like the family (Mowrer, 1932) and recreational pursuits and urban life styles (Cressey, 1932). A review of the contemporary qualitative research literature over the last 10 years shows that informal conversations are still used by researchers today (e.g. see recent work from Angotti & Sennott, 2014; Arrazola & Bozalongo, 2014; Berg & Sigona, 2013; Densley, 2013; Korobov, 2018; Thomson, 2018).

While many researchers (e.g. see Bernard, 2011; Kawulich, 2005; Merriam, 1998) have written about the use of informal conversations in ethnography, as part of participant observation, we are advocating that conversations have an application beyond ethnography and can be used in any more general qualitative exploration that occurs in natural, every day, settings where talking is involved: for example, in action research, case studies, ethnomethodology, narrative methods, phenomenology, social anthropology and so on, although a specific research approach does not always have to be identified and deemed to be given a name. An important point to make is that we are not suggesting that using informal conversations has to be the sole, or even the dominant, method of data creation; rather than replacing formal interviews they can be, and frequently are, an additional and complementary source of data to supplement or enhance data produced by more structured, or formal, methods, and this is certainly the case in our own research.

The term ‘informal conversations’ has a number of synonyms: Bernard (2011) has referred to them as ‘natural conversations’, and Patton (2002) as ‘unstructured’ interviews. However, we regard them being different from unstructured interviews, which, however loosely organised, are usually pre-arranged by both parties in terms of time and place, and both parties often know the (rough) area they are going to be talking about.

During his ethnographic exploration of an English secondary school in the early 1970s, Robert Burgess used the expression ‘conversations with a purpose’ (1988; p. 153), although the term can actually be found in the ‘methods’ textbook, as mentioned above, by Sydney and Beatrice Webb in 1932. The name, though, is particularly apposite because, although these linguistic exchanges bear a closer resemblance to the conversations that people engage in during their everyday interactions than formal interviews, we should not be fooled into thinking that there are not asymmetrical power relations in play and that the researcher will usually have an underlying agenda, an ulterior intention and want to control the proceedings. Perhaps they could be renamed as ‘conversations with a motive’? (Swain & Spire, 2020).

Some researchers, such as Hammersley & Atkinson (2007), contend that these dialogues are still a form of interview, albeit an informal one. It is also important to remember that qualitative researchers are, at times, social actors who are performers of Goffman’s (1959) ‘impression management’, whereby they use mechanisms to try and influence others’ impressions of themselves, including, occasionally, techniques of deception (Bernard, 2011). Sometimes, the person is unaware that an interview is taking place, least of all being interviewed. While interviewees might think they are making comments on, seemingly (to them), every day and ordinary issues, the researcher is focused on, and probing for, the elements of talk that answer their research questions (Murphy & Dingwall, 2007). This is an ethical issue that we discuss below.

The Use of Informal Conversations in our Own Work

The first author has used many methods to conduct qualitative research in a range of educational settings for 30 years. We begin this section with a brief look at one example of the first author’s use of informal conversations in his ethnographic research in a school, before using the second author’s recent doctoral study (also an ethnography) to analyse their use in greater depth. We pick up ethical, methodological (and practical) points arising from the exacts in the discussion section.

A lot of Swain’s time as a researcher is spent in schools and colleges, where he observes classes and, sometimes, interviews teachers or tutors. Often, either before or after the class, he will talk with the teacher/tutor about what he is about to see or has just seen: these participatory conversations may take place in the classroom, staff room, canteen or even the corridor, and they are not considered to be formal interviews and not recorded. Their role and purpose are to gather background information, perhaps to find out an opinion, or to deepen understanding, and they will rarely appear in the final text of the project report or research. However, sometimes they provide incipient ideas for future themes to pursue, and therefore have the potential to make a vital contribution to the research study.

However, there are many other times during fieldwork when Swain uses other participatory conversations and observed conversations to generate data that appear in the final text.

The School Classroom

The exchange below happened, some time ago, during an observation of a classroom and the first author wanted to capture the (unequal) power relations between the teacher, Mr. Hughes, and one of his pupils, Richard [1]. In his fieldnotes, he wrote:

The calm/purposeful classroom environment was backed up by the underlying threat embodied in Mr. Hughes and his strict discipline. Many pupils told me that they were a little bit frightened of Mr. Hughes, especially when he shouted at another pupil in the class, and this included myself. However, in the time that I spent in his class, I only saw this happen on a very few occasions. It was noticeable, though, that pupils were given little chance to put their point of view.

The following text appeared in his PhD thesis.

Fieldnotes: Mr. Hughes’s classroom, pm (05.07.99)

Mr Hughes is giving out some unfinished art work [which is one of the pupils’ favourite activities] for them to continue. He’s holds up some examples of their work in front of the class and it is clear that he is not very happy with it – he says he only likes 3 out of the 12 examples that he holds up. Some of the pupils also have to copy out their plays they have written, and one of them, Richard, tells him that he’s already done it on the computer with Miss Blunt (a teaching assistant).

Mr. Hughes: I don’t want it done on the computer

Richard: But Miss Blunt said we can

Mr. Hughes: [Turns on him, suddenly]: Who is your teacher?

Richard [nods]: You are [says softly and with contrition]

Mr. Hughes: [In a harsh tone] Don’t question me. I tell you what to do; you don’t have to think; I tell you what to do and when to do it.

[Richard looks suitably contrite and shame-faced.]

It is the sentiment and meaning that Swain is trying to relate in this observed conversation, which is used to help make his argument about the authority the teacher embodied, and the unequal power relations between teacher and pupil, and between teacher and teaching assistant. There is no effort to try and capture the words verbatim, or to represent them as such, and they are generally viewed in the spirit of representing ‘something “along the lines” of what was said’ (Fine, 1993; p. 278). However, we can also see that the researcher is also attempting to capture the mood and tone of the speech in a little more detail, adding his own impressions, insights and interpretation of what happened. In other words, he was already beginning to examine the meaning of the participants' words and actions in more analytical depth.

We now turn to the second author’s doctoral research about knife-carrying in the lives of young BAME men. The fieldwork for his research took place in 2019 on a London housing estate, which he called Maxwell, and involved around 50 BAME young men, aged between 18 and 22. His research questions included: (a) What resources and strategies do young men use in constructing their masculine identities? (b) How do young men understand and condone knife-carrying? and (c) What role do informal and formal relationships play in young men’s masculine performances?

King is a white male raised on the Maxwell estate in the 1980-90s. He left for university but returned in 2017 to live nearby. His interest in this area of research stemmed from his work with a youth mentoring project. Maxwell was a dangerous place to visit alone and his primary strategy for participant recruitment was through the youth workers who worked on the estate. In addition to supporting recruitment, they offered guidance on how to approach the young men for an interview.

Although the methods used in the fieldwork included community maps, drawn by the young men during focus groups, and photographs they took of unsafe places in and around the area, the main methods of data production were observations and interviews. The great majority of the young men were very reluctant to be recorded during interview; this was mainly because they were worried about the consequences of being seem ‘colluding’ with a white, university researcher, someone who also resembled an outside authoritarian figure (such as from the police, social services, etc.), and this caused them to lack both trust and confidence. And so, King only used six semi-structured paired interviews and around 100 unstructured observations and informal conversations with youth workers and the young men, whom he engaged with frequently over nine months. These conversations with the young men were not always deliberately sought out, or engineered, but when they were judged as contributing data that helped answer the research questions they were recorded as field notes as soon as practicable. This happened on 16 occasions and provided data based on young men’s own words in a naturalistic setting.

We present three extracts of data below: one from an observed informal conversation at a pre-arranged football tournament with a key informant, a young man called Bankz, and a participatory informal conversation inside a flat with a mother of another main research participant from Maxwell. The final extract is again with Bankz, but this time comes from a recorded interview, which allows us to compare the two forms of data generation.

The Football Tournament

The following observation includes some overheard observed conversations, which are integrated into the observation. It was written up in King’s fieldnotes and appears in his doctoral thesis, after he had added further analytical commentary. One of the youth workers had suggested that he attend a football competition, organised by the youth club, as he could expect to observe some of the relations between one of the leading members of the elite group on the estate, Bankz, and youth workers, and, principally, between Bankz and his coterie of followers (or ‘hangers-on’).

Once, at a Youth Club football tournament, Bankz came with his usual entourage, clearly not dressed to participate. When one youth worker asked Bankz what he was doing there, his curt response was ‘watching, it’s a free country plus I don’t see no tickets sold.’ I then observed Bankz regale his group with stories of his football prowess and how he ‘could’ve turned pro[fessional], but mans heart wasn’t in it.’ These descriptions of his capabilities, while having a degree of legitimacy (Bankz was a competent but ill-disciplined footballer according to youth workers), were met with peers’ nods of approval and affirmation. At the same tournament, Bankz would pass comment on others playing, saying ‘Ah this guy’s shit, he can’t play. He should’ve passed it, mans greedy. If I had my boots, I’d show them.’ Invariably, Bankz uttered these comments between puffs of his cigarette or joint – a frequent fixture in his hand with his group gazing on, almost adoringly.

We do not feel that this is an example of a covert, ‘listened-into’ conversation. At this stage in the fieldwork, a few young men had previously seen King around the estate. However, on this occasion, although the youth worker knew that he was there in his capacity of researcher, the other young men, including Bankz, probably regarded this ‘stranger’ as the youth workers’ friend, or, perhaps, a fellow youth worker. We admit, though, that there was an element of deception, with the researcher listening into conversations and observing reactions and behaviours without any prior or post permission. However, no-one seems to have been concerned and his presence appears to have had little effect on their behaviour or the content of Bankz’s talk, which was as King intended to be. It would also have not been practicable (and seemed incongruous) for the youth worker to suddenly stand up and announce the researcher’s presence and explain his role.

The style of writing is typical of that found in many studies using an ethnographic methodology (Atkinson et al., 2007), and the observation and observed conversations enable the researcher to capture a specific event or occasion with greater verisimilitude in a naturalistic setting. He has scribbled down some of Bankz’s comments in his fieldnotes to delineate some of the strategies (e.g. boasting about his athletic accomplishments) that Bankz was employing to give him elevated status amongst his peers.

Visit to a Flat on the Estate

Whereas the example above contains a series of short interjections or quotations, which are relatively easy to remember, the next example below comes from one of the young men’s mothers, Amina. Classic ethnographic research involves participant observation with the researcher visiting and generating data from several situations or contexts. The second author had visited her flat with a youth worker to meet one of the young men, Azeez, who was also a main contributor to his research. While waiting for him, Amina told King some of the problems she had been experiencing with her son, which he recorded in his fieldnotes when he returned home that night. Although this comes from a participatory conversation, King does not include his own injections as part of the dialogue as they are not relevant. Although Amina knew that he was a researcher, she only had a partial appreciation of his intention and study subject matter, beyond that it involved the behaviours and practices of the young men on Maxwell, including Azeez. No consent form was proffered by the researcher to this ‘participant’. These issues regarding levels of consent, and the question of when does a person become a participant? will be picked up in the discussion.

He’s [Azeez] always had challenging behaviours from a young age, but it’s only in the last year that things started to get a lot worse. He was in and out of school. Whenever we found a new school for him, something would happen, and he’d be excluded again. When he wasn’t in school he had nothing else to do. So, he was going out, causing trouble and getting arrested. He was bored and had too much energy. It was when I started seeing him in different clothes, with new phones and gadgets. I found stuff like small knives in his room. It didn’t matter what I said to him, the pull of his friends on the street, I guess was stronger than anything I could do. I knew he was in danger, so when he left once I had to call the police and got him brought home.

Although it is difficult to claim the accuracy of this quotation, King’s aim is to recreate the meanings of Amina’s words, her wearied concerns about the nature of her relationship with her son, the context and the occasion with as much authenticity as possible. As with Swain’s classroom observation, it is the meaning and tone in the conversation that King is trying to convey, and there is no attempt to claim that these were Amina’s exact words. Although the researcher could have summarised the meaning of what Amina told him in a few lines, we feel that this strategy takes the reader more directly into the home setting and hearing Amina’s difficulties and frustrations in her own words (or as many of them as possible) makes the occasion seem more alive, natural and more real.

Confrontation with a Rival Gang

The final extract comes from a more formal interview with Bankz and a friend, which King audio recorded. During part of the interview, he asked Bankz about his older brother, Mo, who, through his connections to organised gangs, was someone that the young men would go to for protection and counsel, although, in many instances, they would go to Bankz first as a gatekeeper. Bankz thus held a role as a conduit, a position he was asked about:

BK: You’re close to your brother, Mo, aren’t you?

Bankz: Yeah … he kinda looks out for me.

BK: In what way? Does he keep you safe?

Bankz: [laughing] I don’t need him to keep me safe, mans can look after myself.

BK: He [Mo] seems to look out for a lot of people around here?

Bankz: Yeah, I heard that. Any beef [assailant] needs squashing then Mo’s on point. He’s got connections, you know what I mean.

BK: Would you say that Mo makes Maxwell a safer place for lots of people?

Bankz: Kinda, I suppose. Mans come to me first though, innit. They speak to me before they speak to him.

BK: But doesn’t that kinda make things unsafe for you?

Bankz: How? Cos I’m Mo’s brother. Nah, fam. Everyone’s tight with me. They look after mans cos they know that they can’t fuck wit’ me. I just gotta tell Mo, innit.

Similar to the informal conversations, the researcher not interested in tonal nuances or emphases, or the pauses; he is concerned with the way stories are told and how meanings are made, and his main intention is to report the words used as reliably as possible. The recording allows the researcher to generate data that, because of language issues (e.g. any beef needs squashing), would have been very difficult to remember and reproduce in fieldnotes, or as a more polished piece of text in a thesis or academic article. As we can see, Bankz’s vernacular was very dense and ‘of the street’, and at times, King needed youth workers to translate some of his speech. We can also appreciate how difficult it would have been for him to note and replicate this exchange accurately and capture Bankz’s specialized way of speaking.

Although the interview was audio-recorded, we also need to bear in mind that even transcriptions from a recorded interview will involve principles of selection, and that they are essentially theoretical in nature and a representative process (Davidson, 2009), rather than 100 percent accurate. When interviews are recorded there is a greater tendency for interviewees to become more cautious but also be more performative, providing responses and answers that they anticipate the interviewer might like to hear, and our interpretation is that there is a lot of performativity going on here. However, this does not necessarily always have to be the case and while the interview is essentially an artifice, it can also appear naturalistic, and, of course, as in the case with Bankz’s interview, produce equally captivating and revealing material as from a spontaneous conversation. What the recording also allows for is for the researcher to present longer exchanges of data between two or more people, and to ensure that there is a greater degree of accuracy in the words used, especially again, with Bankz’s idiosyncratic style of talking.

Discussion

We use the above four extracts of data to further discuss several issues. We organise our discussion under ethical and methodological issues, although they also contain some corresponding practical issues. This approach is more of an analytical device and some of these issues blur and overlap.

Ethical Issues

The general consensus amongst a number of scholars (Canella & Lincoln, 2011; Roth & von Unger, 2018) is that, rather than being viewed as a set of rules that can be added in a top-down fashion from the outside onto previously predicted behaviour, or a specific act, ethics is present ‘in the mundane conduct of everyday life’ (Roth & von Unger, 2018; p. 33) throughout the research process and needs to be continuously and reflexively evaluated and achieved. The ethical considerations arising from the extracts are largely about consent, which, we contend, is not an absolute, nor binary concept, but rather is malleable and contingent. The questions we to pose are: (i) when does a person need to give their consent? and (ii) how much does a person have to be told, and know, about the research project?

Before the recorded interviews King gave the young men, like Bankz, information sheets (which he talked through with them) describing their involvement and asked them to sign consent forms. Information sheets cited the reason for their selection and assurances of their anonymity (through pseudonyms) and the offer for them to leave the research without cause or prejudice. However, the situation is more complex when researchers use informal conversations to generate data. The second author’s overarching aim was uncovering the young men’s views and stories, and to deepen his understanding of their lives and behaviours on Maxwell in the most natural way. Therefore, obtaining informed consent before, during or after informal conversations in everyday, natural, settings, and particularly if it was in the form of a written document, appeared to be obtrusive and impractical (Akesson et al., 2018; Swain & Spire, 2020). While the youth workers had a much deeper knowledge of the study’s aims, the second author avoided giving full details of the research project and did not seek to obtain informed consent for informal conversations with the young men. King was introduced to them by youth workers, who told potential informants that he was interested in finding out more about how they lived their lives on Maxwell. While the minutiae of the study were not initially shared, the researcher did not actively withhold details, and as researcher-participant relations developed he became more open and expansive about the research aims. The critical dynamic was trust and while it is true to say that many of the young men never fully understood the purpose and motivation of the research, it is also correct to say that they did not want to. Many were happy to have someone to talk to, ‘show off’ to, or perform to. Perhaps the most important point was that the researcher was careful to make sure that he preserved the young men’s confidentiality and anonymity outside of the community, except in the case of any safeguarding or welfare concerns.

In the case of Swain’s example, although Mr. Hughes was aware that Swain was a researcher, he knew little beyond the fact that he was interested in pupil identities and what it was like to be a boy and girl at school. However, Mr. Hughes had signed a consent form when he was formally interviewed and had given permission to be observed in his classroom. Understanding is always partial, which, we maintain, should be seen as being along a continuum from full to none, and therefore we argue that this brief outline about the study was enough at this particular time and in this particular context. During his fieldwork, Mr. Hughes’s class was observed by the researcher on several occasions, and he was considered an important participant, if not a main one (Swain was essentially interested in the views of the boys’ themselves). In King’s research, Amina was a ‘peripheral’ (Delamont & Atkinson, 2018; p. 126) or transient ‘participant’ at best. King did not know if he would use her data in his thesis, and it would have again seemed peculiar, if not bizarre, to her if he had asked her to sign a consent form, just in case he was going to use her words at some future point. Researching in natural everyday settings does not work like this, and this is where ethical concerns overlap with practical considerations; moreover, and crucially, Amina was anonymised, as was the place where she lived.

Bankz was a main participant in King’s study. Like Mr. Hughes, he had been interviewed and so knew at least something of what the study was about, even though he almost certainly had a very imperfect understanding of the research’s intent. The majority of the other young men on Maxwell knew very little about the research but, again, the same issues above apply: it would seem strange to them to be offered a consent form which they could only sign if they knew what the study was about, and what their role in it was – the latter, even King did not know at the time. Is it realistic or practical to suggest that he should have handed out multiple consent forms to Bankz’s followers at the football tournament? The event was, essentially, a casual ‘hang-out’, it would have ruined the dynamic of informality and made King seem ‘uncool’. He did not know at the time whether a particular young man, or a group of young men, would be activated to the status of ‘participant’ because their words were going to be used in a published work, a point we pick up in the following section.

Methodological Issues

The methodological questions we ask are: (i) how does a researcher react to chance or surprising moments that may not have been thought about, or covered, in a pre-prepared ethical review; (ii) when does a ‘person’ become a ‘participant’ in a research study, and (iii) what is the status of the data captured and reported from informal conversations as opposed to recorded interviews?

The methodology in our work is ‘bottom up’ and inductive; this approach views the research process as an active enterprise of knowledge construction and is characterised by a flexible research strategy. There are, therefore, likely to be more moments that cannot be pre-anticipated or predicted and so are impossible to be covered in an ethics application to an Ethical Review Board, such as the one at the university where we work. Researchers and scholars such as Denzin (2010); Murphy & Dingwall (2007), and van de Hoonaard et al. (2016) contend that the need to seek approval before commencing a study are making many forms of inductive inquiry, such as ethnography, much more difficult and, in some cases, impossible to carry out. When carrying out fieldwork in naturalistic settings researchers need to make decisions in-the-moment and in-situ, about whether to take opportunistic advantage of a specific event or conversation when they present themselves, which Guillemin and Gillam (2004; p. 265) refer to the ‘micro-ethics’ of the fieldwork Researchers often do not have the luxury of being able to take their time to consider all the issues (think back to the interview scenario in the introduction or King’s attendance at the football tournament, or when he was introduced to Amina in her flat). A number of other researchers have written about these: Fujii (2015) calls these unplanned opportunities that ‘pop up’ outside the boundaries and constraints of formal fieldwork, ‘accidental moments’ (p. 526) or ‘revelatory moments’ (p. 527), while Pinksky (2015) refers to them as ‘incidental ethnographic encounters’ (p. 281). Having said this, King had held meetings with youth workers before the fieldwork began and was able to ‘problematise’ the young men’s involvement – what topics they would likely discuss and who to avoid talking to, for example. This approach enabled a degree of intelligence gathering before conversations started.

Although the incident recorded in Mr. Hughes’s classroom between teacher and pupil was not an accidental moment, and the observation was pre-planned and relatively structured, Swain made an in-the-moment decision that the interaction between Mr. Hughes and Richard provided an opportunistic moment to explore one theme, that of power relations. Swain did not take notes of every conversation he heard – only those that were deemed, in-the-moment, as containing information which may, or could, have contributed to answering his research questions. And, of course, only a fraction of the fieldnotes make it into a thesis or published work. As with the classroom, the football tournament for King was a pre-arranged occasion, a chance to see the young men in a specific social setting; he did not know that Bankz was attending but when he did this again provided an opportunistic moment that the researcher judged might help him answer one of his research questions (about relationships and resources used to gain status). Conversely, Amina’s flat was an accidental moment (King had gone there to meet Azeez); however, when he was introduced to Azeez’s mother, and she began to talk about knives, he made an impromptu decision to turn this into an opportunistic moment, engaging her in conversation, which he then wrote up in his fieldnotes.

Earlier on, when we wrote about the second author’s visit to Amina’s flat, we enclosed the word participant in quotation marks. We posit that a person becomes a ‘participant’ in a study as soon as their data are recorded as fieldnotes; if Amina’s words had not been written up as they were not deemed relevant, we are happy to refer to her a ‘person’. However, she certainly exerted a significant influence on one of the main informants, her son. She was also closely aligned to one of the research questions, about relationships, so would have featured in King’s thinking, albeit implicitly. In short, she was an agent of influence in the study but not a main player. As we have mentioned above, we believe that there are degrees, or layers, of participation along a continuum from ‘full’, ‘central’ or ‘primary’ participant to ‘peripheral’ participant and then ‘non-participant’ (Delamont & Atkinson, 2018; p. 126). As we written above, Amina was categorised as ‘peripheral’.

Another pertinent question is: what is the status of data generated from informal conversations against more formally captured data from interviews? Although the latter can appear to be a more privileged and exclusive source of knowledge (Pinsky, 2015), we maintain that, although these data are different, the data produced through both the observed and participatory informal conversations is as valid (in the sense that the results correspond to real properties and events in the social world), and should have the same epistemological status, as data obtained through more formal interviews. Both informal and formal conversations are a co-construction between researcher and participant, both sources of data are subject to same caveats and limitations in the sense that people can be evasive, mis-remember, mis-lead and lie (Douglas, 1976). The academic community has to accept the honesty and the integrity of the researcher on a basis of trust. The intention is to capture, at the very least, the gist of what the conversation is/was about. It is not as accurate as data represented from a recorded conversation, but neither is it an invention. Of course, it helps if the conversation is written up as close to the event as possible. (However, in principle, we are not against presenting excerpts that are recalled from memory years later, as long as this time lapse is acknowledged, so allowing the reader to take this into account and make their own judgement about the accuracy and validity.) It is also important that when conversational data from field notes is used in the main body of text it needs to be acknowledged (including the date and time), whether this comes from after either observing, or having been part of, a conversation. In the second author’s thesis, he distinguished between Recorded Interview (RI) and Researcher notes from informal conversation (RNIC). As far as the process of analysis is concerned, the main distinction between a transcript from a more formal and recorded encounter and a record from an informal conversation is, as we have acknowledged, one of accuracy, but the process of coding units of meaning to produce themes is the same.

Although using informal conversations has many advantages, we acknowledge that one of the main caveats is memory. Memories are famously unreliable, prone to contamination and degradation, and no one would pretend that any researcher would be able to write down a reasonably lengthy (or even a short) exchange verbatim, such as the one with Bankz about his brother. The general intention is usually to represent a point of view or line of argument, rather than the exact wording but there can be a danger that humans have a natural tendency to misremember what has been said (Lampinen et al., 1997). The researcher will bring their own expectations and assumptions into the conversations and reflexivity is needed to reflect on their own biases (Symon & Cassell, 2012). They also may add things which were not said or forget things that were. Indeed, sometimes, when reading a transcript of a recorded interview the researcher will be surprised by something that was said as they had forgotten it or be surprised because something they thought they remembered being said was in fact not there.

Conclusion

The aim and purpose of the article has been to promote a greater use of informal conversations in qualitative research and show how they can be used to aid data generation in naturalistic settings to enhance findings. Although we believe the article will be of interest and use to ethnographers, the targeted audience is researchers working with more general qualitative approaches and methodologies, who may be less familiar with this method. Although informal conversations have a long tradition, we contend that they are an under-used and rather unsung method in qualitative, social research and can complement and add to more formal types of data created through interview.

We are not arguing that informal conversations invariably produce better, in the sense of more authentic or richer, data in less artificial contexts than formal interviews, but, in our experience, they can, and frequently do so; there is often less performativity, and so they may get nearer to the reality of individuals' experiences, values and perceptions. They are also a very flexible method and open up a greater range of possibilities for data generation. Firstly, they have the advantage that they can take place almost anywhere (e.g. walking down a school corridor, walking across an army parade ground), and sometimes they can be used to generate ideas to be pursued in more formal methods (e.g. interviews and more structured observations), while at other times extracts from the conversations can be incorporated into, and presented in, the findings; secondly, extraneous background noise is not such an important consideration as when a conversation is being recorded (e.g. in a noisy café or a raucous school canteen). Sometimes informal conversations are not only the best way, they are the only way to generate data.

Although informal conversations can be a means of reducing the imbalance of authority between the investigator and those being studied, we acknowledge that they are different from what we may call ‘everyday conversations’ with a friend or stranger. Here, people are less likely to have ulterior motives or be in search of answers to specific themes or questions. They are infused with power relations, the agenda is usually set and controlled by the researcher, and so, as we have suggested, they may be referred to a ‘conversations with a motive’. There may, also, be issues with trying to remember, and faithfully reporting, long exchanges of conversation and of the researcher unconsciously mis-remembering particular words or phrases.

We believe that we have made new conceptual and methodological distinctions between accidental moments and opportunistic moments in researchers’ observations, and between observed conversations that occur during researcher observations and communal or participatory conversations between two or more people. We hope these new terms or phrases will become helpful to future researchers and become part of the vocabulary in qualitative research. Although observed conversations documented in fieldnotes may have a longer-standing tradition in social research, in our experience these are rarely included and interrogated as a source of data in academic articles.

We are not advocating obtaining data from ‘listened-into’ conversations where ethical issues are harder to defend. This is not to say that covert research is never justified, but rather it is a separate issue, and we do not work in this way. Our perspective on ethical practice resembles Whiteman’s (2012) embedded approach to research where the researcher is often required to make ethical and methodological decisions in-the-moment, based on situated and sometimes unpredictable, contexts, rather than strictly adhering to a set of pre-determined ethical principles. One of the repercussions of conducting inductive research in this way is that it complicates the process of obtaining prior informed consent. King’s work necessitated the use of informal conversations as very few of the young men from Maxwell were willing to be interviewed and formally recorded. It may have been necessary and important for youth workers to be given information about the study, and for them to give their consent to be involved, but it would have been impracticable to ask most of the young men to sign a written consent form and to provide information about the study (giving different amounts of information) in natural settings such as a football tournament or in someone’s flat. This does not mean that the young men in King’s study were placed under any undue burden or come to any harm, they were still accorded respect and dignity, and their names and residence were suitably anonymised.

We do not wish to gloss over these issues and recognise that some of them are genuinely problematic for researchers, and we recognise that in this changing landscape some research approaches may step outside the bounds of both methodological, and also formal ethic procedures, including the rather generalised and nebulous guidance from the various ethical frameworks, we therefore suggest that researchers may need to invent and incorporate new strategies to overcome some of the issues. One new and practicable suggestion may be to ensure that there are always gatekeepers involved in the research – like the youth workers mentioned in this paper – who keep the researcher accountable; another way would be to create a pool, or cohort, of researchers working in the same tradition to which the researcher can present their research dilemmas on a regular basis for discussion and gain ‘group’ confirmation. They can also hold each other to account.

We began this article by pointing out that many situations in qualitative research (including epistemological approaches, methods and methodologies) have advantages and disadvantages, but we hope that the arguments we have advanced above have highlighted the benefits of using informal conversations as a major method of data generation, and that more qualitative researchers will consider using them in their future fieldwork.