Researching Individuals’ Media Repertoires: Challenges of Qualitative Interviews on Cross-Media Practices

Researching individuals’ media repertoires is challenged by the problem that respondents are often unaware of their media use because it is part of their daily routines. In an exploratory study, we consider different interviewing strategies that represent varying degrees of explicitness when stating our media-related research interest, different levels of detail in interview questions targeted at individuals’ changing media repertoires and different points in the course of the interview when we state the respective questions. We compare four different strategies based on ten semi-structured interviews with members of the middle class. An interviewing strategy which implicitly states the interest in the respondents’ media repertoires and follows up on this with the help of related enquiries at the end of each sub-theme appears to be the strategy that best suits the purposes. The set stimulus is subtle and, thus, does not dominate the interviewees’ response behaviour; yet it is strong enough to contain the presence of the media topic throughout the interview. Most importantly, this interviewing strategy allows us to capture the respondents’ individual relevance structures with respect to media and media use as part of their daily routines.

awareness makes it difficult to research the role these transforming communications play in the individual's conduct of life.
Therefore, the purpose of this methodical chapter is to identify a qualitative interviewing strategy that meets the requirement of openness-as a core principle of qualitative research-so as to ensure the respondents' freedom to set their own relevance structures, while at the same time maintaining the thematic focus on the interviewees' media repertoire. In the same vein, we aim to reconstruct the relevance individuals attach to their media repertoires and to media change in their conduct of life with respect to disturbances and coping.
To this end, we consider different interviewing strategies representing varying degrees of explicitness when stating our media-related research interest, different levels of detail in interview questions targeted at the individuals' changing media repertoires, and different points in the course of the interview when we state the respective questions. We explored four different strategies in a pretest based on ten semi-structured interviews with members of the middle classnine couples and one single person. The interviews were conducted within the thematic scope of investigating the communicative figurations of German middle-class couples in respect to media-related disturbances and coping practices. We included the life spheres of work, intimate relations, parenthood, long-term asset building and civil society engagement. Our main research question is to analyze how the individuals' media repertoires and the changes thereof shape both the disturbances experienced and the applied coping strategies in the five life spheres. Based on these pilot interviews, we will eventually examine which of these interviewing strategies is most suitable to the research interest of our project.
Even though we developed this approach in order to deal with a specific problem of our research project, the scope of this chapter is a far more general one. We aim to make a methodological contribution to handling the problem of the inherent tension between openness und thematic focus that emerges in qualitative interviews. In particular, we will argue that our findings can deliver profound insights for the ongoing discussion in the so-called 'non-mediacentric' media studies (e.g. Morley 2009;Tosoni and Ridell 2016).
Following this introduction, we first locate our study in the wider realm of qualitative social research and previous discussions concerning the problem of openness in qualitative interviews in methodical standard works, as well as the handling of the phenomenon of media change and media use in empirical studies focusing on the impact of media in individuals' everyday life. Afterwards we describe the different interviewing strategies we considered for researching the role of transforming communications for disturbances and coping of members of the middle classes before presenting our findings and drawing conclusions.

QualItatIve (MedIa) research and the challengIng tensIon Between openness and theMatIc Focus
Our research project is committed to a qualitative research perspective. Thus, it centres on the principle of openness and focuses on the interviewees' subjective relevance system. We translate this orientation into our research questions by asking, firstly, which role does the interviewees' media repertoire play in their conduct of life with respect to disturbances and coping, and secondly, how does the individual experience media change.
An interviewing strategy which explicitly stresses media repertoires and media use would impose the researcher's 'thematic relevance' (Schütz and luckmann 1973: 186ff.). Articulating our research interest in media, we would probably urge the interviewees to talk only about media-related topics. Such a 'mediacentric' approach Stanyer 2014, 2015;Hepp et al. 2015;lunt and livingstone 2016) does not allow us to find out which role media repertoires play in the individuals' general daily conduct of life.
nevertheless, an open approach faces an evident problem. Without a thematic stimulus the interviewees might not talk at all about their media repertoire and if and how they perceive a media change related to their conduct of life. A main problem in this context is the routine character of practices in general. According to Schütz und luckmann, a large part of our daily practices are routinized and based on 'habitual knowledge' (Schütz and luckmann 1973: 107ff.). These more or less 'automatically' conducted practices are primarily a 'means to an end'. Such habitual knowledge has a paradoxical structure of relevance: It is of the greatest relevance and yet of, so to speak, subordinate relevance. It is a determining characteristic of routine that it can be performed without it coming to one's attention, therefore without it becoming thematic in the cores of experience. Routine is continually ready to be grasped without coming into the distinct grasp of consciousness proper. Habitual knowledge is continually, yet marginally relevant. (Schütz and luckmann 1973: 109) In view of these remarks, media use can be regarded as a special routine practice, because media, as the word literally implies, are often used as 'means' and not as 'ends'. For example, using a telephone is usually not an end but a means to get in touch with someone. Therefore, media use as a form of habitual knowledge can be highly relevant for individuals' conduct of life, but not as part of their conscious minds. This implies that an 'activating' thematic stimulus set by the interviewer is necessary in order to reconstruct this habitual knowledge of the interviewees.
The principle of openness is discussed in most of the methodical literature as the core principle of qualitative empirical research. It is the main factor that distinguishes qualitative from quantitative social research, ensuring the respondents' freedom to set their own relevance structures during the interview (e.g. Przyborski and Wohlrab-Sahr 2009: 140;Gläser and laudel 2010: 31;Helfferich 2011: 114). If interview questions reveal the concrete research interest, this directs the interviewees' response behaviour and interviewees are not able to freely state their opinions and experiences, but answer according to the relevance structures set by the interviewer. As a result, the conditions for understanding from an outsider perspective are not given and biases occur, distorting the meaning and interpretation patterns we aim to retrieve from the respondent's interview account (Przyborski and Wohlrab-Sahr 2009: 31).
In order to avoid such biases but, at the same time, to ensure gaining the desired knowledge, scholars suggest different interview forms. Przyborski and Wohlrab-Sahr (2009), for instance, and nohl (2012) advocate narrative interviewing strategies with open ended questions in the beginning and more precise ones in the end. Similarily, Froschauer and lueger (2003) suggest dividing the interview into an initial narrative phase for exploration followed by a second phase of enquiry for clarification (69ff.). According to them, the thematic focus can thus be ensured by enquiring more precisely about the topic of interest as the interview proceeds and when the risk of influencing the respondents' answers is less severe. Although when using these narrative interviewing strategies researchers can adhere to the principle of openness, nohl (2012) argues that even with this interviewing form, habitual knowledge as the aspect we are interested in cannot be retrieved as interviewees cannot reflect this kind of knowledge at all. According to nohl, as an advocate of the Documentary Method, researchers uncover this implicit knowledge in a heuristic analysis of the interview data.
As opposed to nohl (2012), Witzel and Reiter (2012) argue that the respondents are indeed able to reflect upon implicit phenomena. In order to stimulate this reflection, they aim at producing a most natural, everyday conversation (see also Helfferich 2011: 115). Thus, authentic statements are generated that provide the condition for interpretation through others. nevertheless, in spite of this common awareness of the difficulty of balancing openness and thematic guidance, neither Witzel and Reiter (2012), nor nohl (2012), nor Przyborski and Wohlrab-Sahr (2009) describe how to concretely introduce the researcher's thematic interest in the interview-whether during the course of the interview or towards its end. They do not discuss the fact that by stating their research interest in their interview question they might steer the interviewees' response behaviour.
In the same vein, authors of empirical studies on media use neglect the problem of influencing their respondents by introducing their research interest to them. As Röser states, the thematic focus on media use is openly revealed in most of the studies (2016: 491). Hence, there is hardly any problem awareness for the implicit character of media use and, although in communication studies qualitative interviewing is a well-established (loosen 2016) and according to Röser (2016: 490) the most often used method for researching media use, the difficulty of studying media use without imposing one's own relevance structures on the respondent is hardly reflected on in respective empirical studies. Most commonly, researchers dedicate whole sets of questions in their interview guides and specific enquiries to the use of certain media such as communication media (see e.g. Döring and Dietmar 2003;ling 2005;linke 2010;Kirchner 2014). Others do not specify how they introduced the media topic in their interviews, nor give any information on the concrete interview topics. Examples include Clark (2013), who conducted in-depth and focus group interviews as well as observations to research how individuals negotiate the introduction of new media in their home lives, Röser and Peil (2012), who investigated the domestication of the internet with the help of joint partner interviews and representative data on internet use, and Voß (1999), who looked at work and everyday practices of individuals working in autonomous work arrangements. This neglect of precise information about the questions posed during the interview again shows a lack of awareness of the problems around the balance between thematic guidance and the principle of openness.
With regard to implicit knowledge, von Streit (2011) stresses the difficulty in asking individuals about their routine practices. In order to yield this kind of knowledge, she therefore employed an open interview question to start the interview and subsequently asked the interviewees to describe typical days, for instance, in their work life. With the help of these 'experience questions' (Patton 2002: 350), she managed to retrieve implicit knowledge that the interviewees were unlikely to reflect upon otherwise. Through this, she intended the respondents to re-live their daily routines and thus recollect activities that they were usually not aware of, such as media use. However, von Streit is not the first researcher who called attention to the implicit character of media use. Kübler (1987) earlier highlighted this phenomenon. When discussing the enquiry of media use in biographic interviews, he stated that in biographical reconstruction media play only a marginal role. According to him, respondents are not conscious of media use and do not readily recollect media use. Media have become a natural part of their daily routine, but have not reached the deep, biographical dimension of remembrance. Whereas events that have changed, for example, time structures or leisure activities (such as the purchase of a TV) are remembered more easily, slow changes remain unnoticed (1987: 56f.). Therefore, he perceives questions targeted at the share of media reception in the constitution of daily life as the production of a scientific artefact, since this methodical procedure predetermines the meaning media have for the individual (1987: 57). The stimuli effect of, for instance, associations connected with characters and idols prominent from TV could be used by the interviewer to trigger memories of and experiences with media. However, these should only be used to locate or enrich a statement, and not in order to provoke a certain response. Therefore, again, the question of how to precisely introduce the topic of interest and, hence, to maintain the thematic focus remains unclear.
For the international 'non-mediacentric' media studies which emerged in the 1980s, the methodological problem of how to reconstruct media repertoires and relevance frames of media use as integral part of everyday practices is a main focus (Morley and Silverstone 1991;Berker et al. 2006;Morley 2007Morley , 2009Tosoni and Ridell 2016). Paradigmatically, David Morley states in opposition to mediacentric approaches that 'we need to "decentre" the media, in our analytical framework, so as to better understand the ways in which media processes and everyday life are interwoven with each other' (Morley 2007: 200).
Most studies prefer open-ended interviews as an adequate method to enable respondents to articulate their individual relevance frames (Gray 1992: 21;Krajina 2014: 51-57). Most empirical studies following this approach apply an ethnographic oriented 'methodic triangulation', and use different forms of data collection such as observation, media diaries and also qualitative interviews. The latter is attributed a 'pivotal role' (livingstone 2010: 566) to reconstruct how individuals use media in a broader cultural and social context. There are methodological reflections about the problems that the artificial interview situations and the interviewer's interventions have the effect of limiting the possibility that respondents will tell their own 'stories'. Furthermore, there is obviously a more or less explicit awareness of the structural tension between an open-ended interview approach and the media-related research question in mind (Morley and Silverstone 1991: 155). But as far as we see, so far no systematic method has been worked out to address this problem. As a typical example, in his study about individuals' 'everyday encounters with public screens' in cities, Krajina characterizes his interview method as 'in-depth, unstructured conversation, loosely anchored around themes such as routines and interaction with screens' (Krajina 2014: 57). nevertheless, the discussion explicitly poses questions of how to 'unlock' implicit or prereflective practices and knowledge in a phenomenological perspective; implicit media use which is discussed above as a special routine practice is hardly reflected in the ongoing discussion.
To summarize, we can conclude that problem awareness in the methodological literature and the media studies exists with regard to the balancing act between adhering to the principle of openness, on the one hand, and maintaining the thematic focus, on the other. What is less discussed is the implicit nature of media use as a routine practice as was outlined by Kübler in 1987, but which has been neglected subsequently. Although scholars found ways of enquiring about such routine practices with the help of experience questions that encourage respondents to reflect upon their daily routines, in most of the empirical studies researchers do not reflect upon the risk of influencing the interviewees' and elaborated upon sufficiently by the interviewees before. Thus, the relevance of media set by the respondents can be controlled for, while at the same time there is no risk of fully omitting it in cases where the couples do not bring up the subject themselves. However, to the detriment of this approach, questions focusing on the respondents' media repertoires after the first set of questions could influence the relevance interviewees attach to media in the subsequently discussed life spheres, resulting in biased response behaviour. Consequently, the relevance the interviewee attaches to the role of media repertoire is distorted. The second option is to ask for media repertoires only at the end of the interview and after all the different life spheres have been addressed. These questions would encompass all domains and would make the respondent reflect upon the media impact in a comparative way without disturbing the flow of the interview for those who do not mention by themselves media as relevant to their conduct of life. nonetheless, a disadvantage of this strategy could be the detachment of these media-related questions from the respective life spheres, requiring each interviewee to recollect them themselves. A third option is, again, not to ask for the middle-class couples' media repertoires at all. On the one hand, this strategy bears the advantage of leaving it solely to the interviewees to determine the media's relevance to their conduct of life. On the other hand, this approach runs the risk that media repertoires are not mentioned by the respondents at all.
Based on these two variables with three different options each, nine different strategies arise to investigate the role of media repertoires and media change as both sources of disturbances and as strategies of how to cope with them (see Table 15.1). These are neither to state the research interest, nor to ask questions targeting media, or not to state the mediarelated research interest, but to ask such questions either at the end of the entire interview or after each set of questions. Other strategies are to explicitly state the media-related research interest, not to further enquire unless the respondents address the topic themselves, or to again ask corresponding questions at the end of the interview or after each set of questions dealing with one of the identified life spheres. Finally, media change can be stated implicitly as one research interest among others, and then again, questions dealing with this area can be asked not at all, at the end of the interview or after each set of questions.
From these options we decided against the strategies of not stating our interest in media use and media change. nevertheless, asking focused questions on the topic either at the end of the interview or after each set of questions appeared illogical and would have confused the interviewees. In the same vein, we dismissed the reverse strategies of explicitly or implicitly stating media change as the research interest (thus raising awareness) but then not at all or hardly following up on the topic, since we deemed this to be inconsistent. Consequently, four strategies remained. Strategy 1 involves neither stating media change as the research interest, nor following up on the topic by asking questions on it. This represents the most open interview form as it leaves it completely to the respondents whether or not they address the topic of media change and media repertoires. Strategy 2 means explicitly stating the media focus and enquiring about this after each set of questions with regard to the respective life sphere. This represents the least open interview form and has a strong focus on media repertoires. Strategy 3 involves implicitly stating the media interest and posing corresponding questions at the end of the interview. This is a looser form, giving more space for the couples to set their relevance structure. Strategy 4 implicitly states the interest in the respondents' media repertoires and follows up on this with the help of related enquiries after each set of questions on one of the chosen life spheres. This provides both orientation towards changing media repertoires as well as room for the respondents' own relevance structures.
In order to explore these different interview forms, we conducted ten semi-structured interviews between June 2015 and January 2016. One of these interviews was conducted with a single person, whereas the others were joint partner interviews. 1 All interviewees were living in a shared household with their partners in Bremen and its surroundings and were between 34 and 64 years old. Moreover, they were biological parents or caregivers of children between the age of five and 19, of whom at least one attended school. Three of the interviews were conducted in university offices and the others took place in the interviewees' private homes. All interviews were conducted in German and transcribed verbatim.
Having conducted ten interviews, we explored interviewing strategy 1 three times, strategies 2 and 3 twice each, and strategy 4 three times. Strategy 1 was conducted three times owing to practical reasons that followed from the research process, while strategy 4 was conducted three times because it turned out to be the preferred option. Although ten interviews constitute only a small sample, they provide a sufficient basis for our purpose of exploring a topic that researchers have hitherto not explicitly reflected on.
15.4 eMpIrIcal FIndIngs: how to FIgure out a proper IntervIew strategy For a non-MedIacentrIc MedIa study

Interviewing strategy 1 neither stating media-related research interest, nor following up on the topic during the course of the interview
With the first interviewing strategy, we neither mentioned media change as the research interest, nor posed corresponding questions during the course of the interview. Thus, we left it entirely to the respondents to address the topic if they perceived it as relevant for their conduct of life, and introduced the topic of our project as follows: We are interested in how this conduct of life has changed in your personal perception and your experience, 2 but also in practical questions of how you arrange your everyday life against this backdrop, and how you deal with changes and challenges.
Two remarkable cases demonstrate the ambivalent outcome of this interviewing strategy. In the case of a middle-aged couple with both partners working in information technology (IT), media use was mentioned as both disturbances and coping strategies in several of the identified life spheres without being asked. This applied, for instance, to both the domains of work and family: As these quotations illustrate, the media environment and the couple's media repertoire play a crucial role both in their work lives, as well as in how they reconcile family and work. Accordingly, the issue of media use came up naturally without any incentive being required. However, this did not apply to another interview that was conducted employing the same interviewing strategy. Hence, one respondent interviewed individually did not mention media in any respect throughout the entire course of the interview(P4). This was the case, in spite of her being a medical professional working in her surgery in the third generation. Based on this long tradition of running the practice in her family, we can assume that she must have witnessed major media-related changes and developments, at least at her workplace. not having been informed about our interest in media environments and repertoires, and hence, not having been encouraged to reflect upon respective changes, she did not consider these in her account. Therefore, it is likely that a stimulus drawing her attention to media change would have stimulated her to reflect upon the topic and led to a different, more yielding outcome for our purposes-despite the apparently low relevance she attaches to media in her conduct of life. As a consequence, the risk of media change not being mentioned at all seems to be too high when conducting interviews according to this first interviewing strategy.

Interviewing strategy 2 explicitly stating the media-related research interest plus enquiries after each set of questions dealing with one of the life spheres
Conducting the interviews according to the second interviewing strategy, we explicitly stated media change as the main research interest using the following formulation: We are interested in how this conduct of life has changed in your personal perception and your experience, but also in practical questions of how you arrange your everyday life against this backdrop, and how you deal with changes and challenges. We are particularly interested in the role that media change, i.e. increasing mediatization, e.g. in the form of an increase of digital media, such as email, or SMS impacting the individual's everyday life, plays for these changes as well as for dealing with these changes.
Subsequently, we enquired about the respondents' media repertoire after each set of questions dealing with one of the identified life spheres. The interviews conducted in this way showed that highlighting the interest in the respondents' media repertoires and their change over time when introducing the topic led to the respondents strongly focusing on the media aspect, particularly for the first set of questions.
Interviewer: All right, I would like to start with the life sphere of occupation and career, and in the media it's always discussed that there are many changes: they talk about acceleration, more mobility, more flexibility are discussed and I'm first of all interested in how you experience this in your everyday working life?
IP03: Do you want to start, yes? IP04: Well, there is constant accessibility, right? Just through the mobile phone you're always available or at least contactable and reachable. And through email contact, well that's all much faster and, yes, graspable. (P2: 1-11) The example shows that although the initial question was not clearly targeted at media, the respondent focused on media use which had been explicitly stated as the main research interest in the introduction to the interview. Although this choice could also reflect the relevance the respondent attaches to media in this specific domain, it seems unlikely since other interviewees chose a broader start when discussing this life sphere, or talked about further disturbances later on. In the present case, the media focus tended to limit both interviewees' responses to the topic of media, leading them to neglect other factors that might have changed and caused disturbances in their work life, and likely overshadowing their own relevance structure. Although this constraint dissolved or was less prominent with regard to domains discussed later, the narrow concentration initially limited the respondents in their reflections and response behaviour. This initial focus on the explicitly stated research interest in media repertoires is particularly evident, since the couple in this interview did not bring up the media topic in other life spheres again unless specifically asked for it. Thus, enquiring about their media use after each set of questions discussing one life sphere proved to be useful.
Interviewer: How about media communication in the family? Do you use new media there? IP03: Oh, we have a lot of media ((laughing)). IP04: What do we have? IP03: We have a lot of media ((laughing)), the two of us, don't we? Well, so that we can also communicate through email, we'll organize family issues through email. Here another enquiry was necessary to again stimulate reflection upon media use in the life sphere of parenthood and family life, although the partners widely used media to organize their daily life. In spite of this substantial and daily presence, the media focus faded into the background over the course of the interview, showing that these enquiries are beneficial to remind the respondents and assure continuation of this thematic priority.

Interviewing strategy 3 implicitly stating the media-related research interest plus enquiries at the end of the entire interview
Following the third interviewing strategy, we mentioned media change implicitly as the research interest applying the following formulation: We are interested in how this conduct of life has changed in your personal perception and your experience, but also in practical questions of how you arrange your everyday life against this backdrop, and how you deal with changes and challenges, and also which role the increasing mediatization of everyday life might play for this.
We followed up on the topic only at the end of the entire interview. In some cases, this implicit stimulus and the lack of frequent enquiries on the media subject led to the respondents forgetting about this research interest over the course of the interview. neglecting the topic might or might not reflect the relevance the respondents attach to media in their conduct of life, but could also indicate that this focus fades into the background if not taken up occasionally on the part of the interviewer, as discussed above. However, in other cases respondents interviewed with this third interviewing strategy actually did talk about the impact of a changing media repertoire on their conduct of life, referring to the implicit stimulus set in the introduction to the interview topic.
IP05: I think this topic of flexibility, mobility and so on has many qualities that I appreciate, but one real burden is that the possibilities are SO gigantic. You used to have a phone from Telekom, it wasn't called Telekom, it was called the Post.
Usually it worked, but if it didn't, you called somewhere and someone came, tightened a bolt and left again. Or you didn't have a phone, then you had different problems. But nOW. Oh God! Which provider? Which call rate? With a mobile phone there are three million call rates and I realize, basically, this is too much for me. I don't have time for that. And then I realize-and this was an aspect you also mentioned at the beginning: digital world. I believe everything we accelerate: transcribing, typing with the computer and so on. All this always returns as a problem in the form of this flood of possibilities. (P3: 1712-1721) Possibly, the interviewee would have mentioned this irritation in her conduct of life irrespective of the impulse to reflect upon a changing media environment and her own media repertoire. However, there are indicators that the implicitly stated research interest stimulated the consideration of this. The reference to the impulse given in the introduction to the interview implies that this inspired the respondent to make the connection between her reflections and the changing media environment. At the same time, the implicitly set stimulus did not overshadow her own relevance structure, which is supported by the fact that the topic came up in the course of her statement and was not triggered by a direct enquiry. Consequently, the implicit naming of the media-related research interest can produce an adequate balance. With this third interviewing strategy, we further observed that the enquiry about media-related topics only at the end of the interview came as a surprise for the interviewees, who were attuned to the announced five sets of questions, and tended to be tired by the end of the interview. Thus, it is likely that they answered the respective questions in a less elaborate and committed way than previous ones. Moreover, it can be assumed that at the end, the respondents were no longer aware of, or were not able to recollect, all aspects of the previously discussed life spheres and, hence, responded in a way that does not allow for separate conclusions for the different life spheres under study. Consequently, the interviewees might neglect crucial aspects, meaning that the retrospective questions yield less detail.

Interviewing strategy 4 implicitly stating the media-related research interest plus enquiries after each set of questions dealing with one of the life spheres
Applying the fourth interviewing strategy, we stated our interest in the role of changing media environments and repertoires for the respondents' conduct of life only implicitly using the same formulation as in strategy 3 (see above).
We enquired about the respondents' media use after each set of questions dealing with one of the previously identified life spheres. The implicit statement that media was one among other research interests led some respondents to take up the topic in their narrative. Others did not talk about their media use spontaneously. This differing response behaviour implies that the stimulus was not so strong that it overshadowed their own relevance structures, but that it left room for them to discuss individual chances and disturbances. When respondents mentioned media use in the depictions of their everyday life, these statements provided the interviewer with links to further enquiry on the topic, thus enabling natural conversation.
IP17: (…) The new media enable me to do a lot of coordination work of my job myself-independent of space and time. Therefore, I'm more flexible, but therefore I'm also more out and about, and out and about in shorter intervals. I would say there are many more…, or basically it applies to me that the options of doing several things increases, with it the problem of choice increases-what I am doing is rather becoming too much. But the coordination work necessary for doing these things has all become much easier owing to the new media. Interviewer: Could you describe again what are, or which are, these media that play a role for you? (P9: 22-32) However, also for those who did not initially mention media during the interviews, media-related enquiries did not come as a surprise. In such cases, asking for media repertoires was not perceived as unnatural and did not disturb the course of the interview, since these questions were not posed before the topic was raised by the interviewees themselves or at the end of each section of interview questions. Consequently, the media focus did not artificially create supposedly desired responses.
Interviewer: (…) You've mentioned before that sometimes there are business trips. Do you use something like Skype or any… IP19: no. Interviewer: … other media? IP18: no. We're, I, well, ((laughing)), a friend of mine recently said that I'm a media dinosaur.
Interviewer: OK. ((laughing)) IP18: So in the sense of, that actually already distinct, doesn't exist any more. I'd also be so difficult to get hold of and so on. (P10: 856-869) Another reason for the lack of this 'interviewer's desirability' could be the fact that media use was initially mentioned only as one among other research interests. Hence, interviewees might have perceived it as more legitimate and might have felt more comfortable about not being able to extensively talk about media. At the same time, enquiring about media use with regard to every life sphere ensured that respondents who did not attribute high relevance to media did not drop the subject entirely, as was sometimes the case with strategy 3, when such questions were posed only at the end of the interview.
Consequently, this fourth interviewing strategy allowed interviewees to express their individual relevance structures. Illustrative evidence for this claim is the interview account of a couple in which one partner is to a much higher degree involved with media and media use than the other. They readily stated their different weighting of media for their daily routines in the different interview sections. This can be shown in relation to the life sphere of work. Asked about the changes in his everyday work life, the husband, an employee of an internationally operating industrial enterprise, immediately identifies changed media repertoires as a driver for changed working practices: In contrast to her husband, for her, media use is not a crucial part of her everyday work life. She explicitly describes media usage as marginal and of minor relevance ('I use it a little bit'). In comparison with her husband, her media use corresponds with a generally lower subjective attention to media-related topics during the interview.
The reconstruction of the couple's different relevance structures shows that strategy 4 is well balanced between openness and thematic focus: on the one hand, it provides the interviewees with room to express their own relevance structures and, on the other hand, it allows the researcher to carefully examine the role of media use and media change. Thus, it is possible to reconstruct the degree to which mediatization and media change shape the individuals' lifeworld.
After having explored four different interviewing strategies, one of them, namely strategy 4, can be identified as the most suitable for our purposes.
Strategy 1 was successful in one, but not in the other two interviews conducted in this way. Depending on the respondents' involvement with media, without stating the media-related research request at all they attached more or less relevance to the changing media environment and their own media repertoire. However, the risk that media are not mentioned at all is too high for our project targeted at researching mediarelated changing communicative figurations in middle-class couples' conduct of life.
Strategy 2 proved to be ambivalent. While prominently highlighting the media focus led to an initial bias towards media-related disturbances and overshadowed the respondents' own relevance structure, the frequent enquiry about media environment and repertoires ensured recollection of the topic.
Strategy 3 again had advantages and disadvantages. On the one hand, the implicit impetus provided orientation without impairing interviewees' own relevance structures; on the other hand, some interviewees forgot about this focus in the course of the interview. This neglect was further induced by the lack of follow-up questions during the interview. Additionally, the occurrence of the media-related questions only after having discussed all the different life spheres was perceived as tiresome on the part of the respondents, who had been attuned to the announced five sets of questions.
Consequently, strategy 4 appears to be the one that best suits the purposes of our project. The set stimulus is subtle and thus does not dominate the interviewees' response behaviour; yet it is strong enough to contain the presence of the media topic throughout the interview. Most importantly, this interviewing strategy allows the respondents' individual relevance structures with respect to media and media use as part of their daily routines to be captured.

conclusIon
Studying media use in middle-class individuals' conduct of life as routine practice, we were confronted with the difficulty of how to retrieve habitual knowledge through interviewing without imposing our relevance structure onto the respondents. Therefore, we were looking for an interviewing strategy that meets both the requirement of openness as the main principle of qualitative research and of thematic guidance throughout the interview. We explored four different interviewing strategies with the help of ten semi-structured interviews covering different life spheres, namely work, intimate relations, parenthood, longterm asset building as well as civil society engagement, and found one strategy that meets our purposes. By naming media use as one research interest among others and enquiring into it only after the respondents had brought up the topic themselves or at the end of each set of questions, this strategy allows for the interviewees' habitual knowledge to be addressed and at the same time provides room for the individuals' own relevance structures. As stated in the introduction, our findings do not only deal with a specific problem solution for our particular research project but in general can provide significant methodological insights and impulses for qualitative research approaches applying interview methods. First of all, our findings address in a pragmatical perspective a basic question of qualitative research, that is, how to 'unlock' individuals' routine practices by retrospectively asking them about it. This is, as we discussed (see Chap. 2), of particular importance for media studies because of the highly implicit character of media use as a routine practice. Furthermore, we argue that this chapter-representing a media-centred approach-can contribute to the ongoing discussion of 'media-centric' versus 'media-centred' approaches in media and communication research Stanyer 2014, 2015;Hepp et al. 2015;lunt and livingstone 2016). Our findings can provide a systematic interview guide for the examination of relevance structures that media repertoires have in individuals' everyday life from a subjective perspective. By finding the right balance between openended interview questions and thematic focus imposed by the researcher, our approach promotes 'a nuanced and critical grasp of the reciprocity of media and everyday life' (Tosoni andRidell 2016: 1286). Up to now this challenge has been largely delegated to the individual researcher's improvisation skills in concrete interview situations. Finally, the interview strategy we developed here could be especially useful for the examination of cross-media practices because, as our empirical discussion made clear, our approach is able to highlight the relevance and diversity of media use and media technologies in certain life spheres such as work, family or civic engagement as well as in these spheres' interconnections. notes