Qualitative Interview Research in Multilingual Contexts—A Comparative Discussion of Language-Related Decisions in Two Empirical Studies

Abstract: Research settings are often multilingual in the sense that researchers and participants speak different first languages. This should be explicitly considered in qualitative research in order to enable or facilitate communication, promote understanding, and allow for context-sensitive analyses of intent and meaning.

In this article we comparatively address methodological considerations in two research projects conducted in Austria and Germany. Based on the project-specific context, research interests, sampling strategies, and participants, we developed different methodological approaches to address the respective settings and linguistic needs. We argue that there is no generic guideline but aim to provide orientation for context-specific project design. Based on the assumption that in every research endeavor, a contextualized strategy needs to be developed for the entire project from idea development and project design to the dissemination of results, we highlight language-related decisions throughout qualitative research projects.

Key words: multilingual research; translation; second language interview research; language-related research design; interpreters; migration research; refugee studies; multilingual interviews; socio-translational collaborative approach; episodic interviews; research ethics; positionality; study design

Multilingualism is a characteristic of super-diverse societies (VERTOVEC, 2023). In social sciences, this often leads to constellations in which researchers and participants do not share the same first language(s) or even a common language. Also, international research can make multi-language approaches necessary (BADING, KAZZAZI & WINTZER, 2025, BANERHEE & SOWARDS, 2022). Whether and how language and translation are addressed varies in practice and depends, among other things, on the research question, research approach, and the (financial) framework conditions. [1]

During an informal discussion of our respective experiences in migration research, we realized that we had made very different methodological and translation-related decisions. However, our choices seemed plausible in the respective research environments. Hence, we had the idea to structurally compare our respective contexts and empirical approaches to multilingual research. This resulted in a collaborative workshop held in Vienna in Spring 2024. The workshop gave us the opportunity to familiarize ourselves with each other's projects in extraordinary depth. We aimed to understand the questions and methodological approaches, to discuss the broader context and practical experiences and to critically reflect on our decision making. Both reference projects are situated in the field of migration studies. However, in order to investigate intent and meaning making, we emphasize the importance of considering societal multilingualism and related issues of communication, language mediation, and translation in all kinds of qualitative research projects. [2]

First, we will provide an overview of the state of literature on research in polyglot settings and translation in research projects (Section 2). In Section 3, we proceed to introduce both of our research projects to provide a more detailed understanding of their respective aims, contexts and empirical considerations. This will be followed by a discussion of language-related points of decision in research projects, which is meant to provide a starting point to develop individual, project-specific strategies to address multilingual research contexts (Section 4). We conclude by highlighting key insights that are based on our research experience and joint reflection (Section 5). [3]

While multilingualism is relevant for qualitative research in all social science fields (see e.g., the contributions in HOLZINGER & DRAXL, forthcoming), methodological debates primarily arise in migration research, e.g., around various approaches to interview research (KRUSE, BETHMANN, NIERMANN & SCHMIEDER, 2012) or the translation of data (BERGER, 2023; BURKHARD & PARK, 2019; LARKIN, DIERCKX DE CASTERLÉ & SCHOTSMANS, 2007). HOLMES, REYNOLDS and GANASSIN (2022) emphasized that power relations, the significance of languages, and language hierarchies in the research process, as well as the role of interpreters and translators, are still too rarely and insufficiently reflected upon. In recent years, however, there has been an increasing awareness of power imbalances and the choice of interview languages, especially in empirical forced migration research (TREIBER & KAZZAZI, 2020). [4]

In qualitative interview research, the choice of language can result in different linguistic and personal constellations. A direct exchange between interviewers and interviewees can rely on the multilingualism of either the researcher or the participant (who speaks the first language of the other)—or on both parties' multilingualism by using a lingua franca (SCHITTENHELM, 2017). Alternatively, interpreters can be involved to facilitate communication. They can allow for using the interviewees' first language regardless of the interviewers' linguistic competencies. It should be considered that their presence can change the social interaction during the interview (LAUTERBAUCH, 2014). In practice, the degree of interpreters' professionalism varies greatly, from academically trained translators to lay interpreters (BRÄMBERG & DAHLBERG, 2013; INHETVEEN, 2012; LITTIG & PÖCHHACKER, 2014). Each approach has been shown to have advantages and disadvantages. For example, using the main language of the host country in interviews with migrants (who speak a different first language) has been criticized for potentially reproducing hegemonic power relations in the research interaction (THIELEN, 2009). On the other hand, the same practice has been discussed as an opportunity to recognize the language skills that migrants have acquired or to account for their lifeworld linguistic experiences (BERG, GRÜTTNER & SCHRÖDER, 2019a; SHERIDAN & STORCH, 2009). Instead of strictly presuming one language, translanguaging, code-switching and -mixing have also been addressed. For example, GARCÌA and KLEIFGEN (2020) discussed the potential of translanguaging to address polyglot language literacies and the inter-connectedness of linguistic systems and "participants' simultaneous use of multiple semiotic resources that mutually elaborate each other" (p.557; see also MacSWAN, 2017). [5]

The use of interpreters has been controversial, especially in the context of forced migration research. BOGNER and ROSENTHAL (2014, §14) described a trust-building effect and argue that time needed for translation could provide pause for reflection and to deal with emotionally challenging or sensitive topics during the interview (HAUG, LOCHNER & HUBER, 2017, p.7). By contrast, FRITSCHE (2016) has argued that not using interpreters allows for "direct communication and [...] underlines the active speaker role" (p.173).¹⁾ Furthermore, interpreters might create parallels with interviews conducted in settings like the asylum procedure (BERG et al. 2019a). Additionally, as UÇAN (2019, p.135) emphasized with reference to postcolonial perspectives, power and inequality dynamics in interview situations are not necessarily resolved when the interviewee's first language is used, but the ambivalences contained in the interview situation have to be faced. Likewise, the literature has critiqued a monolingual habitus in research (CAN, 2020; RECKINGER, 2020). Instead, interview practices can strive to allow for translingual practices and to recognize their analytical potential (HAVLIN, 2022; UÇAN, 2019). [6]

Different forms of written or verbal translations may be required throughout a project (e.g., preparation, data collection, data processing, data analysis, dissemination) and can be done by either professional translators (LARKIN et al., 2007), interpreters, or the researchers themselves (BURKHARD & PARK, 2019), who engage in "paraprofessional translation" activity (KOSKELA, KOSKINEN & PILKE, 2017, p.466). In this context, it should be considered to what extent implicit meaning, intention, and the social positioning of knowledge are comprehensible and translatable (GOITOM, 2020). Involving interpreters or researchers with the same first language as the interviewees can help access their "cultural competence [i.e.,] knowledge of culture- and milieu-specific connotations of expressions and the competence to convey these" (UÇAN, 2019, p.119). However, the risk that translations might (unconsciously) reproduce inequalities and power imbalances within the speaker group must always be scrutinized and reflected upon. For example, speakers of the same first language could be stratified in their life-experience or belong to different (i.e., political) in-group fractions. TEMPLE, EDWARDS and ALEXANDER (2006) have argued that cross-language data analysis can have strong similarities to secondary data analysis. More generally, we have previously argued that polyglot research settings require specific reflection processes, documentation, and transparency in regard to language-related decisions (HOLZINGER & DRAXL, 2024a). In this article we realize this claim by comparatively discussing decision points in the course of two empirical projects. We confine our discussion to spoken multilingualism (for a methodical approach to spoken and signed multilingualism see KUSTERS & DE MEULDER, 2019). [7]

In this section, we provide a brief overview of the context and participants of Wedge and DEMICO, the projects we reflected on during our collaborative workshop. [8]

We conducted the Pathways of Refugees into Higher Education (WeGe) research project at the German Centre for Higher Education Research and Science Studies from April 2017 to March 2021. Based on SEN's (1985, 1999) capabilities approach, we investigated challenges faced by prospective refugee students in preparing for their studies and factors influencing their successful transition to higher education (GRÜTTNER, SCHRÖDER, BERG & OTTO, 2021). Against the background of the incomplete literature, the project was not exclusively explanatory, but also exploratory and interpretative, and we thus pursued a mixed-methods approach (SCHOONENBOOM & JOHNSON, 2017). In addition to a quantitative longitudinal survey and expert interviews, we primarily aimed at interpreting the experiences and constructions of meaning of refugees participating in preparatory courses at German higher education institutions that conclude with a certification of sufficient German language skills. We focus here on the qualitative part of the study, as the quantitative methods are beyond the scope of this paper. [9]

The sample for the qualitative panel study consisted of refugee students in study preparation courses. Given the access requirements for those courses, we assumed the participants to speak German at a B2 level, preparing for C1, based on the Common European Framework of Reference for Languages (CEFR)²⁾ (SCHRÖDER, GRÜTTNER & BERG, 2019). Further sampling criteria were gender and the country of origin. Although language was not an explicit criterion, our sampling process indirectly led to the selection of people with multiple first languages, such as Arabic, Dari, French, or Urdu. As we chose participants based on residence status without specifying countries of origin, we were unable to predict the exact first languages we would encounter, but we could expect relatively high levels of German proficiency. Before beginning any empirical fieldwork, we developed an awareness strategy for quantitative and qualitative research with refugee students (BERG, GRÜTTNER & SCHRÖDER, 2019a, 2019b). Our data collection methodology was closely aligned with participants' presumed characteristics and needs. We conducted episodic interviews (FLICK, 2000). In contrast to biographical-narrative interviews, this type of interview is not to explore big (life) stories, but rather small stories, relating to specific periods of life, events, daily routines, etc. (DAUSIEN & THOMA, 2023). To express appreciation for the obtained language skills and to be able to interview the course participants ourselves, we chose to offer interviews in German and/or English (BERG et al., 2019a). All materials used for data collection, including the interview guideline and informed consent forms, were made available in German and English. We discussed them with experts in the field, such as study preparation teachers, and all pre-study interviews were concluded with a few meta questions, e.g., on wordings and language choice. The data analysis was based on qualitative content analysis procedures. Additionally, we used positioning analysis as a tool to reflect upon self- and third-party positioning throughout the interviews, including our own positionings (DEPPERMANN, 2013a). At the start of the project, we emphasized an intensive exchange with study participants in order to align our readings with their perceptions. [10]

The research project Investigating the Social Construction of Deskilling Among "New" EU Migrants in Vienna (DEMICO) (University of Vienna, 2021-2025) investigated the social construction of deskilling among highly educated EU migrants from Central and Eastern European EU countries in Vienna. To uncover individual motives and migrant agency in coping with (the risks of) deskilling, we applied a qualitative‐interpretative approach that drew on the methodological principles of constructivist grounded theory methodology (CGTM) (CHARMAZ, 2014). Paying special attention to the temporal dimension of deskilling, we conducted a qualitative panel study with migrants across three interview waves (HOLZINGER & DRAXL, 2025). [11]

Our sample consisted of Czech, Hungarian, and Romanian citizens with academic degrees from tertiary institutions outside of Austria who currently live and/or work in Vienna. In total, we conducted 49 interpretative interviews (SCHEIBELHOFER, 2023) in three waves. Given the importance of communication in interpretative research, we followed the socio-translational collaborative approach suggested by LITTIG and PÖCHHACKER (2014) and collaborated closely with a team of six interpreters and translators³⁾. 15 of the interviews were conducted in the first language of the interviewees (Czech, Hungarian, or Romanian) with the help of an interpreter, ten were conducted in English, and 24 in German. The interviews were recorded⁴⁾ and later transcribed in all the recorded languages. The Czech, Hungarian, and Romanian passages from these multilingual transcripts were subsequently translated to German⁵⁾. Our interpretative data analysis was oriented towards initial and focused coding techniques from CGTM, as proposed by CHARMAZ (2014). [12]

In the following analysis, we discuss selected characteristics relevant to multilingual qualitative interview research that emerged from the critical reflection of our research experience in WeGe and DEMICO. We aim to address key issues to provide guidance for future research projects in polyglot contexts. [13]

Based on our research experience within the two above-described projects, we argue that methodological aspects related to the field's multilingualism should be considered when a research team is initially assembled. On the one hand, experience in polyglot settings and research, as well as language proficiency, can become relevant during recruitment processes (ENZENHOFER & RESCH, 2019). On the other hand, methodological, translation-related, as well as culture and language sensitivity trainings can help promote personnel development. [14]

However, human resource management cannot replace professional translation expertise. It is difficult to choose appropriate project members if the participants' first languages are not predetermined by the sample or project design, as in the WeGe project. Furthermore, there might be a lack of eligible candidates to cover multiple first languages. If various languages are spoken across the team, the question remains whether and how the data collected in different languages should be made accessible (partly or in its entirety) for the purpose of intersubjective analysis to those team members who do not speak these languages. Besides these practical limitations, recourse to multilingual qualitative interviewers without proper translation expertise may pose further problems. For example, linguistic or translational expertise is not automatically given by someone's native speaker status. [15]

Given the complexity of translational tasks, it has been argued that in multilingual settings, the social researchers' professional expertise is thus ideally complemented by professional translation expertise (LITTIG & PÖCHHACKER, 2014). This was our chosen approach in the DEMICO project, because our sample allowed us to plan for the interviewees' first languages and for long-term collaboration with a team of translators. The core team's composition was not based on language-related considerations (all three of us speak German as a first language and have no or only very limited knowledge of the interviewees' first languages), but on previous collaboration experiences. From the project's onset, we thus planned to systematically involve qualified interpreters and translators throughout the entire research process. After receiving funding, we expanded our team to include certified interpreters and translators (most of whom work at the Centre for Translation Studies, University of Vienna). Basic knowledge of qualitative research was a recruitment criterion and this knowledge was expanded systematically during continuous workshops throughout the entire research process. Beyond translations and interpretations, the translation team's professional linguistic and cultural expertise⁶⁾ contributed to the construction of interview guidelines, interview participant recruitment, as well as establishing field contacts. Additionally, it was valuable for the analyses (see Section 4.6). [16]

Following the socio-translational collaborative approach (LITTIG & PÖCHHACKER, 2014), as we did in the DEMICO project, is indeed time-consuming as well as cost intensive and consequently, not feasible in all research contexts. Limited personnel and financial resources may necessitate outsourcing translation work. However, based on our experiences, we argue that translation is not a one-off, purchased product. Rather, it is processual and interactive: Not only should the translation product be commissioned, but ideally (organizational and financial) arrangements will be made so that the translators' expertise is available for consultancy, inquiries, and discussions during the whole research process. [17]

Sampling involves identifying characteristics relevant to the research question which also impact the project's implementation. It requires knowledge of and reflection upon the research population, which is then incorporated into further methodological considerations. Sampling in multilingual contexts is closely linked to the participants' language competencies. This is always the case, regardless of whether language is explicitly addressed during the sampling process or not, as it has relevant implications for the study's accessibility and the inclusion of potential participants in the sample. Each project should include reflections on whether and how communication strategies, including language selection, act as a criterion for inclusion or exclusion, motivation, or the deterrence of potential participants. [18]

When the sampling criteria explicitly include language skills, the project team can plan accordingly to pre-determine the interview mode and interview language. During the DEMICO project, researchers did this by deliberately relying on interpreters for their chosen interpretative methodological approach: To achieve a profound understanding of the participants' perspectives, we opted to give them a choice between first language, English and German, which would best allow them to express themselves as freely as possible from linguistic constraints. Also, we strove to not exclude potential interviewees due to their (self-assessed) insufficient proficiency in English or German. In turn, this methodological consideration had considerable influence on sampling criteria: We decided that a close collaboration with a fixed group of interpreters and translators throughout the research process required limiting the language groups for financial, team-dynamic, and organizational reasons. This meant limiting our sample to three countries of origin, inter alia due to the chosen multilingual approach and ensuing pragmatic constraints. [19]

Inversely, participants' non-linguistic characteristics, like the experience of forced migration, can influence decisions around language selection and interview mode. This was the case in the WeGe project where we chose not to use interpreters, as multiple, previously undeterminable first languages posed challenges for adequately preparing multilingual materials and empirical approaches, and we aimed to prevent parallels to interviews during the asylum procedure. At the same time, all interviewees participated in university-preparation courses, which required a high level of German proficiency (minimum B2 on the Common European Framework of Reference for Languages), which we aimed to acknowledge (BERG et al., 2019a). This goes to show how non-language-related sampling criteria can infer language skills, in this case, relatively high German proficiency levels. [20]

Access to and the recruitment of interviewees is closely related to linguistic needs and competencies in the field. Researchers or interpreters who speak several languages might influence potential interviewees' motivation to participate, while monolingual approaches could exclude people from participating based on their (perceived) language-skills. This underlines the importance of reflecting upon multilingualism in all social science research. Calls for participation should be phrased in a suitable manner for the target group and possibly be translated. In the WeGe project, we relied on gatekeepers who presented our project during study preparation courses and also directly contacted students about participation based on the sampling criteria (gender, country of origin). While using gatekeepers' risks creating a sampling bias (e.g., "success cases"), the interview content did not suggest this was the case. If gatekeepers manage field access, researchers should be very clear about their approach to selecting participants. For example: Which language level should be considered necessary to participate in the study? [21]

In the DEMICO project, the team of translators was of great help when accessing the field and finding (potential) interview partners, as they posted the call for participation in relevant (and language-specific) groups on social media and even recruited participants at events. They also arranged some interview dates in consultation with the research team. This sometimes meant that they had already established a relationship with participants prior to the interviewers. On the one hand, this constituted valuable support; on the other hand, it meant that the interviewers had fewer opportunities to establish relationships and trust with the interviewees prior to the interview. [22]

Discussing the choice of interview language with prospective interviewees requires a reflective approach for ethical and methodological reasons: Gatekeepers, researchers, and interpreters must be aware of how languages are embedded in societal power relations, and that language practices and choices are related to language ideologies, prestige, and status. Offering different language options in an interview setting does not necessarily mean that they are perceived as equally acceptable and socially desirable by potential participants. In the DEMICO project, we continuously reflected upon and adapted how we discussed and chose language options with the participants. Considering everyday prejudices against migrants in public and media discourses (especially regarding the integration through language trope, DIRIM, 2021), we did not want to give interviewees the (false) impression that making an interpreter available for the interview implied they had poor English or German skills. On the other hand, we assumed that giving the interviewees the full authority for choosing a language might (in the context of racialized language hierarchies in migration societies) make some feel pressured to prove their English or German proficiency to us, when they would actually have preferred to speak in their first language during the interviews. Therefore, in our call for participation we stated that "our research design includes professional interpreting on site (Czech, Romanian or Hungarian language)." Additionally, during our first contact with potential interviewees, we explained that by involving interpreters, we were also interested in further developing a multilingual research approach. [23]

However, while attempts should be made to prevent interviewees from feeling pressured to prove their proficiency in German⁷⁾, it is equally important to respect them as competent speakers who can self-assess their language proficiency—and thus to accommodate their language preferences. In the WeGe project, some interviewees reported that the opportunity to practice German in the interview setting motivated them to participate. [24]

Also (but not exclusively), in regard to research ethics, thought should be given to the languages of the provided, relevant documents (e.g., project description, the informed consent form including information on the use of data, or further information on websites). If these documents are unavailable in the participants' first languages, their adaptation may be required to meet the participants' linguistic competence (beyond the always-necessary intra-language translation from a scholarly to a more colloquial linguistic register). In the WeGe project, we provided those documents in German and English. In addition to hiring a professional translator for the English versions, we asked gatekeepers who were familiar with courses and refugee students to proofread our material, such as project information and interview guidelines, and provide feedback on their linguistic suitability. In the DEMICO project, in line with our multilingual approach, we approached the field in five languages. However, we chose to provide certain information (regarding data handling, anonymization etc.) only in German and English because all interviewees had high reading proficiency in at least one of these languages.⁸⁾ [25]

In qualitative project designs, interview mode, method, and location have to be adapted to the research interest, the sampling population—and also the multilingual context as the project's language choice and linguistic requirements have a central impact on the methodological implementation. This connects to other matters, such as using interpreters to expand the linguistic scope, or using a lingua franca. [26]

We used our awareness strategy to provide orientation for our data collection in the WeGe project. Some of the central topics we considered included language choice, researcher positionality, the potential for re-traumatization, and specific experiences related to forced migration and asylum applications. The strategy was developed based on previous considerations during the project conceptualization, literature research (BERG et al., 2019b), and exchange with other researchers. [27]

We discussed the resulting methodological considerations in an expert workshop where external speakers were invited to give presentations on German as an interview language (THIEL & RIEDL, 2020), language and power relations, as well as (re)traumatization and social research (GRAEF-CALLIESS, 2018). Eventually, we gave interviewees the option of using German—the language of the host country—or English as a lingua franca. This was due to our expectation that participants would have high German proficiency and reflects how we positioned potential study participants as having a competent command of German and English. Additionally, the interviews focused on study preparation courses and experiences with higher education access in Germany. In those programs, administrative matters and course content were usually communicated in German. Most participants chose to do interviews in German. We concluded all pre-study interviews in the WeGe project by a set of meta questions, including questions about format and language choice. Overall, the interviewees gave positive feedback on using German as the interview language, while some interviewees preferred to receive paperwork in English or their first language. However, refugee students' self-assessed language skills and perceived ability to participate in an interview in English or German might have acted as an exclusive mechanism during the study. Finally, as expected, we received feedback that discussing an experience in a German-speaking environment was often easier to do in German, because doing so in a different language would require translations or new vocabulary. However, some interviewees also noted difficulties expressing their full intended meaning and included code-switching. In individual cases, their concentration span was notably exhausted. Both the German- and English-language interviews demonstrate that narrations in a foreign/second language might (but do not necessarily have to) be less comprehensible and coherent than in the interviewees' first languages, which—depending on the research focus and approach—might have implications for the later analysis. This was addressed in the WeGe project by re-inviting pre-study participants to a focus group where they could critically discuss the research team's interpretations and results. [28]

In the DEMICO project, we offered potential interviewees a further language option besides German or English, namely, to conduct the interview in their first language (Czech, Hungarian, or Romanian) with the help of an interpreter. Offering thus each prospective interviewee three options (English, German or interpreter-mediated) allowed us to integrate different language constellations—thereby not limiting our sample to certain levels of proficiency— and also enabled us to address and reflect on language choice (both in the context of interviewees' life worlds and in the research context). [29]

Our experience in this project revealed the consequences that each option has on both a practical and symbolic level: Conducting interviews in German (the dominant language in Austria, and the first language of the researchers, but not of the interviewees) meant an imbalance of language skills through which societal power relations were potentially reinforced and reproduced (for a discussion of how our own social positionings as white, Austrian-born, female academics influence the interview situation—and thus what is said—in the context of a racialized migration society, see SCHEIBELHOFER, HOLZINGER & DRAXL, 2023). Nevertheless, we believe that it would be patronizing and ethically questionable to deny interviewees the opportunity to speak German in the interview setting (see also Section 4.3). Overall, the DEMICO project has shown that conducting interviews in a common language—in English—offers a greater power balance regarding language proficiency. However, this option was only available when interviewees felt proficient enough in English to narrate their (migration) biography. [30]

Involving interpreters in the DEMICO project was arguably the best option for allowing interviewees to express themselves in their first language as freely and in as much detail as possible⁹⁾. Nevertheless, as discovered over the course of our research, this has several implications. First, having an interpreter present alters the relationship-building and dynamic in the interview setting: As previously described (HOLZINGER & DRAXL, 2024a), the interpreters were never invisible intermediaries, but actively involved participants in the conversation—thus influencing the emerging narrations by their presence (LAUTERBACH, 2014). Furthermore, interpreting affects the interview process itself, i.e., the interview length, narrative flow, and the participants' focus and attention span. We observed the relevance of the interpreting mode¹⁰⁾, because simultaneous and consecutive interpreting affect the interview process differently: Consecutively interpreted interviews take significantly longer and are characterized by waiting phases (because the interviewee and the interviewer have to wait for the interpreter to translate the previous passage) during which concentration sometimes wanes. [31]

Furthermore, it was a challenge for us interviewers not to be able to follow the narrations in real time, which meant that non-verbal signals such as nodding or smiling (which are characteristic of active listening to signal attention and interest) were usually delayed and sometimes no longer matched what was being said. However, we also noticed a narrative-generating interaction between narration and interpreting: Narrative pauses caused by interpreting phases were often used to subsequently provide details, additions, or corrections. Simultaneously interpreted interviews, on the other hand, can be challenging due to sensory overload, especially if the interviewees speak quickly and are interpreted without technical aids (consisting of microphones and headphones). While interruptions and waiting phases are less relevant in this mode of interpretation, the narrative flow and the attention span can be influenced here by distractions, which can be attributed to the interpreter speaking in parallel. [32]

Overall, we made the experience that an orientation towards the style of interpreting required in therapy settings can be beneficial (HOLZINGER & DRAXL, 2024a): In both settings, it is important to interpret as closely as possible to the original, since both the interviewee's and "the client's exact expression and the nuances of meaning in their choice of words represent valuable working material" (DABIĆ, 2021, p.274). [33]

After collecting qualitative interview data, they are typically processed from audio files to written text. This is often done externally, e.g., at transcription offices, by student assistants, or increasingly with AI software. In multilingual contexts, data that might be available in multiple languages (both in-between and within interviews), contain recordings of speakers with strong dialects, and/or include multiple speakers and ad hoc translations. If an interview was conducted in multiple languages or used words and phrases from more than one language, this should be correctly and accurately reflected in the transcription. [34]

Furthermore, this stage also creates opportunities for preparing data material in languages not mastered by the whole research group. Interviews conducted in the interviewee's first language (with or without the help of an interpreter) can subsequently be translated in full length by a qualified translator and made accessible to the entire team. Thereby, special attention must be paid to the specificities of the text type transcript which is not an objective representation of the primary verbal data, but as secondary data material itself a construction (KRUSE, 2015, pp.346-347). While the creation of transcriptions is always error-prone, multilingual contexts present specific additional challenges. [35]

First, the transcript quality must be ensured. Multilingual data reveal the need to address linguistic competencies for transcription as well as training for the required transcription style. Currently, this is beyond the scope of available AI-tools. In the WeGe project, we included experience with transcribing second language and multilingual interviews as selection criteria for professional transcription services. Further versions of such criteria could include experience with multilingual interviews, accents, or dialects. In the DEMICO project, our team of translators established contact with students from the Centre of Translation Studies (University of Vienna) who could handle the bilingual transcriptions of the interviews conducted with the support of an interpreter in Czech, Hungarian, or Romanian. [36]

Second, a transcript is not typically a text type that most translators are used to translating. This means that besides the translators being familiar with the transcription style, researchers and translators must also discuss the scope of the translation, i.e., its purpose, which is closely related to intended methods of analysis (see also ENZENHOFER & RESCH, 2019). Ideally, translators should have access to the audio files to consult in case of inconsistencies, ambivalences, and contradictions encountered within the transcript. It could also be of analytical interest to comment on noticeable elements resulting from the first language, such as translations of proverbs or common forms of expression. In the DEMICO project, the translators addressed difficulties with the text type interview transcript and translating conceptual orality (AGUADO, 2021) in written methodological memos that they sent us. We discussed these difficulties together in a reflection workshop with the translation team, emphasizing that features of conceptual orality should also be reflected upon as much as possible in the written translation. We found the translators' comments that pointed out translation difficulties (e.g., regarding the translation of proverbs or equivalences of school types), ambiguities (e.g., in regard to gender¹¹⁾), gaps, and background information to be extremely helpful for the analyses. For example, one of our interviewees talked about a friend whose parents had bought her an apartment, mentioning the name of the square in Budapest where it was located. While the interviewee added no further information to the name of the square, the translator inserted a short comment into the translation, stating that the square was located in a trendy and expensive neighborhood. This information was very helpful for our analysis of the passage which we interpreted as referring to social class and status loss through migration. Furthermore, this passage is interesting because we interpret the omission as related to the presence of the interpreter: As both the interpreter and the interviewee were from the same city, the interviewee probably implied a shared knowledge and thus did not feel the need to spell out what she associated with the address (e.g., class-related positioning). While the interpreters in the interview situation did not always have the time to clarify and explain ambiguities and information gaps, the written annotation allowed the translators to direct our attention to ambivalent passages and/or provide background information that could potentially be relevant for the analyses. [37]

As in all research projects, the chosen method of analysis determines the required data quality in the multilingual context. For example, if the analysis aims to exclusively capture the manifest content of a statement, fine-grain transcriptions are not always necessary. Likewise, less attention can be paid to issues regarding the translation of the interviewees' meaning-making. However, applying interpretative modes of analysis can introduce a range of issues regarding both the translation and the general translatability of the interviewees' narrations and processes of meaning-making. Researchers must scrutinize whether interpretative methods can be adequately applied if transcription quality, translation processes, and/or the proficiency in the interview language are considered. For interpreted interviews, their analyses require translations that also reproduce the orality of the interviewees' statements, including all 'errors' and ambiguities that are characteristic of conceptual orality (AGUADO, 2021) (e.g., slips of the tongue, repetitions, filler words). [38]

Comparing the two projects reveals an interesting finding in this regard: Given our participants' high German language proficiency in the WeGe project, the material was rich enough that re-constructive and interpretative methods of analysis could be applied without consulting translators. Further, we usually triangulated data from different perspectives (e.g., interviews with counsellors, refugee students, and teachers), and qualitative and quantitative project data to ensure the quality of our analysis and continuously reflected upon our results with other researchers and practitioners. As mentioned above, after analyzing the pre-study data, a focus group with refugee students was held to address our pre-study results and methodological features. Hence, the preliminary study was relevant both as an empirical and also as a methodological exploration. [39]

On the other hand, the DEMICO project mainly relied on interviews in multiple languages and followed an interpretative approach. Therefore, rather than only capturing the manifest content of a statement, the analysis was also intended to elucidate the underlying latent patterns of interpretation and orientations for action. From the onset, we thus attempted to minimize translation-related data quality limitations: We collaborated with formally qualified interpreters and translators who had basic knowledge of qualitative research, paid attention to fine-grained transcription quality, and based our analysis both on the transcribed on-site interpretation and the subsequent translation of the original statement¹²⁾. [40]

When comparing the translations of the transcribed statements with the on-site interpretations, we noticed that the interpreters often summarized and smoothed the interviewees' statements (with regard to their manifest content), i.e., they also omitted characteristics of conceptual orality (AGUADO, 2021), such as slips, errors, repetitions, and filler words. Such omissions and condensations are discussed as an integral part of simultaneous and consecutive interpreting in the literature (ANDRES, 2015; DIRIKER, 2015). In the beginning, we thus deemed the subsequent translation of the transcribed utterances in the first language as more suitable for the analysis (HOLZINGER & DRAXL, 2024a). We believed them to be closer to the original statement. However, as the project progressed and we discussed the data with our translation team during the methodological workshops, we learned to appreciate the added analytical value of the on-site interpreters: The interpreters often made stronger transmissions of para- and non-verbal interview features, including tone and atmosphere which are often lost in the transcription process. For example, when comparing our different versions of the data in a reflection workshop with the translation team, we noticed that the transcribed oral interpretation was significantly longer than the translation posteriori of the transcribed original statement. The interpreter present in the interview recalled vividly her attempts to transmit the overall atmosphere in the situation and to translate not only the words the interviewee used but also to convey how she (the interviewee) said what she said in order to get across the feeling of frustration she sensed from the part of the interviewee to the interviewer. Illustrating the analytical added value of multilingual research designs, our experience thus confirmed that comparisons between the on-site interpretations and the translations a posteriori can constitute an important heuristic device (INHETVEEN, 2012), The comparisons enable researchers to immerse themselves more deeply into the data by analyzing potential contradictions and tensions between the modes of translations¹³⁾. [41]

The dissemination of research findings in non-academic formats has received increased attention in contemporary academic and political debates (BUCCHI & TRENCH, 2021). In addition to academic publications, both project teams used various formats, such as blogposts, media articles, podcasts, and contributions to professional trainings, to transfer their findings to policy, practice, and the public. Additionally, researchers should consider the linguistic needs of their target audience; for example, choosing appropriate languages, linguistic registers, and formats for study results that will be made accessible to the participants and/or their communities. [42]

In both projects, we emphasized knowledge transfer activities that either addressed practitioners and/or a broader public in German and/or English. By publishing in these two languages, we made the findings available to an international audience and to interested study participants, who were all proficient in reading at least one of these two languages. Additionally, addressing how publications include interview quotes is also important when disseminating qualitative research in multilingual contexts, e.g., considering in which languages to publish them (in the language of the publication or in the original version including a translation) or whether quotations should be corrected for grammar and coherence. [43]

We made the multilingualism of our research from the DEMICO project visible by flagging translated passages in presentations and publications. This ensured making translation processes visible even when space limitations prevented us from printing quotes in more than one language. Likewise, we usually translated citations from the project as literally as possible in order to align with the publication format (English or German). [44]

In the previous sections, we have summarized our experiences in multilingual research projects and introduced general recommendations about how to adequately deal with multilingual challenges. However, achieving these methodological idea(l)s is often directly connected to available research funding and closely linked with personnel resources. Nevertheless, adequate funding goes beyond financial means to employ professional translators or interpreters. As shown above, multilingual research may require additional steps that address how work packages are planned. Examples include maintaining networks and organizing workshops with translators, developing methodological strategies, training research staff, as well as translating materials and data like informed consent forms, interview guidelines, or transcripts. Further, researchers should account for ample reflection upon methodological decisions and implications when developing the project timeline, which requires adequate resource (i.e., time and finances) allocation. Because most third-party funders have strict guidelines on eligible expenses, this has implications for researchers when they decide for which funding schemes to apply. This could have easily created problems for the WeGe project, i.e., if it had turned out in retrospect that interpreters would have been needed but could not be made available. In the case of the DEMICO project, it proved to be beneficial that the research had been conceptualized as a multilingual study from its onset: From the beginning, we had planned to rely on language mediation provided by certified interpreters and translators. Thus, we were able to calculate the budget accordingly, which was granted in its entirety by the funding authority. [45]

Planning and creating space for methodological reflection were key elements of our approaches for both research projects. This was explicitly the case during the initial phase of the WeGe project, where we developed an awareness strategy and discussed pre-study results and empirical considerations with experts (e.g., practitioners), other researchers, and research participants. As mentioned above, the WeGe team initially developed an awareness strategy for conducting qualitative and quantitative research with refugee students and held an internal project workshop with other researchers and practitioners to address topics such as language selection, power imbalances, (re-)traumatization, as well as interviewee and project team member well-being. [46]

The cyclical conceptualization of grounded theory methodology was an important starting point for methodological reflection in the DEMICO project, as were recurring reflection workshops with the translation team. These led to methodological adjustments at every stage of the project. Unlike the WeGe project, the DEMICO research team did not initially develop an awareness strategy for research-ethical aspects of the participants' vulnerability. However, our first interview experiences underlined that reflecting upon the positionalities and positionings in the interview setting is of crucial importance. We did this both with colleagues at conferences and in a methodological paper on the positioning strategies and power asymmetries in the interview setting (SCHEIBELHOFER et al., 2023). Thus, issues regarding vulnerability as well as other ethical aspects were addressed more thoroughly in the course of the research project. However, language-related issues received in-depth consideration from the project's outset. Besides the preparatory kick-off workshop, we also held regular methodological workshops¹⁴⁾ with the translators to continuously evaluate and discuss the methodological and ethical dimensions of our interdisciplinary collaboration. Likewise, interviews that were conducted with interpreter support were concluded with a meta question about how the interviewee perceived this specific interview setting. Additionally, after each interview, we briefly discussed the interview situation with the interpreters and asked them to send us a brief summary of their impressions as soon as possible after each interview. [47]

A further way to plan and create spaces for methodological and ethical reflection is to exchange ideas with researchers from other projects, e.g., at (interdisciplinary) conferences or through publication projects and talks that present and critically discuss language-related aspects of research designs—or through collaborations such as the one that led to this publication. In both projects, our methodological reflection closely overlapped with guiding conceptual considerations in three interconnected thematic areas: Analyzing forms of in- and exclusion arising from different approaches (e.g., through language choices), scrutinizing power relations and positionings between research participants (both researchers and researched), and reflecting upon risks to data quality. [48]

Conducting qualitative research projects presents various opportunities and needs for methodological and ethical reflection. This ranges from project design, constructing shared meaning in data collection, reflection and collaboration in the project team, to decisions about target group-oriented dissemination. Although methodological considerations regarding language choice and translation are mostly discussed in the context of migration research and international research projects, they are particularly relevant to any qualitative research in post-migrant societies where researchers and interviewees speak different first languages. [49]

The exchange of experiences between the two project team members exemplified that there is no one-size-fits-all approach to multilingual research. Instead, project-specific solutions should be tailored to the research question, participants, and available resources. Issues like inclusion and exclusion, positionality, and positionings, as well as impacts on data quality, should all be considered from the project planning stage to its implementation. This includes reflecting upon which perspectives are required to comprehensively investigate the research question, addressing power differences and dynamics that might occur throughout the research project, and considering the implications for data quality. [50]

Based on the joint discussions of the two research teams, we attempted to formulate central insights that might be of use for future research in polyglot contexts. Most importantly, we recommend focusing on the following three overlapping thematic areas as initial guidance to identify as well as address both needs and opportunities for reflection. [51]

As a starting point, all communication includes and excludes something. This assumption is vital to qualitative research in a multilingual society when considering interview language(s). By making proficiency, e.g., in the majority language or in a lingua franca, a prerequisite for participation, researchers create criteria for inclusion and exclusion. As mentioned above, we expected a high German proficiency of the WeGe projects participants. However, there is a risk of students' self-exclusion based on their perceived language skills, which could be a limitation of this approach. Furthermore, certain experiences and narratives may only be available and narratable by research participants in a specific language in spite of their proficiency. Therefore, following a monolinguist approach could exclude certain social groups (or at least identity components of multilingual individuals) from research, and consequently certain perspectives in knowledge production processes. Thus, when making decisions about interview languages and professional translation, researchers should always ask whose perspectives must be included in order to answer a given research question—and which language-related arrangements are likewise necessary. Researchers must scrutinize whose voice will be excluded by the defined interview language(s) and/or the non-provision of interpreters. [52]

Furthermore, in both projects, it became critically evident that while people can generally express themselves in greater detail and depth using their first language, the participants' life-world multilingualism must be taken into account; for example, interviewees might be more used to—and therefore more comfortable—discussing job searches, courses, and working experiences from Germany or Austria in German or English because these are the languages of their everyday (working) life (HOLZINGER & DRAXL, 2024b). Some WeGe participants reported that it was easier to describe certain aspects related to their courses in German, which was typically the language spoken in this context. This was supported by observing interviewees from the DEMICO project code-mixing or code-switching, i.e., they switched (word-wise or even in passages) from one language to another (if our language skills or an interpreter made understanding possible) depending on the content and location of what they described to us. [53]

In conclusion, when developing research designs, academics should ideally always consider the interviewees' multilingualism, enable translingual practices, and allow them to draw on most of their linguistic repertoire (see also HAVLIN, 2022; UÇAN, 2019, 2021). This extends beyond data collection. The WeGe study participants preferred to be interviewed in German and some of them pointed to the opportunity to practice the language as a key motivator for participation. However, some preferred to receive paperwork, such as project information and informed consent forms, in English or their first language. Thus, trans- and multilingual practices can not only support the interview situation, but language is a tool that should be considered throughout project-design. As mentioned before, such practices need to be accompanied by appropriate strategies of documentation and data processing, which need to be considered early on in resource planning and can be restricted by practical limitations. [54]

Verbal communication functions as a mode of communication that uses language to co-construct mutual understanding, but also to negotiate social positions and power relations on a symbolic level (BOURDIEU, 1982). When speaking, we always position ourselves and others (HARRÈ & VAN LANGENHOVE, 1991). From the perspective of HARRÈ´s positioning theory, moral positionings that negotiate who speaks to whom and in what relationship to each other are inevitable in any conversation (DEPPERMANN, 2013b). In multilingual contexts, additional dimensions regarding the negotiation of social positions and relationship building should be considered (SCHEIBELHOFER et al., 2023). [55]

Research design and its implementation incorporate assumptions about the social groups being researched, as exemplified by debates about using interpreters in forced migration research. These, too, express positionings and imply presuppositions about language use and language skills. In the research process, positionings take place within the research team (associated with responsibilities) and between researchers and participants. In particular, researchers often hold a privileged position in migration-related, multilingual contexts (ibid.). This includes, but often exceeds hegemonial status of nationality, but could also refer to first- and second-language speakers or the status of academics vs. prospective students, as in the WeGe project. This has implications about the extent to which the choice of interview language and the decision to use interpreters or not interferes with or reproduces social power relations in the respective research context (THIELEN, 2009). Which positionings are desirable or to be avoided—and what does it mean to choose a certain language in this respect? [56]

These reflections are of special importance when conducting research in contexts characterized by racialized social hierarchies which is the case in migration societies like Austria or Germany. Positionalities have important implications when discussing sensitive topics like discrimination with interviewees—as in the DEMICO project—and must be critically and continuously reflected upon. [57]

Additionally, when relying on language mediation, the "classical" interview setting is extended by a third person—namely, the interpreter—which multiplies and alters the positionings within the conversation: "The interview partner has to position herself/himself then not only in relation to one but two interlocutors (researcher and interpreter)—and the respective positioning strategies might be difficult to reconcile" (SCHEIBELHOFER et al., 2023, p.83). [58]

How researchers approach the multilingualism of a given field (e.g., the selection of the interview language) has implications for data quality, which may be accompanied by specific risks of bias. In light of the inclusion and exclusion argument above, bias can occur when certain groups of participants or relevant experiences are systematically excluded due to language barriers or (perceived) language literacy. This may create blind spots, lead to the under- or non-representation of certain voices, and therefore create considerable limitations to a study. Additionally, language choice as well as the researchers', interpreters', and the participants' social positionalities influence what can be said in an interview setting. [59]

The required data quality is partly contingent on the research context, the research interest, and the chosen method of data analysis. Translations from one language to another can either be seen as a source of bias and loss of meaning, or as an additional interpretation that can be utilized during data analysis. We strongly advocate the latter: Research teams should give special attention to how consciously engaging with multilingualism can enrich empirical work (BOCHMANN, 2023; HAVLIN, 2022; MIJIĆ, 2025; REICHERTZ, 2020; SCHITTENHELM, 2017; TEMPLE & EDWARDS, 2002). Translators can support the identification of implicit meaning, reference to figures of speech, or common knowledge among first language speakers, such as the aforementioned socio-cultural situatedness of a city square, as in the DEMICO project. Multilingualism underlines the usefulness of the classic guiding principles openness, reflexivity, and pragmatism (ERHARD, JUKSCHAT & SAMMET, 2021) as well as the importance of intersubjectivity and object reference as quality criteria in interview research (SCHITTENHELM, 2017). It also demonstrates that, despite many researchers' reluctance to relinquish control, quality depends on seeking, allowing, and reflecting upon cooperation (KRUSE et al., 2012, p.36). [60]

In this paper, we have identified multiple occasions for reflection upon and planning throughout the research process in polyglot contexts. We have also shown fundamental issues to consider when designing qualitative research projects in order to provide guidance for developing specific solutions for the respective project context. We argue that it is essential to understand methodological and ethical reflexivity as interrelated. Drawing on VON UNGER´s (2021, p.200) work on ethical reflexivity "as an ongoing practice," we conclude that language-related considerations in multilingual empirical research "can best be met in dialogue with scientific peers, key informants, and partners in the field." Hence, "[p]positionalities must be reflected vis-à-vis the field and the subject matter, not only in scientific terms (that is, on methodological grounds), but also on moral grounds regarding the smaller and larger implications of research practice" (Ibid.). [61]

The project "Pathways of Refugees to German Universities" (WeGe) was funded by the German Federal Ministry of Education and Research under the grant number 01PX16015. The DEMICO research project was funded by the Austrian Science Fund (FWF) under the grant number P 33633‐G.

We thank our respective project teams, all research participants, the anonymous reviewers and editors for their helpful comments on an earlier draft of the manuscript, Leora COURTNEY-WOLFMAN for her thorough proofreading as well as the Faculty of Social Sciences of the University of Vienna for financial support for proofreading.

2) "The CEFR […] has developed a description of the process of mastering an unknown language by type of competence and sub-competence, using descriptors for each competence or sub-competence […] The descriptors specify progressive mastery of each skill, which is graded on a six-level scale (A1, A2, B1, B2, C1, C2)" (COUNCIL OF EUROPE, n.y.). <back>

3) Members of the translation team were Ladislava BAXANT-CEJNAR, Cinzia HIRSCHVOGL, Michaela KUKLOVÁ, Anna LEDÓ, Corina NIŢU and Izabella NYARI. <back>

4) If the interpretation mode was simultaneous, we used two recording devices (one with a headset worn by the interpreter and one for the room) to record both speakers and "disentangle" the parallel voices. <back>

5) The translations were always provided by the person from the respective 'language tandem' who had not interpreted the interview. <back>

6) The latter not being understood as essential but as fluid, and in need of continuous social interpretation. <back>

7) It remains contestable whether this is feasible in a context of racialized language hierarchies. <back>

8) We disseminated calls for interview participation in Czech, English, German, Hungarian, and Romanian. While the project website contains information in English and German for a general audience, interviewees also received an informative handout at the beginning of the interview that was specifically directed towards them. It provided information about the broader research focus, data handling, and the interviewees' rights concerning data protection. We chose to provide this information in German and English because all interviewees had high reading proficiency in at least one of these languages. We explained the points covered in the handout at the beginning of each interview; if an interpreter was present, this information was also made (orally) available in the respective language <back>

9) While interviewees mostly had excellent German or English language skills, some nevertheless expressed their appreciation for being able to conduct the interview in their first language via an interpreter. They explained that they either lacked a sufficient German vocabulary or that speaking in their first language relieved them of having to think too much about "grammatically correct" speech during the interviews. Other interviewees, however, declined the option of having an interpreter. <back>

10) We differentiate here between consecutive and simultaneous mode. Consecutive interpreting commonly refers to the rendering of source-language utterances after someone has spoken in another language, prototypically implying (monologic) speeches with a duration of five minutes or longer (ANDRES, 2015). On the other hand, simultaneous interpreting is the mode of interpreting where the interpreter renders the speech as it is being delivered by a speaker into another language with a minimal time lag of a few seconds (DIRIKER, 2015, pp.382-383). <back>

11) For example, while in German, most professions have both masculine and feminine forms, this is not the case in all source languages included in the DEMICO project. <back>

12) For the interpreted interviews, we had three written versions of the interviewees' utterances as the basis for our analysis: The transcription of the original statement in the first language of the interviewee, the transcription of the interpretation on-site, and the subsequent translation of the transcribed original statement. <back>

13) Additionally, if a passage that we deemed important for the analysis remained ambiguous or unintelligible, or if we discovered differences between the transcribed interpretation and the subsequent translation, we consulted the transcribed passage in the original language (and occasionally even the recording), which we translated again with the help of machine translation or online dictionaries—drawing e.g., on our own limited knowledge of Czech and Romanian. <back>

14) One full-day and three half-day workshops were organized during the four-year research project. <back>

Aguado, Karin (2021). Sprachliche Teilkompetenzen (1): Mündlichkeit. In Claus Altmayer, Katrin Biebighäuser, Stefanie Haberzettl, & Antje Heine (Eds.), Handbuch Deutsch als Fremd- und Zweitsprache (pp.253-267). Berlin: J.B. Metzler.

Andres, Dörte (2015). Consecutive interpreting. In Franz Pöchhacker, Nadja Grbić, Peter Mead & Robin Setton (Eds.), Routledge encyclopaedia of interpreting studies (pp.84-87). London: Routledge.

Banerhee, Paulami & Sowards, Stacey K. (2022). Working across languages/ cultures in international and environmental communication fieldwork. Journal of International and Intercultural Communication, 15(1), 36-56.

Bading, Cornelia; Kazzazi, Kerstin & Wintzer, Jeannine (Eds.) (2025). (Fremd-)Sprache und Qualitative Sozialforschung. Wiesbaden: Springer Spectrum.

Berg, Jana; Grüttner, Michael & Schröder, Stefanie (2019a). Entwicklung und Anwendung eines Sensibilisierungskonzeptes für qualitative Interviews mit Geflüchteten – Erfahrungen im Projekt WeGe. In Birgit Behrensen & Manuela Westphal (Eds.) Fluchtmigrationsforschung im Aufbruch. Methodologische und methodische Reflexionen (S. 275-300). Wiesbaden: Springer VS.

Berg, Jana; Grüttner, Michael & Schröder, Stefanie (2019b). Zwischen Partizipation und Retraumatisierung – (Selbst-)kritische Überlegungen zur Forschung mit Geflüchteten. In Nicole Burzan (Ed.), Komplexe Dynamiken globaler und lokaler Entwicklungen. Verhandlungen des 39. Kongresses der Deutschen Gesellschaft für Soziologie in Göttingen 2018, https://doi.org/10.21241/ssoar.100909 [Accessed: September 2, 2025]

Berger, Yvonne (2023). Sprachliche und kulturelle Übersetzungsprozesse in der Forschungspraxis–Herausforderungen im Kontext qualitativer Forschung. In Jasmin Donlic (Ed.), Qualitative Methoden in der Forschungspraxis: Perspektiven, Erfahrungen und Anwendungsfelder (pp.138-162). Opladen: Verlag Barbara Budrich.

Brämberg, Elisabeth Björk & Dahlberg, Karin (2013). Interpreters in cross-cultural interviews: A three-way coconstruction of data. Qualitative Health Research, 23(2), 241-247.

Bochmann, Annett (2023). The sociology and practice of translation: interaction, indexicality, and power. Qualitative Research, 23(6), 1781-1799, https://doi.org/10.1177/14687941221124736 [Accessed September 2, 2025].

Bogner, Artur & Rosenthal, Gabriele (2014). The "untold" stories of outsiders and their significance for the analysis of (post-) conflict figurations. interviews with victims of collective violence in Northern Uganda (West Nile). Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 15(3), Art. 4, https://doi.org/10.17169/fqs-15.3.2138 [Accessed: December 19, 2024].

Bourdieu, Pierre (1982). Ce que parler veut dire. L'économie des échanges linguistiques [What it means to speak. The economics of linguistic exchanche]. Paris: Fayard.

Bucchi, Massimiano & Trench, Brian (Eds.) (2021). Routledge handbook of public communication of science and technology. London: Routledge.

Burkhard, Tanja & Park, Su Jin (2019). Translating (in) the margins: The dilemmas, ethics, and politics of a transnational feminist approach to translating in multilingual qualitative research. Qualitative Research, 24(3), 486-504.

Can, Halil (2020). Multilinguale multi-sited ethnography. In Angela Treiber, Kerstin Kazzazi & Marina Jaciuk (Eds.), Migration Übersetzen. Alltags- und Forschungspraktiken des Dolmetschens im Rahmen von Flucht und Migration (pp.23-44). Wiesbaden: Springer VS.

Dabić, Mascha (2021). Dolmetschen in der Psychotherapie: prekäres Gleichgewicht. Tübingen: Narr Francke Attempto.

Dausien, Bettina, & Thoma, Nadja (2023). "Small stories" as methodological approach. Reflections on biographical narratives based on an ethnographic research project with refugee school students. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 24(1), Art. 3, https://doi.org/10.17169/fqs-24.1.3784 [Accessed: December 19, 2024].

Deppermann, Arnulf (2013a). Interview als Text vs. Interview als Interaktion. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(3), Art. 13, https://doi.org/10.17169/fqs-14.3.2064 [Accessed: September 2, 2025).

Deppermann Arnulf (2013b). Editorial: Positioning in narrative interaction. In Michael Bamberg (Ed.), Narrative inquiry (pp.1-15). Amsterdam: Benjamins.

Diriker, Ebru (2015). Simultaneous interpreting. In Franz Pöchhacker, Nadja Grbić, Peter Mead & Robin Setton (Eds.), Routledge encyclopedia of interpreting studies (pp.382-385). London: Routledge.

Dirim, Inci (2021). Sprache und Integration. In Claus Altmayer, Katrin Biebighäuser, Stefanie Haberzettl & Antje Heine (Eds.), Handbuch Deutsch als Fremd- und Zweitsprache. Kontexte – Themen – Methoden (pp.88-101). Stuttgart: J.B. Metzler.

Enzenhofer, Edith & Resch, Katharina (2019). Unsichtbare Übersetzung? Die Bedeutung der Übersetzungsqualität für das Fremdverstehen in der qualitativen Sozialforschung. In Michael Roslon & Richard Bettmann (Eds.), Interkulturelle Qualitative Sozialforschung (pp.199-225). Wiesbaden: Springer VS.

Erhard, Franz; Jukschat, Nadine & Sammet, Kornelia (2021) Lost in translation? Openness, reflexivity and pragmatism as guiding principles for cross-language qualitative research. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 22(3), Art. 19, https://doi.org/10.17169/fqs-22.3.3722 [Accessed: May 26, 2025].

Flick, Uwe (2000). Episodic interviewing. Qualitative researching with text, image and sound. In Martin W. Bauer & George Gaskell (Eds.), Qualitative researching with text, image and sound: A practical handbook for social research (pp.75-92). London: Sage.

Fritsche, Andrea (2016). Kultur(en) und Sprache(n) der Asylwirklichkeit – Herausforderungen empirischer Forschung im Kontext von Unsicherheit, Verrechtlichung, Interkulturalität und Mehrsprachigkeit. Österreichische Zeitschrift für Soziologie, 41(S2), 165-190, https://doi.org/10.1007/s11614-016-0227-5 [Accessed: December 19, 2024].

García, Ofelia & Kleifgen, Jo Anne (2020). Translanguaging and literacies. Reading Research Quarterly, 55(4), 553-571.

Goitom, Mary (2020). Multilingual research: Reflections on translating qualitative data. The British Journal of Social Work, 50(2), 548-564.

Graef-Calliess, Irisi Tatjana (2018). Migration und Transkulturalität: Neue Aufgaben in Psychiatrie und Psychotherapie. Stuttgart: Klett-Cotta.

Grüttner, Michael; Schröder, Stefanie; Berg, Jana & Otto, Carolin (2021). Erfolgreiche Studienvorbereitung? – Ergebnisse einer Mixed-Methods-Studie zum Erfolg von Geflüchteten in der Studienvorbereitung. In Martin Neugebauer, Hans-Dieter Daniel & Andrä Wolter (Eds.), Studienerfolg und Studienabbruch (pp.235-258). Wiesbaden: Springer VS.

Harré, Rom & Van Langenhove, Luk (1991). Varieties of positioning. Journal for the Theory of Social Behaviour, 21(4), 393-407.

Haug, Sonja; Lochner, Susanne & Huber, Dominik (2017). Methodische Herausforderungen bei der quantitativen und qualitativen Datenerhebung bei Geflüchteten. Beitrag zur Ad-Hoc-Gruppe "Geflüchtete als Herausforderung für die empirische Sozialforschung". In Stephan Lessenich (Ed.), Geschlossene Gesellschaften. Verhandlungen des 38. Kongresses der Deutschen Gesellschaft für Soziologie in Bamberg 2016 (pp.1-11), https://publikationen.soziologie.de/index.php/kongressband_2016/de/article/view/402/pdf_107 [Accessed: December 19, 2024].

Havlin, Tetiana (2022). Multilingualism and translanguaging in migration studies: Some methodological reflections. Forum Qualitative Sozialforschung / Forum Qualitative Social Research, 23(1), Art. 2, https://doi.org/10.17169/fqs-23.1.3730 [Accessed: December 29, 2024].

Holmes, Prue; Reynolds, Judith & Ganassin, Sara (2022). Introduction: The imperative for the politics of "researching multilingually". In Prue Holmes, Judith Reynolds & Sara Ganassin (Eds.), The politics of researching multilingually (pp.1-28). Bristol: DeGruyter/Multilingual Matters.

Holzinger, Clara & Draxl, Anna-Katharina (2024a). Sozialwissenschaftliche Forschung mit Dolmetscher*innen – Praktische Erfahrungen und methodische Reflexion. In Katia Iacono, Barbara Heinisch & Sonja Pöllabauer (Eds.), Zwischenstationen/Inbetween: Kommunikation mit geflüchteten Menschen/Communicating with Refugees (pp.219-291). Berlin: Frank & Timme, https://www.frank-timme.de/de/programm/produkt/zwischenstationen_inbetween?file=/site/assets/files/6928/9783732990009.pdf [Accessed: August 25, 2025].

Holzinger, Clara & Draxl, Anna-Katharina (2024b). A racism-critical look at language-related discrimination in access to welfare state benefits and the labor market. Momentum Quarterly, 13(1), 16-32, https://doi.org/10.15203/momentumquarterly.vol13.no1.p16-32 [Accessed: August 25, 2025].

Holzinger, Clara & Draxl, Anna-Katharina (Eds.) (forthcoming). Multilingual challenges: Empirical social research in migration societies, transnational spaces, and international contexts. Social Inclusion, https://www.cogitatiopress.com/socialinclusion/issue/futureissues#i535 [Accessed: August 25, 2025].

Koskela, Merja; Koskinen, Kaisa & Pilke, Nina (2017). Bilingual formal meeting as a context of translatoriality. Target: International Journal of Translation Studies, 29(3), 464-485.

Kruse, Jan (2015). Qualitative Interviewforschung. Ein integrativer Ansatz. Weinheim: Beltz Juventa.

Kruse, Jan; Bethmann, Stephanie; Niermann, Debora & Schmieder, Christian (2012). In und mit fremden Sprachen forschen. Eine empirische Bestandsaufnahme zu Erfahrungs- und Handlungswissen von Forschenden. In Jan Kruse, Stephanie Bethmann, Debora Niermann & Christian Schmieder (Eds.), Qualitative Interviewforschung in und mit fremden Sprachen. Eine Einführung in Theorie und Praxis (pp.27-68). Weinheim: Beltz Juventa.

Kusters, Annelies & De Meulder, Maartje (2019). Language portraits: Investigating embodied multilingual and multimodal repertoires. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 20(3), Art. 10, https://doi.org/10.17169/fqs-20.3.3239 [Accessed: December 19, 2024].

Larkin, Philip J.; Dierckx de Casterlé, Bernadette & Schotsmans, Paul (2007). Multilingual translation issues in qualitative research: Reflections on a metaphorical process. Qualitative Health Research, 17(4), 468-476.

Lauterbach, Gwendolin (2014). Interpreter-mediation in qualitative interviews: About the requirements and effects. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 15(2), Art. 5, https://doi.org/10.17169/fqs-15.2.2025 [Accessed: December 19, 2024].

Littig, Beate & Pöchhacker, Franz (2014). Socio-translational collaboration in qualitative inquiry: The case of expert interviews. Qualitative Inquiry, 20(9), 1085-1095.

MacSwan, Jeff (2017). A multilingual persperctive on translanguaging. Academic Educational Research Journal, 54(1), 167-201.

Mijić, Ana (2025). Navigating the boundaries of understanding: Methodological reflections on cross-cultural challenges in reconstructive social research. International Journal of Qualitative Methods, 24, https://journals.sagepub.com/doi/full/10.1177/16094069251337002 [Accessed: August 25, 2025].

Reckinger, Gilles (2020). Die sprachlichen Spuren der Migration. Zum Umgang mit Multi- und Translingualität in der ethnografischen Forschung mit Geflüchteten. In Angela Treiber, Kerstin Kazzazi & Marina Jaciuk (Eds.), Migration Übersetzen. Alltags- und Forschungspraktiken des Dolmetschens im Rahmen von Flucht und Migration (pp.13-22). Wiesbaden: Springer VS.

Reichertz, Jo (2020). Grenzen der Interpretation. Wie kann und soll man in der qualitativen Sozialforschung mit interkulturellen Daten umgehen?. In Halyna Leontiy & Miklas Schulz (Eds.), Ethnographie und Diversität. Wissensproduktion an den Grenzen und die Grenzen der Wissensproduktion. Wiesbaden: Springer VS.

Scheibelhofer, Elisabeth (2023). The interpretive interview. An interview form centring on research participants' constructions. International Journal of Qualitative Methods, 22, 1-8, https://doi.org/10.1177/16094069231168748 [Accessed: August 25, 2025].

Scheibelhofer, Elisabeth; Holzinger, Clara & Draxl, Anna-Katharina (2023). Confronting racialised power asymmetries in the interview setting: Positioning strategies of highly qualified migrants. Social Inclusion, 11(2), 80-89, https://doi.org/10.17645/si.v11i2.6468 [Accessed: August 25, 2025].

Schittenhelm, Karin (2017). Mehrsprachigkeit als methodische Herausforderung in transnationalen Forschungskontexten. Zeitschrift für Qualitative Forschung, 18(1), 101-115, https://doi.org/10.3224/zqf.v18i1.08 [Accessed: August 25, 2025].

Schoonenboom, Judith & Johnson, R. Burke (2017). How to construct a mixed methods research design. Kölner Zeitschrift fur Soziologie und Sozialpsychologie, 69(Suppl 2), 107-131, https://doi.org/10.1007/s11577-017-0454-1 [Accessed: February 27, 2025].

Schröder, Stefanie; Grüttner, Michael & Berg, Jana (2019). Study preparation for refugees in German "Studienkollegs"—interpretative patterns of access, life-wide (language) learning and performance. Widening Participation and Lifelong Learning, 21(2), 67-85, https://doi.org/10.5456/WPLL.21.2.67 [Accessed: August 25, 2025].

Sen, Amartya (1985). Well-being, agency and freedom: The Dewey lectures 1984. The Journal of Philosophy, 82(4), 169-221.

Sen, Amartya (1999). Development as freedom. New York, NY: Oxford University Press.

Sheridan, Vera & Storch, Katharina (2009). Linking the intercultural and grounded theory: Methodological issues in migration research. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(1), Art. 36, https://doi.org/10.17169/fqs-10.1.1217 [Accessed: December 19, 2024].

Temple, Bogusia & Edwards, Rosalind (2002). Interpreters/translators and cross-language research: Reflexivity and border crossings. International Journal of Qualitative Methods, 1(2), 1-12, https://doi.org/10.1177/160940690200100201 [Accessed: February 27, 2025].

Temple, Bogusia; Edwards, Rosalind & Alexander, Claire (2006). Grasping at context: Cross language qualitative research as secondary qualitative data analysis. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(4), Art. 10, https://doi.org/10.17169/fqs-7.4.176 [Accessed: December 19, 2024].

Thiel, Barbara & Riedl, Alfred (2020). Forschung mit Geflüchteten – methodische Erfahrungen aus zwei Projekten. In Britta Matthes & Eckard Severing (Eds.), Zugang zu beruflicher Bildung für Zuwandernde – Chancen und Barrieren (S. 48-64). Bonn: BIBB, https://www.agbfn.de/de/agbfn_152646_149685.php [Accessed: August 25, 2025].

Thielen, Marc (2009). Freies Erzählen im totalen Raum? – Machtprozeduren des Asylverfahrens in ihrer Bedeutung für biografische Interviews mit Flüchtlingen. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(1), Art. 39, https://doi.org/10.17169/fqs-10.1.1223 [Accessed: August 25, 2025].

Treiber, Angela & Kazzazi, Kerstin (2020). Migration Übersetzen. Methodische Zugänge, epistemologische Fragen, theoretische Ansätze. In Angela Treiber, Kerstin Kazzazi & Marina Jaciuk (Eds.), Migration Übersetzen. Alltags- und Forschungspraktiken des Dolmetschens im Rahmen von Flucht und Migration (pp.1-12). Wiesbaden: Springer VS.

Uçan, Yasemin (2019). Sprachen und Sprechen in der qualitativen Migrations- und Fluchtforschung. In Birgit Behrensen & Manuela Westphal (Eds.), Fluchtmigrationsforschung im Aufbruch: Methodologische und methodische Reflexionen (pp.115-139). Wiesbaden: Springer VS.

Uçan, Yasemin (2021). Potenziale einer translingualen Interviewführung am Beispiel erziehungswissenschaftlicher Migrationsforschung. In Angela Treiber, Kerstin Kazzazi & Marina Jaciuk (Eds.), Migration Übersetzen. Alltags- und Forschungspraktiken des Dolmetschens im Rahmen von Flucht und Migration (pp.71-87). Wiesbaden: Springer VS.

Vertovec, Steve (2023). Superdiversity. Migration and social complexity. Abingdon: Routledge.

Jana BERG is a senior researcher at the University of Applied Sciences Technikum Wien and conducts transdisciplinary research on participatory energy transition, climate communication, and international higher education. She steers the social science research area in an interdisciplinary team and leads several empirical research projects. In addition, she teaches qualitative methods and academic writing at the University of Vienna. She received her Dr. phil in sociology at Leibniz University Hannover in 2021 and studied sociology and linguistics (B.A.) at the Universities of Mainz and Vienna. She has been actively involved in several national and international empirical research projects and has completed several international visiting scholarships, e.g., in Belgium, the USA, and South Africa.

Clara HOLZINGER is a post-doctoral researcher and lecturer at the University of Vienna, Department of Sociology, and is part of the INMI research group. She has a disciplinary background in sociology (PhD), German as a foreign and second language (MA) and history (BA). She has been involved in various national and international empirical research projects and recently received an ESPRIT grant of the Austrian Science (FWF). In her research, she focuses on qualitative methods, migration, racism and multilingualism.

University of Vienna, Department of Sociology
Rooseveltplatz 2, 1090 Vienna, Austria

Michael GRÜTTNER is senior researcher at the German Centre for Higher Education Research and Science Studies (DZHW) in Hannover. He received his Dr. in sociology at Leibniz University Hannover in 2021. In his research, he focuses on participation of refugees and asylum seekers in higher education, continuing and adult education. He is trained in qualitative, quantitative and mixed methods.

German Centre for Higher Education Research and Science Studies (DZHW)
Lange Laube 12, 30159 Hannover, Germany

Anna-Katharina DRAXL is a senior researcher at the Department of Sociology, University of Vienna, Austria. With a background in cultural and social anthropology (MA) and German as a foreign and second language (MA), she has experience in various interdisciplinary research projects in educational and social sciences. In her research, she focuses on migration and language(s) in the fields of education, work and language policy as well as multilingual contexts. She is also working as a lecturer for German as a foreign and second language at the Academy of Fine Arts, Vienna.

University of Vienna, Department of Sociology
Rooseveltplatz 2, 1090 Vienna, Austria

Berg, Jana; Holzinger, Clara; Grüttner, Michael & Draxl, Anna-Katharina (2025). Qualitative interview research in multilingual contexts—A comparative discussion of language-related decisions in two empirical studies [61 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 26(3), Art. 15, https://doi.org/10.17169/fqs-26.3.4367.

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

ISSN 1438-5627

Funded by the KOALA project

Creative Commons Attribution 4.0 International License