Only Second Best? Interview Reports as a Method of Documenting Qualitative Interviews

Abstract: Qualitative interviews should be audio recorded and transcribed word by word. This has been recommended in most methodology texts. Relying on interview reports as a means of interview documentation is typically described as a second-best solution if audio recording is not possible. In this contribution, we question this position on the basis of theoretical considerations and research experience.

The advantages and disadvantages of "audio recording + transcription" versus "note-taking + reporting" are systematically compared. Two studies are presented as exemplary cases in which interview reports have been chosen as the most suitable means of interview documentation. In both studies, we dealt with sensitive topics and aimed at reconstructing routines and procedures.

Generally, reports can be adequate when researchers are less interested in the interpretation of what was exactly said and more interested in what interviewees intended to say. In addition, note-taking should be considered if audio recording prevents the participation of particularly interesting interviewees or if it hinders an open conversation. With this contribution, we aim at stimulating further debate about when interview reports are appropriate and how they should be designed.

Key words: interviewing; transcription; recording; minutes; interview reports; telephone interviews; expert interviews; qualitative interview; undocumented migration

2.3 The advantages and disadvantages of documentation methods: Transcription versus interview reports

3.1 Short telephone interview reports in a study on school enrollment for undocumented children

In most methodology texts on qualitative interviews, audio recording of interviews and subsequent transcription (i.e., converting them into text through verbatim transcription) is recommended. For example, LAMNEK (2010, p.325) referred to recording devices in his textbook as indispensable. To some extent, accuracy and comprehensibility are mentioned as reasons why recording and verbatim transcription are necessary (KUCKARTZ, DRESING, RÄDIKER & STEFER, 2008, p.25). [1]

Nonetheless, interview reports²⁾ as a means of interview documentation—based on the notes taken by one or more researchers—are widely used in practice; however, they are often considered highly problematic from a methodological point of view (LIEBOLD & TRINCZEK, 2009, p.41). Methodological discussion is usually limited to short paragraphs in long methodological texts. DAVIDSON (2009, p.36) considered transcriptions as a neglected topic in qualitative research in her literature review, which included more than 80 sources, we can state that this is even more the case for interview reports. [2]

Interview reports are often seen as a second-best solution, only recommended in the case of refusal to record. They should be completed immediately after the interview, and researchers should be aware that these reports are only of secondary quality (LIEBOLD & TRINCZEK, 2009, p.41). Mostly, the discussion of quality requirements for interview reports is limited to short references, for example, that they should be prepared as soon as possible after the interview. Especially in the case of expert interviews, it is assumed that recordings and at least the transcription of relevant passages are necessary (MEUSER & NAGEL, 1991, 2009, p.50, 2013, p.466). [3]

The emphasis on the problem of interview reports is not equally pronounced in all disciplines. Especially in anthropology, documenting and reflecting one's own and others’ observations is an essential part of field research (BREIDENSTEIN, HIRSCHAUER, KALTHOFF & NIESWAND, 2013; LOUBERE, 2017; STRECK, UNTERKOFLER & REINECKE-TERNER, 2013). In this paper, we argue that interview reports for documenting interviews can be the method of choice even beyond ethnographic research, and that they are by no means considered only a second-best solution. [4]

First, we explain our understanding of transcription and interview reports (Section 2.1) and distinguish between informant and respondent interviews (Section 2.2). Then, we identify dimensions that can be used in the research planning of qualitative interviews to help decide between "audio recording and subsequent transcription" or "note-taking and reporting of interviews" (Section 2.3). After that, using the example of two of our studies, we show how these dimensions can be used as a criterion for decision-making (Section 3). Finally, we summarize that the decision for "note-taking and reporting" can also be the first-best solution and conclude that there should be a more intensive methodological discussion about scientific quality standards in interview reporting (Section 4). [5]

Transcription is the rule-based textualization of the spoken word into written language (FUß & KARBACH, 2014, p.15). Transcriptions, in the positivist understanding, are about verbatim reproduction of the utterances of the interviewer/moderator and interviewee (ARBEITSKREIS QUALITATIVE MARKT- UND SOZIALFORSCHUNG, 2007, p.24). Transcriptions can be made more or less comprehensive. For example, in addition to the spoken word (including words or sentences that have not been spoken to the end), phonetic expressions can be transcribed. Furthermore, one can transcribe other audible aspects of the interview situation such as non-verbal utterances (laughing, crying, coughing, smacking), background noise, variations in the sound of speech (volume, intonation), and pauses. Within a research context, researchers agree on transcription rules to strive for a transcription that is as uniform as possible. These rules determine which linguistic occurrences are identified with which signs (FUß & KARBACH, 2014, pp.15-19). The different possibilities of creating transcripts already point to their constructivist character and make clear that an objective representation of a conversation is not possible: "[A] transcript is a text that 're'-presents an event; it is not the event itself. Following this logic, what is represented is data constructed by a researcher for a particular purpose, not just talk written down" (GREEN, FRANQUIZ & DIXON, 1997, p.172). [6]

In this respect, the transcription implies a theoretical, interpretive, selective, and representational process (ASHMORE & REED, 2000, §11; DAVIDSON, 2009, p.37). A transcription is always a selective perspective construction in a new medium of representation (BREUER, 1999, p.252). The focus is no longer on what was "really said" but on the extensive provision of ways of hearing and understanding an interview situation (p.254). [7]

Interview reports, which are also about the textualization of conversations, can also be written in various ways. For instance, during an interview the main statements can be noted directly, and to some extent, even word-for-word (simultaneous log). However, it is equally possible that a report is written based on notes and memories, usually immediately after the event (memory report or log) (ARBEITSKREIS QUALITATIVE MARKT- UND SOZIALFORSCHUNG, 2007, p.24). For transcription, we detail above why it gives a selective view of a conversation situation, whereas this is obvious for interview reports. "Key messages" can only be documented if the question and thus the selection criterion is clear. The choice is also influenced by the note-taking person's mental capacity to listen and remember. [8]

Interview reports can be organized as chronological or summary reports. The former documents an interview over the course of its time as precisely as possible and thus reflects an understanding of an interview situation. In summary reports, the essential contents are outlined in a structured manner. Therefore, text fragments from the notes are organized into coherent units (BREIDENSTEIN et al., 2013, p.100). Any interview report can be done by a single person or in a team of two or more people who complement and correct each other (team report). [9]

Regardless of the form of the report or the number of persons taking notes and reporting, researchers paraphrase which represents a later step of analysis in a transcribed interview. By paraphrasing, we mean a textual summary of a statement in our own words with the aim of preserving the original information, interpretations, and opinions (KAISER, 2014, p.96). In this respect, reporting involves an even more active process than audio recording with subsequent transcription—a process which already belongs to the analysis of the objects under investigation through word choice and sequencing, through emphasis and omission, through the creation of order and coherence (BREIDENSTEIN et al., 2013, p.103). Thus, interview reports have not only a documentary but also a communicative function—content is not only captured but also prepared in such a way that it can be understood by third parties (p.106). In contrast to transcriptions, it is often neither about the reconstruction of what was said nor about the variety of ways of understanding an interview situation, but about a written agreement on a way of understanding. [10]

When we discuss the advantages and disadvantages of documenting interviews by transcription and interview reports comparatively below, the focus is on the main documentation method. We are aware that even in the case of recorded interviews, supplementary notes are often made, e.g., in the form of postscripts commenting on the interview situation (WITZEL, 2000). On the other hand, audio recordings may also be available for interview reports (LOUBERE, 2017, §8). These are more likely to be used to eliminate gaps in memory, uncertainties about individual parts of the interview, or to transcribe individual sentences verbatim and insert them into the report. Thus, the decisive criterion is not whether audio recordings exist but whether research is predominantly continued with transcripts or with interview reports. [11]

Whether a detailed transcription or the recording of an interview appears necessary depends first and foremost on the epistemological interest and the associated role of the interview partner. In ethnological research, statements in interviews are treated both as a resource (as information) and as an object of investigation (as representation) (BREIDENSTEIN et al., 2013, p.83). In the first case, interviewees are understood as informants—as observers for the observer or proxy observers. In the second case, the focus is on the respondents themselves as subjects with their experiences and interpretations. In this respect, a distinction can be made between informant and respondent interviews. Even in a typology of expert interviews, BOGNER and MENZ (2009) distinguished whether the experts are used as a source of information (pp.64-65) or whether additionally the subjective dimension of knowledge and thus the interviewees themselves become the subject of the research. [12]

When informants are interviewed, they may provide different types of information in the interview:³⁾

If informants are additionally interviewed as respondents, the focus is on subjective aspects that go beyond the information given above:

In research with qualitative interviews, respondent interviews are certainly more common than informant interviews. MEY and MRUCK (2010, p.431) explained that qualitative interviews are mainly suitable for generating narratives, arguments, and justifications, but not for querying (factual) knowledge, since a well-constructed questionnaire can do this much better and more reliably. We do not share this view because qualitative interviews are an important and sometimes the only possible source of event- and process-related information. Potentially, this is the reason why there is relatively little methodological literature on interview reports, as these types of qualitative interviews account for only a narrow segment of qualitative research. [15]

2.3 The advantages and disadvantages of documentation methods: Transcription versus interview reports

In the literature, when interview reports and transcription are compared, it is usually assumed that a long interview is documented by a single person. However, short interviews and team reports are also considered as possibilities in the following sections. In team interviews, one person conducts the interview, and another takes notes, as in the study presented in Section 3.2; alternatively, the interviewers take turns asking questions and taking notes (LOUBERE, 2017, §24). [16]

On this basis, we subsequently discuss in which research situations an interview report might be the best method to document oral interview situations in writing.⁴⁾ The following aspects are compared: Time and financial cost, probability of an interview, quality of the interview, and quality of the documentation. [17]

In comparison to transcriptions, interview reports are usually less time-consuming. For transcription, KUCKARTZ et al. (2008, p.29) estimated four to eight times the actual interview duration.⁵⁾ To the best of our knowledge, there are no benchmarks for how long it takes to produce an interview report. In our own experience, at least concerning memory reports, twice the interview time for a single person is realistic—implying significantly less time than for a transcription. In the case of team interview reports, the time needed for the initial note taking, the time for revision and amendment by the person conducting the interview, and the time for the discussion of differences must be taken into account. A team report may require as much time as a transcription. However, it contains an additional step of analysis since the interview is already focused on the research question. According to the experiences of LOUBERE (2017, §25), such a team report can be prepared in one-fifth of the time required for verbatim transcription. [18]

We assume here that the interviewer is usually part of the research team; summary reports require a high level of expertise and knowledge of the research goals, so the researchers themselves usually have to be involved. Transcriptions, on the other hand, can sometimes be created by student assistants or externally.⁶⁾ The financial costs per hour can be lower for transcription than for a qualified summary report. [19]

Overall, it can be assumed that interview reports are often less time-consuming and costly, but depending on the report's form, this does not necessarily have to be the case. If it requires less time, it also means that relevant material for further research is available more quickly. [20]

If it is possible to motivate the desired interview partners to be interviewed, the sample's quality can be considered high. This applies to both quantitative surveys and qualitative interviews. If the documentation method affects the willingness to participate, it will affect the quality of the sample. [21]

According to PATTON (2002, p.381), most people are so familiar with audio recordings that they agree to it, so that taking verbatim notes during an interview has become the exception. WITZEL (2000, §7) also assumed general acceptance. In the case of video recordings, REDLER (2006, p.28) assumed that they may create resistance among the participants. [22]

However, RUBIN and RUBIN (2005) also stated that audio recordings are not appropriate for informal interview situations and sensitive topics: "You certainly do not record a secret informant" (p.101). For successful recordings and transcriptions of sensitive issues, one must first establish trust, which does not always work out, and in the end, only a few interviews result from many contacts. In the 1990s, for example, CYRUS contacted many people several times before he was able to interview individuals without valid residence permits and also tape the interviews (CYRUS & VOGEL, 2006). These interviews have provided rich material for understanding the lives of undocumented immigrants. However, one cannot rule out that interviews with people who were not available for a recorded interview would have added further insights. [23]

Deniability—the ability to deny having said something—can be a condition for saying anything at all. Therefore, RUBIN and RUBIN (2005, p.100) deliberately did not use a recorder in discussions concerning sensitive topics, while they preferred audio recordings for less sensitive topics. In our experience with government employees, the willingness to be interviewed is higher if the interviewees do not have to fear that they will be blamed for an unfortunate wording—be it in public or by superiors.⁷⁾ This concern can also lead someone to refuse an interview and providing only written answers to questions. In our own experience, such written answers usually only contain information on the legal situation, regardless of how the question was asked. [24]

When a person does not agree to a recording, there are two possibilities: Either the interview does not take place, and the researcher tries to recruit another person with similarly valuable insights, or it takes place but with note-taking as a so-called second-best alternative (LIEBOLD & TRINCZEK, 2009, p.41). In this case, producing an interview report based on notes is better than leaving it out, especially if a person appears essential for a research question. This might be because that person has a unique position in a social structure (e.g., the head of an organization to be examined). [25]

If a small number of respondents are to be interviewed from a large number of potentially interesting interviewees, requesting an audio recording can systematically bias the sample. The reason for this is that the people who do not have sensitive information are more willing to be recorded than people with information they consider to be sensitive. [26]

In this section, we assume that a person would agree to both a taped and transcribed interview as well as an interview documented by note-taking and preparing an interview report. In such cases, a recorded and transcribed interview is often recommended in the methodological literature. [27]

The first reason usually provided is that the interviewer can fully concentrate on conducting the interview and observing situational conditions (KVALE & BRINKMANN, 2009, pp.178-179; WITZEL, 2000, §7). Verbatim note taking can distract the researcher in such a way that the interviewee only receives secondary attention (PATTON, 2002, p.381), which could have a negative impact on the willingness to provide information. This criticism refers to a specific interview situation in which one person conducts a long personal conversation and tries to make verbatim notes. If these conditions do not apply, the criticism does not apply either. [28]

In team reports, for example, one person can focus on conducting the interview and the second person on notes. The load of taking notes is lower when the main concern is to write down notes in such a manner that one can reconstruct the information from memory, and only individual words or sentences are to be written down verbatim. In the case of short telephone interviews, by which the interviewer is equipped with headphones to take notes directly on the computer, the distraction is supposedly marginal. Even so, it requires adequately fast typing skills. [29]

The more important question is whether interviewees express themselves differently when the conversation is being written down instead of being recorded. [30]

Several authors of methodological articles estimated the disruptive effect of a recording device during the interview as rather low. FLICK (1991, p.161) already argued that recordings are part of everyday life in many institutions, e.g., for supervision purposes, and are therefore not perceived as disturbing.⁸⁾ MEY and MRUCK (2010, p.431) also attributed this to the small recording devices that are now available. However, FLICK also said that it cannot be ruled out that the statements made by those involved are influenced by recording (1991, p.161). MEY and MRUCK (2010) indicated that the device can be switched off if the insecurity does not subside in the course of the interview. [31]

Based on their experiences, RUBIN and RUBIN (2005, p.100) emphasized that interviewees were influenced in different ways by recordings: some reacted in a shy and hesitant manner, while others spoke important answers explicitly into the microphone to feel assured that their answers would not be misquoted. GLÄSER and LAUDEL (2009, p.157) stated that the most important disadvantage of recording is that interviewees could become self-conscious and withhold information due to the presence of a recording device. [32]

Postscripts also indicate this. They were often used to document statements relevant to the research questions after the recorder was switched off. It is open to debate whether this information would have come up earlier in an interview report based on notes. Especially with members of hierarchical organizations, a recording can induce special caution. Even if anonymity is guaranteed, hierarchically subordinate persons may fear that they can be identified and sanctioned through insider knowledge in the organization. [33]

Government employees do not mainly encounter interviews through supervision but through the media in which thoughtless statements are sometimes used to cause a scandal. They are, therefore, often very anxious to represent their work in such a way that it reflects, for example, the prevailing legal situation. Words are, hence, chosen carefully in recorded interviews, while non-recorded conversations allow them to speak more freely so that the deniability already mentioned in §24 plays a role here: "They can always say 'I never said that' if they feel they need to, and there is no proof that they did" (RUBIN & RUBIN, 2005, p.100). [34]

Even with non-sensitive topics, note taking can have benefits. When stressed interviewees are interviewed on a complex topic—e.g., bureaucratic processes that have changed over time—they sometimes get flustered, make mistakes and correct themselves immediately or later in the interview. Of course, careful preparation should reduce the likelihood that an interview will take place in a stressful manner or with too little time. However, this is not always possible or foreseeable by the interviewers. In such cases, chaotic transcripts are often the result. Nevertheless, the process to be determined may have become so clear to the interviewers that they can summarize it at the end of the interview. In such a case, it is about explicitly intended messages, e.g., about a typical process, which are more easily captured by paraphrasing than by verbatim transcription. [35]

While WITZEL (2000, §8) assumed that recording and transcription allow for an "authentic and precise record of the communication process," this is by now mostly questioned in the methodological literature, which emphasizes the constructivist character of transcriptions (DAVIDSON, 2009). However, the recording and complete transcription of a qualitative interview retains more verbal information than the interview report and has fewer interpretive elements: Even though it is inevitable that transcriptions sometimes contain gaps and misunderstood words, they essentially also allow third parties to understand interviews, specifically if the audio recording can be accessed (KVALE & BRINKMANN, 2009, pp.178-179), and relevant contextual information is added by means of a postscript. Compared to an interview report, an audio recording has the advantage that the statements of fast-talking interviewees can be recorded completely. Furthermore, the passages at the end of an interview, when the ability of the listener to concentrate has decreased, can be completely replayed at a later time (PATTON, 2002, p.380). [36]

When the aim is to capture what was said as authentically as possible, the transcription of an audio recording is superior to the interview report. Complete and accurate simultaneous note-taking is almost impossible today, especially with the decline of stenographic skills. Also, memory reports always remain incomplete. Listeners quickly forget exact phrases. Active listening can not only distort memories but also act as a selective filter, especially if memories do not seem relevant to the topic and purpose of the interview (KVALE & BRINKMANN, 2009, p.179). [37]

The longer the interview, the greater the advantage of transcription as compared to the interview report; in the case of short, e.g., telephone conversations and immediate notes, the loss of content is less than in the case of long conversations. In addition, the distortion and filtering effect with regard to research-relevant content is lower with team reports than with individual reports. [38]

The further interpretation steps associated with interview reports (GLÄSER & LAUDEL, 2009, p.157) should be subjected to additional checks. The process of preparation and evaluation should be carried out in a controlled and systematic way and not be dependent on the respective memory performance of the interviewer or note-taker (LIEBOLD & TRINCZEK, 2009, p.40). We consider this particularly useful when the respondents, as subjects with their own interpretations, are the focus of the analysis. A recording and at least a partial verbatim transcription are indicated in order

"to convincingly demonstrate the validity of their statements on the material. An interpretation must be compelling and comprehensible for the reader, i.e., the readers are involved and guided through the interpretation process; ideally, they should go through the researcher's own cognitive process described in the text" (BERGMANN, 2009, pp.500-501; our translation). [39]

However, if the interview is conducted as an informant interview, this is not so obvious. With clearly defined questions about one's own field of action and observation, recording and transcription harbors the probability that data material will be generated but remain largely unused. A summary report provides an additional option for communicative validation by the interviewees after the interview (WITZEL, 2000, §16): In this case, the researchers write down how they understood the presented aspects. Of course, this is also possible after recording and transcription but it requires an additional step of analysis.⁹⁾ The interviewees are given the opportunity to correct and supplement the understanding presented in the interview report. [40]

If an interviewee has approved recording and no negative impact on the content of the interview is to be expected, the resulting audio files can also be used selectively for interview reports, by listening to individual passages in the case of incomplete notes and memories. If necessary, a recording can be used to insert verbatim quotations in more detail into an interview report or to listen to an interview excerpt in the case of differing understandings in teams. However, the interview report remains the main data material with which further work will be conducted during the research process. [41]

If the merits of interview reports are compared to transcripts in methodological literature, the balance is usually in favor of recording and transcription and against interview reports. GLÄSER and LAUDEL (2009, p.158), for instance, summarized that the probability of a significant loss of information outweighs the risk of a biased (or self-conscious) interview situation and therefore considered audio recording to be indispensable. Undoubtedly, this is true for many research questions. Nonetheless, after a differentiated consideration subsequent to our reflections, we state that this does not apply to all research questions:

However, if interview reports are the mode of choice in certain cases, then the methodological demands on interview reports should also be reflected and discussed. [43]

In this section, we introduce two forms of interview reports that we used in studies. Within this context, we present why the interview report was chosen and what was done to ensure quality, as well as our practical experiences with it. The remarks conclude with a brief summary of appropriate application areas and desirable quality assurance measures for the respective type of interview. [44]

3.1 Short telephone interview reports in a study on school enrollment for undocumented children

3.1.1 Aim and research design of the study on school enrollment opportunities for undocumented children in Germany

We conducted a study on school enrollment to examine the extent to which the right to education is implemented for all children regardless of their residence status, i.e., also for children without valid residence documents (sans-papier, undocumented, irregular migrants)¹⁰⁾ (FUNCK, KARAKAŞOĞLU & VOGEL, 2015). This legal right to schooling is not only required by the International Convention on the Rights of the Child but was also affirmed nationally in 2011 by amending the German Residence Act. Within the framework of this amendment, schools were exempted from the obligation to provide information to the authorities so that even children lacking papers could attend school without the fear that their irregular residential status would be discovered (FUNCK et al., 2015, p.11). At the political level, this was seen as sufficient. However, staff members of Non-Governmental Organizations (NGOs) told us about their experiences with schools unwilling to enroll undocumented children. We wanted to determine whether these were individual cases or whether the non-admission of undocumented children in schools was a widespread problem, which raised a quantitative question. If the latter was the case, we further wanted to ascertain what was preventing school attendance, and this was a qualitative question. [45]

Anybody seeking answers to these questions faces several methodological problems, especially stemming from the sensitivity of the topic, the relative rarity of undocumented children, and the diversity of school types and school policies across the states. Finally, a design was developed for telephone interviewing at 100 public primary schools in 22 cities nationwide¹¹⁾, with a manual containing three simple, closed-ended questions while simultaneously allowing for further explanations, comments, and follow-up questions (see below). For the study, primary schools were selected because they exist nationwide, whereas the diversity of school types at the secondary level in Germany would have made it challenging to identify comparable schools in different states. [46]

The reason for conducting a survey at schools was because NGOs reported to us in advance that undocumented parents call schools or have trusted persons who reach out—often by telephone—to see whether it is possible to enroll their child in school. Therefore, we wanted to conduct a telephone survey to reach persons who are also confronted with comparable inquiries in reality. The target person was the one who had confidence in answering the questions, and this could be the school secretary. However, she could also refer the call to the school administration or to another office.¹²⁾ In that case, the referred person was also contacted and thus became the next target person for the school in question. In some schools, the first phone call was successful. However, in other schools, FUNCK needed to make multiple phone calls due to following up on references to other times or people. [47]

The interviews conducted within the framework of the study were informant interviews (see Section 2.2): Interviewees were questioned because they were expected to provide information about local knowledge as well as routines and orientations for actions. The initial contact was with secretaries or school principals; SchulrätInnen [administrative officers at higher levels] were approached when further referrals were made to them. One can assume that they are not necessarily better informed than secretaries about what practices actually prevail in a city's schools, but they should be better informed about what practices should prevail. [48]

The structure of the manual was as follows: A short standardized introductory sequence was used to make the research institution and its area of interest transparent. The introduction announced three questions about school enrollment in non-ordinary cases that could be answered in three to five minutes. Later, some personal information was requested if this had not already become clear during the course of the interview. [49]

School staff were asked three questions starting generally and becoming more specific. The first question was intended to be easy to answer, so as to motivate cooperation. The questions were:

Thus, the questions were asked in such a way that the interviewees could simply answer yes or no. Nevertheless, it was possible to explain and comment on the answers. The interviewer could then ask further questions. The closed question was meant to be used for a minimum assessment of the problem. The explanations and comments were intended to offer insights into obstacles concerning the enrollment of undocumented children in school. [51]

3.1.2 Recording and transcription versus interview report based on simultaneous notes

Subsequently to the decision on the basic research design, it was necessary to determine whether the telephone calls would be recorded and transcribed or whether an interview report would be created based on simultaneous notes. In Table 1, a comparison is made between the expected advantages and disadvantages of recording and transcribing, and simultaneous notes of telephone calls.

Table 1: Evaluation of documentation methods for a short telephone survey on school enrollment [52]

For the purposes of our study, the decision against recording and transcription was primarily influenced by the expectation that the quality of the sample and the quality of the interviews would suffer. Based on experience from informal preliminary interviews with school staff, we expected that many school secretaries would have responded to parents immediately and routinely. However, it is likely that if a study requesting permission for an audio recording had been announced, they would have immediately referred to school administrators or declined to be interviewed. Beside our expectation of an increased rate of refusals, we expected extra caution with the wording. Conversely, the advantages of recording and transcription with respect to the quality of the documentation appeared to us less significant, especially since the interviews would generally be very short. In order to keep the disadvantages as low as possible, quality assurance measures were implemented. [53]

VOGEL undertook a pilot study with students of educational science. Unfortunately, for the purposes of this paper, a recorded interview and an interview documented by interview reports based on notes were not comparatively tested as the primary focus was on the basic feasibility of the method.¹⁴⁾ Nonetheless, the students conducted some interviews as two-person teams and some as individual interviewers. Accordingly, they observed a high accordance between the notes taken as a team interview and traced that back to the short length of the conversations during the telephone calls. Therefore, the students felt that there was no loss of quality resulting from individual interviews. [54]

FUNCK conducted all interviews in the main study. She was equipped with a headset that enabled her to have both hands free so that she could enter as much information as possible verbatim into the virtual questionnaire (computer-assisted telephone interviewing, see ENGEL, PÖTSCHKE & SIMONSON, 2005, p.1217). It was essential to have the ability to type quickly without looking at the keyboard.¹⁵⁾ FUNCK prepared for the interview with multiple test interviews and practiced computer-assisted note-taking of answers. [55]

During the interview, she documented the answers by ticking them off and writing them down in the free fields, partly in keywords and partly in the original wording. As the content was essential, repetitions, for example, were left out when taking notes. [56]

Whenever she was unsure whether she had understood an explanation correctly or if the interviewee had made many statements in a short time that seemed relevant to the research, she asked further questions ("Did I note/understand correctly that ...")¹⁶⁾. If necessary, further additions or corrections were made. Therefore, a communicative validation of the notes took place directly on the phone. Furthermore, one of the project leaders (VOGEL) was present from time to time during the interview phase. [57]

At the end of each interview, notes were immediately edited, i.e., sentences were completed and, for example, any typos, grammatical errors or punctuation mistakes were corrected. Any special expressions made in the interview were taken over verbatim ones and marked with quotation marks. Alternatively, if the exact wording could not be reproduced, indirect speech was used. [58]

Overall, the experience with reports based on simultaneous notes of short telephone interviews was positive. A response rate of 73% was achieved. 137 schools had to be contacted to reach the desired target number of 100 schools. In as many as 29% of the cases, the questions were answered directly by school secretaries and not referred to the school management or higher-ranking offices (FUNCK et al., 2015, pp.22-23). [59]

In only 10% of the cases, the respondents merely answered the short, closed questions. In the remaining cases, the respondents took the opportunity to explain and comment, sometimes with remarkable openness. For example, two school principals, after being assured of the anonymity of their statements, reported on their experiences with the school enrollment of undocumented children. There were some participants who showed their insecurities in dealing with school admissions: "Oh God! We've never had that before! I would check with the education authority." In the case of others, it became apparent that they had inaccurate legal knowledge that jeopardized school attendance. For instance, they had mistakenly assumed that it was permissible or even necessary to inform the police: "From a human point of view, I would enroll the child, but if it is illegal, I have to ask the police." Others provided justifications desirable under the current law: "It must not fail because of papers, we have to enroll the child. Education must not be denied." There was also documentation of statements with a clearly negative attitude, which was also reflected in the way they were expressed: "That [referring to the school enrollment of an undocumented kindergarten child] is not possible at all, no go, no go, I would say! (laughs) And then, I would call the brain of the kindergarten teacher into question. What a call! (laughs)." [60]

We regard these examples of short quotations as documented with high quality. Longer explanations could not be incorporated verbatim. During the review of the free-text entries for coding, there were some cases with insufficient documentation. Furthermore, in some cases, the interviewees talked fast and at length, so that only the most significant aspects with regard to the research questions were documented. In such cases, it cannot be ruled out that recording and transcription would have shed light on further explanations of interest that were not documented in the report. From our point of view, however, it is quite likely that an important aspect could also have been found in another, better-documented case and thus would have been available for the qualitative result of the study. We, therefore, consider the overall quality of the documentation to be sufficient for the purpose of the study. Also, from a retrospective point of view, we consider the choice of documentation method to be justified. [61]

A study about the living conditions of undocumented persons in Hamburg was finalized in 2009 (DIAKONISCHES WERK HAMBURG, 2009). It had been commissioned in 2008 by Diakonisches Werk Hamburg [a protestant social welfare organization], in cooperation with Nordelbische Kirche [a protestant church] and ver.di [the trade union of the service sector]. The purpose of this study was to identify the need for action and support, and to present empirical evidence concerning the effects of measures on the living situation of undocumented people. As part of this study, VOGEL and AßNER (2009) at the Hamburg Institute of International Economics (HWWI) undertook the sub-study described below. [62]

3.2.1 Aims and design of the sub-study on the number of undocumented people living in Hamburg

Prior to conducting the study, journalists cited a wide range of estimates of the size of the undocumented population in Hamburg—between 10,000 and 100,000 people (BLASBERG & BLASBERG, 2007). However, these figures were not supported by documented and substantiated justifications. In light of this, the HWWI researchers had the task of estimating the size of the undocumented population in a more precise and comprehensible way since proposed measures always raise the question as to how many people might be affected. [63]

Although studies on other cities have also provided figures, none of them provided a substantiated estimate of the number of undocumented immigrants. Methods used abroad, which JANDL (2008, 2011) described in a comprehensive overview, could not be applied for the most part (CYRUS, 2008).¹⁷⁾ Consequently, the estimation required extensive theoretical and methodological consideration, which eventually led to the development of a Logicom method estimating the upper and lower limits (VOGEL & AßNER, 2009, p.44).¹⁸⁾ [64]

For this purpose, researchers discussed with experts about which data are available or can be made available, how they are generated in the work process and what needs to be taken into account when interpreting them. Among others, interviews took place with responsible persons from the federal and state police, the Finanzkontrolle Schwarzarbeit [Financial Control of Undeclared Work], the Repatriation Program, and Medibüro [initiative for medical care of refugees]. The reason for this selection of interviews was that, based on the current state of research, it could be assumed that a subgroup of the population without valid residence papers was included in their process-produced individual data. Each interview was based on a separate guide. Due to the complexity of the issues involved, an interview duration of at least one hour was targeted. [65]

Based on the experience of previous studies, we assumed that government employees would consider the topic to be extremely sensitive. Furthermore, we expected group discussions involving several people with different levels of knowledge, including practitioners with detailed knowledge on the one hand and their supervisors with overview knowledge on the other. In an earlier study on a similar topic, VOGEL conducted and documented the interviews alone. In one case, she unexpectedly conducted a group interview with ten people, although she had expected an interview with three people. Interviewees ranged from simple practitioners to employees of different hierarchical levels and functional contexts to the person in charge at the ministry.¹⁹⁾ Based on this experience, the interviews were planned in advance as team interviews. [66]

It was vital for the study that the researchers would correctly understand complicated procedures and data limitations as presented by the interviewees and subsequently be able to represent their understanding in public without any revocation by the interviewees. Therefore, validation by interviewees was planned. [67]

Similar to the school enrollment study, the interviewees for this study were questioned as informants who have knowledge in a specific functional context. For these interviews, a documentation format had to be chosen. Therefore, in Table 2, the recording and transcription are compared to a paraphrasing team report, which summarizes the obtained understanding of data, data collection routines, and orientations.

Table 2: Evaluation of documentation methods for expert interviews on irregular migration data [68]

The team report was prepared collaboratively: One person carried out the interview, maintained eye contact with the interviewees and took only a few notes. The second person took extensive notes and posed only supplementary questions. On the same day, the second person prepared a preliminary draft of the interview report, which was then revised by the first person the following day. This immediate preparation of the interview report was intended to achieve the best possible memory performance. Subsequently, both discussed the report, which was revised in Microsoft Word's change mode, and produced a joint version that was sent to the interviewees for review. Any feedback from the interviewees was discussed and incorporated into the interview report. Communicative validation by the interviewees aimed at eliminating any misunderstandings in the content, thereby increasing the interview quality for answering the research question.²¹⁾ [69]

In selecting an interview report as the documentation method, three aspects were pivotal: First and foremost, it was expected to encourage more detailed and open statements, allowing researchers to have a better understanding of organizational activities and the data generated therein. In addition, the time and cost of transcription were saved, and a paraphrased summary was produced immediately afterwards based on notes. This was effective for the course of the study and, in addition, facilitated the timely availability of a report to the interviewees for validation purposes. Since a validation by the interviewees was planned regardless of the documentation method, a verbatim transcription of the interview appeared to offer relatively little advantage. Furthermore, it might be difficult to understand a transcription when interviewees provided handouts or showed data on screens. However, it cannot be ruled out that a paraphrased thematic summary based on transcripts would have pointed out other aspects that were missed in the note-based summary. Nevertheless, by using the team report method, this possible disadvantage or the risk of losing content-related aspects or observations was intended to be reduced. [70]

All in all, the experience with the paraphrasing team report was positive. It was possible to arrange four individual and two group interviews. In one case, there was a refusal to be interviewed due to time constraints, but a written response to questions was promised. This contained mainly legal information and statistical data. It can therefore be assumed that caution also played a role in the refusal of the interview. Upon follow-up, additional questions were answered by telephone. A transcribed interview would certainly not have been possible, at least in this case. With the follow-up questions and the provision of the paraphrased team report, some assurance could at least be obtained that those available statistics were interpreted in a meaningful way. [71]

Interviewees in the remaining interviews explained work contexts and described the formation of statistics openly and in detail. Furthermore, they agreed to review paraphrased formulations of the researchers that were not written from the perspective of their organization. The interview partners engaged in discussing some case studies and, in one case, demonstrated how data are entered on a computer, which was also helpful. The interviews were an important element in assessing meaningful reference values and directions of bias needed for the estimation of the number of irregular migrants in Hamburg. [72]

In this paper, we have tried to show why interview reports are not always a second-best solution for documenting the content of qualitative interviews but can also be the preferred method for some research questions. If interview topics are potentially sensitive from the point of view of the interviewees, it can make a difference in terms of willingness to participate or impartiality in the interview situation, whether interviews are recorded or interview reports are prepared based on notes. Even though more accurate documentation is made possible by recording and transcription, it is not superior if, as a result, the most appropriate interviewees are unwilling to participate or to discuss matters relevant to the research question. [73]

Furthermore, an interview report can be the method of choice if it accomplishes the same research goal with less effort, e.g., if the information on routines and processes is collected in informant interviews. Nonetheless, interview reports remain a second-best solution on occasions when they are inadequate for the research goal, e.g., when the focus is on the interviewees themselves and their subjective beliefs are in the foreground. In such cases, the use of interview reports should be avoided. Here, a transcript can at least offer a better starting point for further analyses, even if it also selectively reconstructs the interview situation. [74]

In the two studies presented (school enrollment study and statistics study), the interview report procedure was chosen as the first-best solution. In both cases, the informant interviews concerned sensitive topics (residency without valid documents). Interviewees' beliefs were of additional interest in the school enrollment study which were documented by noting verbatim quotes. In the statistics study, a verbatim record was of relatively little interest. A paraphrasing summary report was prepared and made available to the interviewees. Such a paraphrasing summary report could also have been done after recording and transcription. However, it would have led to higher cost and time, with no added value in answering the research question. With respect to the interview reports made for communicative validation, the aim is not to reconstruct the conversational situation but rather the understanding which the researchers gained based on the conversation. [75]

Yet, when interview reports can also be the documentation form of choice, more intensive thought should be given to methodological requirements. In the examples presented here, different quality assurance methods were used: Pilot and practice interviews, technical aids for telephone surveys, the immediacy of writing down information, interviewing and note-taking in teams, and communicative validation. [76]

In general, interview reports can be used in cases where, due to the sensitivity of the subject matter, it is likely that audio recordings will compromise the quality of the sample and/or the course of the interview. Interview reports are more appropriate for informant interviews than for respondent interviews. In certain cases, word-by-word note-taking is also possible for interview reports based on simultaneous notes, for instance, in the case of short telephone interviews with a highly focused research question. [77]

In conclusion, these studies show that interview reports do not always have to be the second-best solution for documenting qualitative interviews but can also be the best form of documentation, depending on the research question. Even if reports are also prepared in a supportive manner, e.g., in the form of a postscript to a transcript, a sociological discussion about the quality of reports makes sense. In our view, this has only been done to a limited extent (e.g., by LOUBERE, 2017 in the context of field research). Therefore, this paper aims to stimulate further discussion on this issue. In the Appendix, a sample interview report can be found as a helpful first step. [78]

During the review process for this paper, we discussed with anonymous reviewers and members of the FQS editorial team whether we implicitly followed a positivist position in our contribution, especially because we initially neglected the constructivist character of transcripts. We, therefore, point out that we assume that both interviewees and interviewers (re)construct social facts, but not in an arbitrary form. We, thus, regard factuality as relevant: "Factuality may be a fiction, but it appears to be a useful one, and attention to how (constructed) facts relate to (constructed) claims and theories is a widely recognized hallmark of good-quality research" (SEALE, 2004, p.411). [79]

We would like to express our sincere thanks to Mona SCHIELE for her support in the literature research and Norbert CYRUS for his critical comments and constructive suggestions. We would also like to thank the anonymous reviewers and members of the FQS editorial board for their suggestions that led to the revision of some parts of the text. The completion of this text was delayed due to the birth of a child and the completion of a project.

We would also like to thank Lalit SINGH NEGI for the translation and to Katrina STOLLMANN for the proof-reading of the text.

There had been prior communication on the topics of the interview by e-mail. The interview took place in the office of [Name 1]. Ms. [Name 1] answered all questions and consulted Mr. [Name 2] when she did not have detailed information on hand. Both interview partners were very interested and cooperative. On several occasions, graphs were brought out for explanation, which were scanned and incorporated into the interview report.

On three occasions, the interview was briefly interrupted by people knocking on the door to clarify something with Ms. [Name 1]. These interruptions could be used for additional notes. Based on notes taken by Funck, the interview report was prepared on the same day and revised by Vogel the following day. The thematic outline does not necessarily follow the course of the interview. References and notes for queries are listed in footnotes, and for some, there may still be a need for clarification.

[Name 1] has a degree and has been working in [Institution 1] for ten years. [Name 1] first worked as an interviewer, then as a project manager, and has been working for the last two years as a managing director. [Name 2] studied business administration and after graduating, started working as an executive assistant six months ago.

In the interview, the causes and processes of cooperation between Institution 1 and other institutions were discussed.

Researchers agreed with the interviewees to send them an interview report within one week after the interview and include potentially open questions.

1) This text is a translation of the original German article published in 2018. <back>

2) What LOUBERE (2017, §4) called "systematic interview report" is very similar to our Protokoll in the original German version of the article. By comparison, the English term "interview protocol" is used in the sense of interview guideline. <back>

3) These categories are based on a classic paper by ZELDITCH (1993, p.121) but are sorted and named differently. <back>

4) This question does not arise with internet-based written interview forms (OPDENAKKER, 2006, §24). <back>

5) The time and cost of transcription have changed significantly between the publication of the original article in 2018 and the publication of this translation in 2024. Automatic transcription based on artificial intelligence has seen notable improvements, providing a draft transcription of good quality in approximately 30 minutes for a one-hour interview, as stated by f4x (https://www.audiotranskription.de/f4x/). The transcription quality is satisfactory for interviews with a single person, significantly reducing the required time even after manual control and correction. Additionally, in group interviews under challenging conditions (involving many people, similar voices, accents), which require more time for control and correction, the overall transcription time can still be slightly reduced with the use of automatic transcription. <back>

6) As transcriptions also involve interpretation, there is a discussion in the qualitative literature about the use of paid transcribers, addressing the role of briefings, quality control, confidentiality issues, and the benefits of transcription by research participants (DAVIDSON, 2009, pp.43-44). <back>

7) These experiences stem from the project by CYRUS and VOGEL (2002) on external controls in the labor market and work permit procedures, as well as from VOGEL and AßNER (2009) on the statistical recording of cases involving illegal residence. In surveys with public authority employees, FUNCK (2015) also found that the more sensitive information was usually given when the recording device was already switched off. <back>

8) According to LAMNEK (2002, p.190), the interviewers' familiarity with recordings ensured that after a very short start-up phase, interviewees ceased to find them irritating, unusual, or even disturbing. However, it is neither immediately apparent to us nor did we find any explanation for why an interviewee should feel more comfortable with an interview recording if the other person is familiar with it. <back>

9) Transcripts are hardly suitable for this purpose because the speakers do not usually recognize transcripts as a correct reproduction of their spoken contributions (BREUER, 1999, p.251). <back>

10) The European non-governmental organization PICUM (Platform for International Cooperation on Undocumented Migrants), which campaigns for the rights of undocumented migrants, accepts terms such as "undocumented" or "irregular" migrants. They refuse using the term "illegal" because it is often associated with criminality and suggests a lack of rights (PICUM, 2017). <back>

11) The selection of cities was intended to provide a reasonable estimate of the share of unenrolled children. Within the cities, the schools were selected randomly (FUNCK et al., 2015, p.15). <back>

12) In all contacted schools, the school secretaries introduced themselves as females, and therefore they are referred to as female here. <back>

13) This is based on 100 interviews, each lasting an average of five minutes and a transcription time of five times the length of the interview. As a guideline for transcriptions, four to eight times the interview time is assumed (KUCKARTZ et al., 2008, p.29). <back>

14) It was evident from the pilot study that timing and the amount of time required were crucial factors in willingness to participate. Both secretaries and school administrators participated. There was also evidence of particular caution before answering the questions. In some cases, for instance, there was a request for confirmation from the university with more detailed information about the study. Furthermore, there were significant differences in willingness to answer, the causes of which remained unclear but may also have been due to interviewer effects. <back>

15) The questionnaire had been programmed with the software Epi-Data which facilitated an easy input and rapid switching from question to question. <back>

17) The residual method used in the U.S., for instance, assumes that population data for certain countries of origin are higher in the census than in control data, whereas the opposite is the case in Germany. The center-sampling method used in Italy presumes that interviewees in migrant communities admit undocumented residency. Based on experience with qualitative interviews, this is not realistic in Germany. <back>

18) It is basically a matter of understanding which subgroup of undocumented migrants is covered in a database and whether they are over- or under-represented in relation to another group for which data are available in the same database. In that case, it is possible to calculate an upper or lower limit by applying a simple rule of three. <back>

19) The most significant results of the interviews have been included in working papers and essays, in which methodological aspects of conducting the interviews were not described in detail (VOGEL, 1998, 2000). <back>

20) This is based on seven interviews lasting an average of 100 minutes and a transcription time of six times the interview length. As a reference value for transcriptions, four to eight times the interview time is assumed (KUCKARTZ et al., 2008, p.29). <back>

21) It should be pointed out at this point, however, that this is not always advantageous—depending on the particular research concern. For example, it is possible that the statements made or the terms used may be revised because they are subsequently classified by the interviewee as problematic or socially unacceptable. <back>

Ashmore, Malcolm & Reed, Darren (2000). Innocence and nostalgia in conversation analysis: The dynamic relations of tape and transcript. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 3, http://dx.doi.org/10.17169/fqs-1.3.1020 [Accessed: December 20, 2023].

Bergmann, Jörg R. (2009). Mediale Repräsentation in der qualitativen Sozialforschung. In Ruth Ayaß & Jörg Bergmann (Eds.), Qualitative Methoden der Medienforschung (pp.489-506). Mannheim: Verlag für Gesprächsforschung, http://www.verlag-gespraechsforschung.de/2011/pdf/medienforschung.pdf [Accessed: December 20, 2023].

Blasberg, Anita & Blasberg, Marian (2007). Sklaven in Altona. Die Zeit, March 8, 11, http://www.zeit.de/2007/11/Illegale [Accessed: December 20, 2023].

Bogner, Alexander & Menz, Wolfgang (2009). Das theoriegenerierende Experteninterview. Erkenntnisinteresse, Wissensformen, Interaktion. In Alexander Bogner, Beate Littig & Wolfgang Menz (Eds.), Experteninterviews. Theorien, Methoden, Anwendungsfelder (3rd ed., pp.61-98). Wiesbaden: VS Verlag für Sozialwissenschaften.

Breidenstein, Georg; Hirschauer, Stefan; Kalthoff, Herbert & Nieswand, Boris (2013). Ethnografie: Die Praxis der Feldforschung. Konstanz: UVK Verlags-Gesellschaft.

Breuer, Franz (1999). Probleme human- und sozialwissenschaftlicher Erkenntnismethoden: Viel Verwirrung – einige Vorschläge. In Norbert Groeben (Ed.), Zur Programmatik einer sozialwissenschaftlichen Psychologie (pp.193-309). Münster: Aschendorff.

Cyrus, Norbert & Vogel, Dita (2002). Ausländerdiskriminierung durch Außenkontrollen im Arbeitsmarkt? Fallstudienbefunde – Herausforderungen – Gestaltungsoptionen. Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 35(2), 254-270.

Cyrus, Norbert & Vogel, Dita (2006). Managing access to the German labour market: Opportunities and restrictions. In Franck Düvell (Ed.), Illegal immigration in Europe beyond control? (pp.75-105). Basingstoke: Palgrave Macmillan.

Engel, Uwe; Pötschke, Manuela & Simonson, Julia (2005). Telefonsurveys: Vor- und Nachteile. Bundesgesundheitsblatt – Gesundheitsforschung – Gesundheitsschutz, 48(11), 1217-1223.

Flick, Uwe (1991). Stationen des qualitativen Forschungsprozesses. In Uwe Flick, Ernst von Kardorff, Heiner Keupp, Lutz von Rosenstiel & Stephan Wolff (Hrsg.), Handbuch Qualitative Sozialforschung. Grundlagen, Konzepte, Methoden und Anwendungen (pp.148-173). München: Beltz.

Funck, Barbara Johanna (2015). Lernförderung im Bildungs- und Teilhabepaket. Eine akteurzentrierte Untersuchung der lokalen Umsetzung in Bremen. Unpublished Master Thesis, Universität Bremen, Germany.

Fuß, Susanne & Karbach, Ute (2014). Grundlagen der Transkription: Eine praktische Einführung. Opladen: Verlag Barbara Budrich.

Gläser, Jochen & Laudel, Grit (2009). Experteninterviews und qualitative Inhaltsanalyse. Wiesbaden: VS Verlag für Sozialwissenschaften.

Green, Judith; Franquiz, Maria & Dixon, Carol (1997). The myth of the objective transcript: Transcribing as a situated act. TESOL Quarterly, 31, 172-176.

Jandl, Michael (2008). Chapter 3: Methods for estimating stocks and flows of irregular migrants. In Dita Vogel & Albert Kraler (Eds.), Report on methodological issues. Research project CLANDESTINO (Undocumented Migration: Counting the Uncountable. Data and Trends Across Europe) (pp.19-52), https://migrant-integration.ec.europa.eu/sites/default/files/2009-07/docl_9061_620418687.pdf [Accessed: December 20, 2023].

Jandl, Michael (2011). Methods, approaches and data sources for estimating stocks of irregular migrants. International Migration, 49(5), 53-77.

Kaiser, Robert (2014). Qualitative Experteninterviews: Konzeptionelle Grundlagen und praktische Durchführung. Wiesbaden: Springer VS.

Kuckartz, Udo; Dresing, Thorsten; Rädiker, Stefan & Stefer, Claus (2008). Qualitative Evaluation: Der Einstieg in die Praxis (2nd upd. ed.). Wiesbaden: VS Verlag für Sozialwissenschaften.

Kvale, Steinar & Brinkmann, Svend (2009). InterViews: Learning the craft of qualitative research interviewing (2nd ed.). Thousand Oaks, CA: Sage.

Lamnek, Siegfried (2002). Qualitative Interviews. In Eckard König & Peter Zedler (Hrsg.), Qualitative Forschung. Grundlagen und Methoden (2nd ed., pp.157-193). Weinheim: Beltz.

Lamnek, Siegfried (2010). Qualitative Sozialforschung: Lehrbuch (5th upd. ed.). Weinheim: Beltz.

Liebold, Renate & Trinczek, Rainer (2009). Experteninterview. In Stefan Kühl, Petra Strodtholz & Andreas Taffertshofer (Eds.), Handbuch Methoden der Organisationsforschung. Quantitative und Qualitative Methoden (pp.32-56). Wiesbaden: VS Verlag für Sozialwissenschaften.

Loubere, Nicholas (2017). Questioning transcription. The case for the systematic and reflexive interviewing and reporting (SRIR) method. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 18(2), Art. 15, http://dx.doi.org/10.17169/fqs-18.2.2739 [Accessed: December 20, 2023].

Meuser, Michael & Nagel, Ulrike (1991). ExpertInneninterviews – vielfach erprobt, wenig bedacht: Ein Beitrag zur qualitativen Methodendiskussion. In Detlef Garz & Klaus Kraimer (Eds.), Qualitativ-empirische Sozialforschung: Konzepte, Methoden, Analysen (pp.441-471). Opladen: Westdeutscher Verlag.

Meuser, Michael & Nagel, Ulrike (2009). Experteninterview und der Wandel der Wissensproduktion. In Alexander Bogner, Beate Littig & Wolfgang Menz (Eds.), Experteninterviews. Theorien, Methoden, Anwendungsfelder (3rd ed., pp.35-60). Wiesbaden: VS Verlag für Sozialwissenschaften.

Meuser, Michael & Nagel, Ulrike (2013). Experteninterviews – wissenssoziologische Voraussetzungen und methodische Durchführung. In Barbara Friebertshäuser, Antje Langer & Annedore Prengel (Eds.), Handbuch qualitative Forschungsmethoden in der Erziehungswissenschaft (4th ed., pp.457-471). Weinheim: Beltz Juventa.

Mey, Günter & Mruck, Katja (2010). Interviews. In Günter Mey & Katja Mruck (Eds.), Handbuch Qualitative Forschung in der Psychologie (pp.423-435). Wiesbaden: VS Verlag für Sozialwissenschaften.

Opdenakker, Raymond (2006). Advantages and disadvantages of four interview techniques in qualitative research. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(4), Art. 11, http://dx.doi.org/10.17169/fqs-7.4.175 [Accessed: December 20, 2023].

Patton, Michael Q. (2002). Qualitative research & evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Rubin, Herbert J. & Rubin, Irene S. (2005). Qualitative interviewing: The art of hearing data. Thousand Oaks, CA: Sage.

Seale, Clive (2004). Quality in qualitative research. In Clive Seale, Gaimpietro Gobo, Jaber F. Gubrium & David Silverman (Eds.), Qualitative research practice (pp.409-419). London: Sage.

Streck, Rebekka; Unterkofler, Ursula & Reinecke-Terner, Anja (2013). Das "Fremdwerden" eigener Beobachtungsprotokolle – Rekonstruktionen von Schreibpraxen als methodische Reflexion. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(1), Art. 16, http://dx.doi.org/10.17169/fqs-14.1.1821 [Accessed: December 20, 2023].

Vogel, Dita (2000). Migration Control in Germany and the United States. International Migration Review, 34(2), 390-422.

Vogel, Dita & Aßner, Manuel (2009). Wie viele Menschen leben illegal in Hamburg? Eine Schätzung der Gesamtzahl und ausgewählter Strukturmerkmale mit der Logicom-Methode. In Diakonisches Werk Hamburg (Ed.), Leben ohne Papiere. Eine empirische Studie zur Lebenssituation von Menschen ohne gültige Aufenthaltspapiere in Hamburg (pp.9-132). Hamburg: Diakonisches Werk Hamburg, https://www.ew.uni-hamburg.de/ueber-die-fakultaet/personen/neumann/files/diakonie-endfassung.pdf [Accessed: December 20, 2023].

Zelditch, Morris (1993). Methodologische Probleme in der Feldforschung. In Christel Hopf (Ed.), Qualitative Sozialforschung (3rd ed., pp.119-137). Stuttgart: Klett-Cotta.

Dr. Dita VOGEL is a senior researcher in the Department of Migration, Education and Society / Intercultural Education at the University of Bremen. In her research, she focuses on educational policy development and school development in relation to migration, mobility and languages in Bremen, Germany and in an internationally comparative perspective.

After studying economics (University of Cologne), VOGEL completed her doctorate at the University of Bremen on the fiscal effects of immigration (1996). There, she also conducted interdisciplinary research and learned qualitative research methods at the Centre for Social Policy from 1989 to 1997. After completing her doctorate, VOGEL was involved as a senior researcher in numerous interdisciplinary and comparative projects in the field of migration and education at the Universities of Bremen and Oldenburg and the Hamburg Institute of International Economics.

University of Bremen
FB12 Unit for Intercultural Education
Postfach 330440
28334 Bremen, Germany

Dr. Barbara Johanna FUNCK works as a research assistant (post-doc) in the Department of Migration, Education and Society / Intercultural Education at the University of Bremen. In her research, she focuses on the local realization of the human right to education, the accessibility of the education system in the context of migration, and different conditions of residency.

After studying social sciences (B.A., University of Marburg) and social politics (M.A., University of Bremen), FUNCK worked in several research projects and as lecturer for education at the University of Bremen. In her doctoral study (completed in 2023), she employed follow-the-people methods to reconstruct school access in the context of migration and different conditions of residency, starting with parent interviews.

University of Bremen
FB12 Unit for Intercultural Education
Postfach 330440
28334 Bremen, Germany

Vogel, Dita & Funck, Barbara J. (2024). Only second best? Interview reports as a method of documenting qualitative interviews [79 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 19(1), Art. 7,
https://dx.doi.org/10.17169/fqs-19.1.2716.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Funded by the KOALA project

Creative Commons Attribution 4.0 International License

	Recording and transcription	Interview report based on simultaneous notes	Comparison
Time and financial costs	Approximately 40 hours¹³⁾ of transcription time; cost of transcription; availability only after transcription	Notes during the interview; short time for revision; immediate availability after the interview	Advantage of report: Time and cost savings with respect to the amount of transcription time and costs
Quality of the sample: Response rate	Expectation that secretaries frequently refuse or refer instead of providing information directly, as they would answering parent calls; lower willingness to participate	Expectation of more frequent responses from secretaries and school administrators directly and thus a higher willingness to participate	Advantage of report: Higher response rate and less biased composition of interviewees
Quality of the interview	Expectation of more inhibited response behavior and more cautious choice of words because of fear of sanctions for careless remarks	Expectation of a more open response behavior	Advantage of report: Less biased because fear of sanctions is less relevant
Quality of the documentation	High accuracy of the reproduction of what was said*	Higher risk of gaps and misunderstandings, as documentation depends on the speaking speed of the interviewees, note-taking and memory skills of the interviewer.	Advantage of Transcription: More comprehensive and more accurate documentation of interviews

	Recording and transcription	Team report	Comparison
Time and financial costs	Approximately 70 hours²⁰⁾ of transcription time; paraphrased summary based on the transcript; availability after transcription and paraphrasing.	A paraphrased summary based on notes and memories; availability after writing the report	Advantage of team report: Time and cost savings with respect to the amount of transcription time and costs; no significant time differences in the paraphrasing time
Quality of the sample: Response rate	Expectation that the people who are contacted will participate	Expectation that the people who are contacted will participate	No significant differences
Quality of the interview	Verbatim reproduction; potentially difficult to understand if, for example, tables are shown to the interviewers for visualization purposes	Documentation subject to the speed of speech of the interviewees, note-taking and memory skills of the interviewers, and the attention of the interview partners reading the report	Advantage of transcription: Comprehensive and accurate documentation of interviews, fewer gaps Advantage of team report: Correction of misunderstandings through team report
Quality of the documentation	High accuracy of the reproduction of what was said	Higher risk of gaps and misunderstandings, as documentation depends on the speaking speed of the interviewees, note-taking and memory skills of the interviewer.	Advantage of Transcription: More comprehensive and more accurate documentation of interviews