Cultural perspective in open-ended interviews – The importance of being adaptable

Psychology as a subject is often investigated from a so-called objectivistic standpoint, implying a belief that the researcher should try to eliminate the subjectivity in the research material to find the relevant information (Kim, 1999). This is done with good intentions, to make research resistant against personal and cultural biases, but the problem with investigating people is that people necessarily come with their own set of preconceptions. To be a person is equivalent to existing according to a pre-understanding, and by eliminating pre-understandings the researcher is effectively trying to root out the social from the social sciences. This is not only true for the respondents in a study, but for the researcher too, who’s preconceptions, cultural background, attitudes, etc. will influence respondents, showing the dialogical character of the social sciences. This subjectivity or pre-comprehension might produce ecologically invalid data. This risk is even more pronounced when the researcher and the respondent are from two different cultures. Even in qualitative research where the researchers’ preconceptions constitute a calculated part of the process, the cultural differences in interpreting the interview situation (or any data gathering situation for that matter) run the risk of becoming an ‘uncalculated’ part of the preconception, and thus the results. Therefore, intercultural research comes with a special set of problems, and as we cannot, or should not, eliminate the social from the social sciences we need to understand these problems and handle them in an adequate manner so that we can successfully compare and interpret respondents from different backgrounds and cultures. We need to use methods where the data we collect are ‘comparable’ and this ‘comparability’, as we shall see, does not necessarily mean using the same method. In fact, it often makes it necessary to apply different methods in different cultures, as the method itself can carry the researchers’ conceptually based preconceptions, and thus be unsuitable, or give a different type of narratives/data when respondents from different cultures interpret the situation differently. When using an open-ended interview in one context, the meaning of the situation might differ in another and that, in turn, suggests that the meaning, interpretation, quality and structure of the narratives will vary and, since we ought not to take the social out of any social science, this has to be taken into consideration when using open-ended interviews as a data-gathering tool. Sometimes it might even be beneficial to change the method slightly or altogether. Naturally the intercultural aspect of the open-ended interview is also an issue of concern to studies staying within one nation’s borders, as people from the same national ethnicity do not necessarily share the same culture; culture might differ depending on age, gender, socio-economic background, etc. (Frederiksen, 1977; Valsiner, 2000).

Most intercultural researchers are aware of the problems one might encounter when conducting interviews in an intercultural setting. For example you might require a translator or use a language you or your respondent has not fully mastered. This might cause parts of the meaning to get lost (e.g. the translator might change the dynamic of the interview situation or choose the same word even when the respondent in fact uses different words with similar meanings, etc.). Apart from the above-mentioned problems narratives/data obtained from interviews are not necessarily comparable over different cultural situations, even though the interview has the advantage over a questionnaire that the respondent has the opportunity to signal or question if something feels uncomfortable or hard to understand. This makes the interview adaptable to cultural differences, but the adaptable nature of the interview can be deceiving. Another difficulty is that the method itself can be hard to use in the manner which the researcher is used to. Although you are familiar with the ‘proper’ way of conducting qualitative open-ended interviews, the method might not be suited in another culture. Much of the method used to make a qualitative interview is developed in the Western part of the world (Brislin, 1976; Laverack & Brown, 2003) and is therefore not necessarily adapted to non-Western cultures or to cross-cultural situations. Difference in status and in the expected form of interaction between the interviewer and the respondent might cause problems for example, especially if the two come from different cultural backgrounds (Ojeda, Flores, Rosales Meza, & Morales, 2011). In short, it is dangerous to assume that any concept, including methodology, is universal. Therefore, it might be necessary to re-think the open-ended interview or at least redefine its practical use.

Although some researchers are aware of the problem and mention a need for new types of methods (Demuth & Fatigante, 2012; Kim, 1999) few of them come with concrete suggestions of solutions or give practical examples of the problems you might encounter. Furthermore, they seem to focus on the interpretation of the data more than the collection and the situation of the interview. In this paper, we try to focus on the interview situation more than the subsequent analysis.

Purpose

This article aims to highlight some potential methodological problems of using the open-ended interview in a Chinese context, as well as promote a heighted awareness and discussion about methodological problems stemming from using Western methods in non-Western cultural settings. It does not aim to provide any general or simple solutions to these methodological problems. Instead, our goal is to contribute to deeper and wider discussion concerning validity by using practical examples as well as discussing possible solutions.

This paper is based on experiences we had during 2012 and 2013 when two of the authors of this article travelled to China to conduct research. We ourselves are from different cultural backgrounds (French and Swedish). The respondents and the subjects we were investigating were different, but our methodological approach was the same, a phenomenological approach called meaning constitution analysis (MCA) (Sages, 2014; Sages & Lundsten, 2009) for which the open-ended interview can be an important tool, as the method emphasizes that the narratives/data should be collected with as little guidance from the researcher as possible. The categories and concepts used should not be decided beforehand but emerge from the narrative itself (Sages, 2003). Although we have experience in the traditional way of conducting open interviews, we encountered problems when conducting interviews in the Western way while in China. Our approach just did not seem to be the ‘proper’ way there. We have combined our experiences into seven areas where the interview contexts in China differ from the ones in Europe. Our third colleague is Chinese herself, has extensive experience in conducting interviews in China, and she has provided us with examples of the same phenomena using an open-ended interview, but with the advantage of speaking the language and being closer to the culture at hand. Although the problems of the intercultural interview have been discussed before (see for example Dermuth & Fatigante, 2012), there have been few concrete examples or discussions about how to handle these types of situations. Drawing from our own experiences led us to question the contextual validity of the open-ended interview in China. We found seven main areas where the interview situation differed from the Western context that we are familiar with. We will look at these areas one by one, hopefully starting the path to a more valid phenomenological understanding of the contextual dependency of the open-ended interview as a method and a better position for accurate interpretation of the narratives collected with the help of this data-gathering method. First, we will present the method used and the research projects which form the basis for these conclusions in order to provide a context.

Method

The issue addressed in this article concerns the method of the open-ended interview which is primarily used in qualitative research. Our research and data gathering has been done from a phenomenological standpoint and the narratives has then been analysed with the MCA method and the software Minerva (for a more detailed account of the Minerva software and MCA see Sages, 2014; Sages & Lundsten, 2009). The idea of MCA is to investigate how the respondent attaches meaning to the world, and therefore (as far as possible) the respondent should be the only source of meaning (Sages, 2003). A broad open format like the open-ended interview is often used because it allows any type of meaning to be discussed, not only the type of meaning attached to pre-prepared or guided questions. It should be said that we believe that these problems we encountered are not solely a problem when using this method of analysis, but can become an issue in all instances where open-ended interviews are used, in particular involving more than one cultural setting.

Epoché

The common method for our research has been the MCA. As in most phenomenological research, the goal for the researcher using the MCA is to achieve epoché (Sages, 2014). It is a state in which the researcher tries to be as conscious as possible to her/his way of interpreting data, and as far as possible let the narrative speak for itself in the way that it appears, hence refraining (as much as possible) from adding the researcher’s preconception (Husserl, 1982; Sages, 2003, 2014). It is, in its ideal form, a pure bottom-up process (Husserl, 1980, 1982; Sages, 2014). To put in other words, epoché is to view the objects as the respondents intended (Husserl, 1982). Complete epoché, that is detachment from all preconceptions, is impossible to reach, as no one can ever be passive receiver of perception; to perceive is inevitably to interpret. One of the aims of the MCA method is to help the researcher to come as close to the epoché as possible (Sages, 2014). The open-ended interview is therefore an appealing methodological tool where the researcher refrains from having prepared questions, instead letting the respondent decide what is relevant to say. Questions should be used carefully and always follow the respondents own words and trail of thought, avoiding to introduce foreign material into the data. This needs a careful examination of the preconceptions that come with the method itself so that we might understand more about how to get a response as open as free as possible. This state of objectivity might also be desirable within the frame of other methods and so the open-ended interview is a multi-faceted tool.

The research projects involved

The data gathered were for four different research projects. None of these had the purpose of examining open-ended interviews as a method, but the experiences from them form the basis of this article. We will briefly present the research here to give the reader a fuller picture of the context.

Project A: Elderly and technology

This study concerned elderly Chinese people and their impressions of, and wishes concerning every day household technology. The data gathering was in both winter and summer time. Eight interviews were conducted, mostly in and around Ningbo and Huzhou (in the Shanghai area), but one also took place in Mengzi in the province of Yunnan. The respondents were both from rural and urban areas and spoke Mandarin or a local dialect. The foreign researcher was in this case Swedish and all the interviews were made with the help of interpreters. The interpreters were students or staff from local universities, and in two cases close relatives of the respondent. The interviews lasted between 15 and 45 min and were taped on a Dictaphone. No notes were taken during the interview. The interviewers asked the respondents about their favourite technological tool. They were also asked what tools they wished for or how they would like to improve the tools they already have. All of these interviews were made in their homes or at day activity centres for the elderly, sometimes in a private room and sometimes in the presence of their family and/or friends. The result of the study is still being processed. The respondents were both male and female and between 55 and 85 years old.

Project B: Moral non-conformers

These interviews were made as a part of a PhD thesis. Twenty-four interviews were made in China in Ningbo, Shanghai, Huzhou, Kunming and Mengzi. The respondents were males and females between 12 and 70 years old and from different socio-economic backgrounds, although the majority were university students in their early 20 s. The interviews were conducted in a variety of places such as the work places of the respondents, at restaurants, the hotel room of the interviewer, in parks, the home of the respondent, etc. Convenience and the wishes of the respondent guided the interview setting. The (not yet finished) thesis concerns people who do not conform to one or more of the norms of the group(s) they belong to because they believe that the group is behaving immorally (examples of this are whistle blowers, people who help the victim of bullying in their work place/school, etc.). The people interviewed were asked to think out loud about this type of behaviour and if they had such an experience themselves tell the interviewer about this. About half of the interviews were made with interpreters who were university students, friends or colleagues of the respondents. In the other half, the respondents knew enough English to participate without an interpreter. However, this did not mean that the interview was conducted with only the interviewer and the respondent present. All interviews were recorded on a Dictaphone and no notes were taken during the interview.

Project C: Children and their cultural vision

This research was about the young child and her/his cultural vision of the world and part of a master thesis. The research was conducted by our French co-author and an interpreter was present during all interviews. The respondents were 15 children, from five to six years old, across several kindergartens, both private and public, in the Yunnan province. The children were from upper, middle or lower class, depending on the type of kindergarten they were attending.

The children were participating three at a time. They were asked to draw two different pictures: one of ‘their family/home’; one of ‘their friend’s family/home’. After each drawing, followed an interview, which were conducted at the kindergarten, separated from the rest of the class. However, every child preferred to stay near his two other classmates for the duration of interview. Everything was recorded (audio and video) (Blanchin, 2013).

Project D: Life quality of the elderly in a rural village

This case study was conducted as a joint effort between Lund and Hong He Universities as part of a cultural research project on the meaning of ageing and its physical and psychological effects on the individual and society. The project was a cooperation between a Swedish–Turkish researcher and our Chinese co-author. Both of them were present during the interviews which were primarily held in an open courtyard of a private home (accessible to friends and family of the owners of the house). The interviews were conducted in a local dialect and in Mandarin, and translated on site into English in place by the Chinese researcher while another Chinese colleague was taking notes. The interviews were transcribed into Mandarin and translated later (the on-site translation was only to benefit the Swedish co-researcher). There were 12 respondents (both males and females). The oldest respondent was 86 and the youngest 15. All respondents were offered 50 Yuan (about 5 €) as compensation for their participation (Sages & Liujin, 2016).

Problems encountered in open-ended interviews in China

We have summarised the difficulties we experienced into seven themes. The order of the themes is not to be taken as any indication of hierarchy, and we also want to emphasise that these are generalisations from a mostly Western perspective, and when we use the term ‘Chinese people’ we are aware that this is a gross oversimplification and generalisation.

What is the correct answer?

In China, we as interviewers often found ourselves in a situation where the respondent tried to find out what we wanted them to say, and then answer the question ‘correctly’. On several occasions we experienced the respondents apologising for not giving us the answers that they thought we wanted. This behaviour also occurs in Western countries, but our experience is that after explaining that it is the experiences and thoughts of the respondent which are the focus of the research, and that there are no ‘right’ answers, it usually subsides, and in the course of the interview the respondent tends to relax and feel more comfortable with the answers they are giving. In China, we have not had the same experience. Even when explicitly told there are no ‘right’ answers, the respondents frequently apologise for not saying what they believe we want to hear or not having good enough examples to give us.

When conducting open-ended interviews for phenomenological research the primary method is to ask broad questions in order not to influence the respondent in any direction. A typical question could be: ‘Can you tell me about your experience, associations or thoughts about this phenomenon? If you don’t have any specific experience, tell me all you come to think of or associate to this’. When asked in Sweden or France, the respondent can sometimes reply with a question such as: ‘What do you mean, this phenomenon, how?’ Then it is usually enough to tell the respondent that ‘I am interested in your point of view. All association thoughts and emotions you want to express are valuable to me’. In China, this is also usually accepted as a response and the respondent starts answering, but unlike in Sweden or France the Chinese respondent will still often interrupt themselves and ask: ‘Is this want you want to hear?’ or ‘Is this example correct?’ or even apologising for not having the ‘correct’ experience, even after such a reassurance.

In Project C the children were asked to draw two different pictures (‘their family/home’ and ‘their friend’s family/home’) and then talk about it. A translator enabled the conversation between the researcher and the child. We noted that they mostly talked about their drawings in a descriptive way, avoiding emotional interpretations; we could see that they were both uncomfortable and unfamiliar with talking about their personal feelings and thoughts. Descriptive answers might have been more comfortable for them as a reassurance against giving the wrong answer, or perhaps it was how they are used to giving answers at school. The translator, despite having had the researcher explain the importance of letting the children express themselves as they wanted, was sometimes embarrassed because she thought the child did not talk enough, or did not say what she thought that the researcher expected. It was necessary to remind her that the content of the interview depends on the child rather than the researcher, that there is no right or wrong answer and that each answer is interesting in its own right. Once, one teacher came to observe her pupil during the interview. The child was clearly not feeling comfortable in this situation; the stress of giving the ‘right’ answer intensified, and at one point the teacher guided the child on what he should talk about. Here, we can see that this question of giving the right answer is present since the very beginning in the Chinese system of education.

Another example is from Project B (where the issue of standing up for one’s own morals in the face of peer pressure was investigated). Many of the answers were so alike and so ‘non-threatening’ that they were likely a symptom of social desirability: ‘I study when all my friends are lazy’ and ‘I always wait for the green light before crossing the street’ (if the response rate of this latest example was any indicator of how Chinese people acted in traffic, the traffic situation in China would be drastically different).

In Project D, interviews conducted with elderly people in a rural village, we intended to ask broad open-ended questions. It turned out few of the respondents knew how to start to answer such a broad question. One of them said to the researcher, ‘I don’t know where to start and what to say. Why not just ask me direct questions?’ Then we switched and started to ask guided questions instead. With this method we obtained rich answers to the questions, but of course took the risk of damaging the validity of the research in favour of the respondent feeling comfortable. On the other hand, it increased the ecological validity of the study, as we adapted our way of questioning to the respondent rather than forcing the respondent to adapt to us.

These are examples of situations where, if no hints about the correct answer was given, the respondent acted nervously and did not talk much at all, or alternatively made a guessing game (guessing the correct answer) out of the interview. One way to solve this was, as presented above, switching to more guiding questions. Another was to wait and let the respondent give all ‘right’ answers, and then let them continue onto their own ideas, associations and experiences. An issue here is of course to distinguish ‘their own’ views from the ones they believe to be ‘correct’.

We suggest that one of the flaws of the open-ended interview in a Chinese setting is that the method and the theory have been developed in Western countries, and therefore it is culturally Western centric. In China’s educational tradition, the role of the teacher is a ‘provider’ and a ‘knower’, and the student is a ‘listener’. This makes it unusual to create an inviting atmosphere between the teacher and the students (Chen & Bennett, 2012; Skinner & Abbott, 2013). The researcher–respondent relationship is much like that of the teacher–student. In such a situation, the respondent might automatically assume that the researcher is the ‘knower’ with ‘correct’ answers and that they themselves are expected to try and give ‘correct’ answers to the questions posed. One could, however, argue that this is not culturally unique for the Chinese and there is research to suggest that the social desirability bias is lower among Chinese than Westerners (Dunn & Shome, 2009), although this is contradicted by other studies that show a greater social desirability bias in highly collectivistic cultures such as China (Bernardi, 2006). Another possible explanation of this** phenomena is that Chinese and Westerners have a different interpretation of what an interview is and what it should lead to. Previous research (albeit concerning job interviews rather than research interviews) has indicated that Chinese managers see an interview more as a ‘fact-finding mission’ whereas Western managers are more interested in the respondent’s own interpretation than an ‘objective fact’ status (Easterby-Smith & Malina, 1999). Thus, it might not be a greater social desirability but a confusion about the purpose of an interview which causes this issue.

Now I have answered your question. Give me a new one.

Often the Chinese respondents try to answer the question given quickly and correctly. They then wait for the next question and an answer is often followed by: ‘That’s all. What is the next question?’ They do not ‘think out loud’ as frequently as in Sweden or France. This limits the amount of information given in each answer and makes it difficult to ask follow-up questions which derive from the respondent’s own words. This affects the validity of the study negatively, as the more questions you ask from a ‘question battery’, the higher the risk of influencing the respondent. One of the possible reasons that the Chinese respondents are more reluctant to ‘think out loud’ than Westerners in general might be due to a different way of thinking in East Asia, a thought process much less based on internal conversation. It has been suggested that for East Asians, thinking and talking are two separate tasks (Kim, 2002) and therefore the ‘thinking out loud’ that we expect from our respondents in an open-ended interview might not come as naturally to Chinese respondents as to Western respondents.

The children in Project C had this tendency to give brief and less detailed answers. This behaviour was encouraged by the adults around (such as the teachers or the translator). The children were encouraged to say what the adults perceived as relevant and avoid ‘losing themselves’ by saying something useless. This might be related to the fact that trite phrases are not appreciated in China and not customarily used in everyday life.

One way to get people to tell you more about an event is to return to it several times during an interview. Whereas in Europe where this return to a previous statement is often followed by a pause for thought and then a new dimension of the issue, in China the most common answer is a short: ‘No, that’s all’ or a confused look seemingly wondering: ‘why would we want to talk about this again?’ Sometimes a follow-up question is taken for criticism: ‘Didn’t I tell the correct story or didn’t you understand me the first time?’

Privacy – a matter of logistics

In China it is somewhat hard to find places to talk privately. In many schools or working places there are few private rooms to conduct an interview in. People often live with many in a small space so it is also hard to find privacy in their homes. Interviews are therefore often conducted in a corner of a room where other people are working, cooking, studying, etc. The privacy issue is also a matter of weather. In China the weather often gets very cold in the winter time. It is therefore not a viable solution to ask people to leave the home to conduct an interview, especially when interviewing children or the elderly. In the summer, on the other hand, the weather can become very hot. It might therefore be necessary to keep the door open during the interview even when you have found a private room. The matter of privacy has to be weighed against the matter of physical comfort, and as point 3b will show, sometimes privacy is not the most prioritised factor. To have a respondent sit in physical discomfort might not be the best environment for a relaxed interview. It might also be unethical, especially if the respondents are very young or very old.

Privacy? But I don’t have any secrets from my friends.

Apart from the logistic problem in finding room for a private interview, many seem unwilling to participate alone. Sometimes there has to be a translator present, but many times other people also insist to be present during the interview. When the researcher asks these other people to leave, some of the respondents insist for them to stay explaining ‘I don’t have any secrets from my friends’. We also find that this might be something beneficial. Many respondents feel nervous in the interview situation and they seem to feel more comfortable when in the company of a friend. However, sometimes this friend is not only a quiet bystander but also many times they interrupt and offer suggestions about what to say and what stories the respondent might have but has neglected to tell. Sometimes they even fill in with their own story. This might pose a threat to the validity of the study. It can affect the person in many ways: (a) The respondent does not say something that they would otherwise have wanted to say because they do not want their friend to hear. (b) The respondent says things they would normally not say because they are influenced by their friend. (c) It might increase social desirability. It can also influence the interviewer by: (i) Making the situation confusing and the interview harder to follow. Splitting the concentration between people might reduce the researcher’s ability to listen to the primary respondent. (ii) The researcher might neglect to ask some questions because he/she does not want to embarrass the respondent in front of their friend(s).

There are, however, positive things about letting their friend attend the interview: (a) The respondent might feel more comfortable. Many Chinese respondents feel shy in an interview situation especially when interviewed by a Western person (see point 7). Having a friend as support might get the respondent to overcome this shyness. (b) When interviewing a person about a difficult subject, the friend can help explain and sometimes remind the respondent of experiences which match the question. This has proven to be very important during some interviews in Project B, where the respondents did not think of themselves as doing something moral, but the friend reminded them and some interesting stories unravelled. Of course being reminded what to say might get the person to think in a certain way, thereby increasing the pre-conceptual bias, but it might also be the springing point where the respondent has a story to tell and from there the interview might run smoothly. We find that getting a person to tell a story opens up the conversation and makes it easier to ask about thoughts, feelings and processes after the story has been told, although this is not always the case in China.

The following experience further illustrates this: in Project C the initial idea was to interview the children one by one, but the Chinese people who were helping the researcher recommended that the children should be in group (groups of three in this case), because they probably would not be able to answer otherwise. The explanation they gave was that the children would not dare to do something personal if they do not see the others doing it. It was like their ‘personal view’ depended on the group view or at least it is the way they were taught to behave in school. One explanation for this behaviour could lie in the collectivistic nature of Chinese society. Collectivistic and individualistic features are something that can be used to define cultures all over the world (Triandis, 1995; Triandis, Bontempo, & Villareal, 1988). In collectivistic cultures people tend to depend on others and do often see this interdependence as a duty or a moral obligation. This might be linked to the children feeling nervous when being interviewed alone. Maybe they actually need the group to allow themselves to think individually (Triandis et al., 1988).

In Chinese culture, privacy can be interpreted in two opposing ways. On the one hand, Chinese people like the way of ‘togetherness’ which gives to them more sense of ‘belonging’. This is why many Chinese people do not mind talking to the doctor about their diseases with other patients standing beside, watching and listening. This might be due to the interdependent qualities of a collectivistic culture where the border between oneself and others is less distinct than in individualistic cultures (Heine, Lehman, Markus, & Kitayama, 1999; Markus & Kitayama, 1991). Other ‘selves’ might in fact be crucial parts of the respondent’s self thus making certain type of privacy unnecessary. In the situation of conducting interviews, if the respondent is put in a private space with a stranger, many respondents would feel insecure as they are isolated from the environment and the people they are accustomed to. Some would even be concerned that their peers would think they are abnormal. Most people, especially elderly people, do not want to look different from others; a trait common in collectivistic cultures (Triandis, 1995). Focus groups could be an alternative to the one-to-one interview, and it seems like it is good to keep the group as homogeneous as possible in order to facilitate discussion. This might have something to do with the vertical culture structure in China; if everyone is basically on the same hierarchical level, everybody in the group can feel freer to talk about an issue without worrying about hierarchy, whereas the presence of group members higher up the hierarchy can inhibit the conversation (De Jong & Van Ommeren, 2002).

Chinese culture also tends to instruct people to be self-controlled and to hide personal feelings, especially negative feelings. Emotions and personal thoughts could be private and should not be revealed to an ‘outsider’, sometimes not even to people close to them. For this reason, unless the researcher has built trust with the respondent, or assured the respondent that ‘I am part of your group’, getting to hear the ‘personal view’ could be a challenge for the researcher who is psychologically seen as an intruder to the respondent’s culture. So while having friends present might increase the respondent’s willingness to talk, it could also inhibit it. However, even being alone with the foreign researcher would not alleviate this; if the respondent deems it too private to tell their friends, then it is probably too private to tell a foreigner.

In sum, expecting privacy in certain cultures might be adding a bias to the research as it makes people more uncomfortable and unable to answer freely. Privacy alone could be a concept that is hard to grasp for people from certain cultures, as it is not sought for, or even a concept that exists. Even in Europe the concept of privacy is a relatively new one (Baumeister, 1987), and so we should expect that privacy is a cultural and relative term. Thus, in China, private one-on-one interviews might not be ideal.

Reluctance to talk about oneself

In the interview situation it has proven hard to convince the respondents that their own perspective is important. The Chinese respondents often speak about abstract examples instead of giving experiences from their own life. This happens even when explicitly asked about their own experiences or asked to give concrete examples. When they do talk about a concrete example they tend to exclude everything except for what happened. Few reflect upon their feelings, emotions and motives. It seems like the respondent is interpreting the situation as, ‘when the researcher asks for an example, they want to know what happened and should not be bothered with personal reflections’. Sometimes the respondent redirects the attention, telling someone else’s story which they perceive as more ‘correct’ in terms of what the researcher has asked about. In that sense this issue is connected to point two; it could be a result of the respondents trying not to ‘lose themselves’.

This behaviour could be a matter of respondents trying to be humble and not presenting their own experiences as more worth than somebody else’s. Modesty is seen as a virtue in the Chinese society (Bond, Lun, Chan, Chan, & Wong, 2012). This difference in willingness to speak about themselves might also be due to differences in personalities between cultures. Chinese respondents tend to be shyer than Western respondents. The Chinese respondent is trying to be modest by not promoting their own experiences as something special, and this might not be a conscious act but rather something built into the culture to be less self-promoting and more self-critical than the average European person (Heine et al., 1999; Kanagawa, Cross, & Markus, 2001; Lee, Aaker, & Gardner, 2000). It could also be attributed to a lower level of extraversion and higher neuroticism previously seen in collectivistic cultures more than individualistic cultures (Allik & McCrae, 2004). This could make respondents timid and less prone to exhibitionism, although there are problems with validity and reliability when measuring personality across cultures, which should be taken into consideration when discussing these differences (Heine, Buchtel, & Noenzayan, 2008).

The Chinese children interviewed did not want to consider, or could not consider the researcher, the interviewer, as a ‘friend’, but wanted to call her ‘teacher’. Even after she had said that she was not there to judge or assess them, and that she just wanted to get to know them, the children kept calling her ‘teacher’. This might be a way to keep a comfortable distance, to not open up too much too soon. It is possible that Chinese people need more time to be able to talk about their feelings and emotions. They would not do it easily with a stranger, even if they want to give the researcher the information he/she wants. Another explanation could be the vertical structure of the Chinese society (Triandis, 1995). In a vertical culture the hierarchy of society is pronounced and one should show proper respect to those which are considered ‘above’ oneself such as teachers, parents, etc. The opposite end of this scale is the horizontal culture where titles are uncommon and everybody should be treated with the equal (lack of?) respect.

But enough about me tell me something about your experience!

One of us experienced that many of the Chinese respondents turned their questions to her. This also happens sometimes when conducting interviews in Western countries, although the impression is that it was more common in China. In the middle of the interview the respondents might throw back the question and ask: ‘What do you think? What’s your hypothesis?’ Here it is hard to know the right course of action. The researcher could chose to ignore the question, saying that they do not know or that they will answer later. This is what they would normally do in Europe. But as they are already in a situation with a suspicious respondent who is reluctant to talk about themselves, they could use this situation to build trust. They could answer the question and thus with the help of the social rule of reciprocity, establish a trusting relationship with the respondent (provided the question is not directly related to the subject researched, as it could pose a problem by directly influencing the subject). This might be a way for the respondent to acquire a more intimate relationship with the interviewer, which would enable them to speak more freely. This type of ‘conversational interview’ has benefits and the method has been used before in a Hong Kong interview setting to make the respondents more relaxed (Yeung, 1995). It might skew the respondent’s answer but it could also be a way to build trust and get them to open up (Ojeda et al., 2011). An alternative to the conversational interview is the focus group technique, where the discussion between the participants facilitates the interview without the researcher’s direct participation (Knodel, 1995).

Insecurity about being recorded

When conducting interviews in Sweden or France, people do not usually mind being recorded. This is a different matter in China, especially when interviewing older people; the recorder can make them feel nervous. They keep looking at the recorder and their mannerisms change noticeably when it is turned on. This sometimes happens in Western countries too, but normally the sideways glances at the recorder subside when the respondent is informed about what will happen with the material, and that it is for the researcher’s ears only. In China, these glances continue and the respondents might show visible discomfort with the device during the course of the entire interview. Recording can be easily associated with ‘proof of something wrong/bad’ in a Chinese context due to historical reasons. This might be one of the reasons why the respondents are uneasy with the recorder. With the presence of a recorder, even if people are ok with the device, they could potentially subconsciously choose to speak about positive things out of social desirability. This is problematic for the validity of the research. The obvious solution to this is not to record the interview, but this option is far from perfect. Notes tend to be unreliable and to miss parts of the conversation. Taking notes might also have the side effect of the respondent feeling like they are taking a test and further strengthen the effects described in point one.

Suspicion of Westerners

In the big cities of China it is not unusual to see Western people these days, but in the smaller towns and villages Western people are quite rare. For some respondents, the meeting with a foreign researcher might be the first time they have spoken to a Westerner. This might make them shy or tense during the interview. For a Chinese respondent, a Westerner is often considered out-group but also treated as higher in status (Easterby-Smith & Malina, 1999). Thus, any interview conducted by a foreigner runs the risk of becoming an interview conducted by a high status out-group member and that is not a good breeding ground for trust. Just as in point six, our experience is that this reaction is more prevalent among the elderly, although younger people also tend to be shy, especially if they are expected to speak English. To some extent the presence of a foreigner could be threatening to the respondent. There is an old saying in China: ‘Do not wash your dirty linen in public’, which means, if there is anything bad happening in your private life, you should not let other people know about it. In the interviews for Project D, several of the older people kept mentioning the merits of CCP, and one of them even warned a Chinese lady who was part of the research team to be cautious about what was being said.

One way to solve this is, of course, to remove the Westerner from the interview situation altogether and let local Chinese associates hold the interviews. This might be a viable alternative but if the interview is to be analysed by the Westerner and if the study will be conducted in other countries too, it might be important to the understanding of the context that the Westerner is present, and as our co-author noted being Chinese does not guarantee that the open-ended interview technique will work smoothly.

Discussion

This article does not claim to have found any perfect solutions to the problems encountered, nor does it aim to do so. Its purpose is to heighten the awareness of the issues that a researcher might face in the context of an open-ended interview as a Westerner in China. We want to point out that the situation of a classic open-ended interviews seems to mean something else to a Chinese person than to, say, a Swedish person. The difference of contextual meaning also changes the nature of the collected narratives. The validity of the method of open-ended interviews is questionable in a Chinese context, and perhaps also in other non-Western contexts, as well as with migrants in Western countries. To adapt the method to the environment it might, for example, in some circumstances be better to hold group interviews to counter point 3b (the matter of having friends present during the interview). This could perhaps reduce the uncertainty of the situation and make the respondent feel more relaxed, also with the fact that the interviewer is a foreigner (see point 7). Although this could potentially raise new validity problems such as problems with social desirability. No ‘data gathering’ is without its own set of potential problems, showing the context dependency of any methodology, and making a strong argument for developing new and adapting old methodologies.

One possible way to tackle some of these problems, such as the tendency to try to ‘answer correctly’ or the respondent’s shyness, is to build a trusting relationship with the respondent, and thus make it less likely that the respondent will feel uncomfortable talking to the researcher alone. However, this is time consuming and could affect the researcher’s, as well as the respondent’s, pre-understanding of the situation (for good or for worse). There might be ways to circumvent this, for example, by using a translator, or by using a co-worker’s own contacts as respondents, thus earning some trust from the respondents, albeit by proxy. Another possible solution is to really take the time to explain the procedure of the open-ended interview. It is, perhaps, more important in Chinese, than in Western contexts to describe in detail, not only the purpose of the research but also of the open-ended interview as a method and how the nature of the interview will affect the analysis.

Another alternative is to not make these transcultural interviews at all, but instead use local interviewers. This can be a good alternative, but it can be hard to find trained interviewers, and the researcher might miss some of what went on during the interview and therefore a vast portion of the information given in the interview (body language, environment etc.), perhaps so much that some of the meaning gets lost. Also, as we noted, a Chinese researcher is by no guarantee that these issues disappear.

The matter of not wanting to respond incorrectly (point 1) might also prove difficult to solve. This type of social desirability bias is not just limited to China but present in more or less all studies. One solution is perhaps to give more time for the interview so that there is room to think and reflect, and also to make sure that it is hard to fill the time with social platitudes only. Extending the time for the interview might lead to other problems though, such as difficulties to find respondents who are willing to invest that amount of time, etc. It might be good to change the way of thinking about the interview situation, if Chinese respondents see the interview as a fact-finding mission it might be better to use an interview technique which is more based on a dialogue or a conversation rather than the ‘one-way’ open-ended interview. This could also potentially serve as a solution to point 5 (‘Enough about me…’) and prevent shyness by building a reciprocal relationship. Informing about the nature and purpose of an open-ended interview or have a dialogue instead might also lead them to talk more about themselves and they might also understand the need to ask about the same subject more than once.

To remedy the scarcity of information given in the interview (as noted in point 2 and 4), and to try to circumvent the problems of privacy, the focus group technique could be a feasible alternative to the one-to-one interview. Focus groups have the advantage of letting the group members discuss the issue, and by this discussion evoke new ideas and broaden the scope of the issue by associating together, and the initiative of the more extravert or outspoken participants can facilitate the more timid respondents’ active participation (Knodel, 1995). Although it is possible that the opposite might occur, that the more timid respondents do not feel comfortable or even able to actively participate in the discussion. The focus group technique comes with other possible problems, such as the respondents being less willing to get into personal problems and/or sensitive (e.g. norm criticising) issues in a group (De Jong & Van Ommeren, 2002; Knodel, 1995), the risk that the personality of the researcher influences the discussion, or that the researcher is not adequately trained in the method (Knodel, 1995). But these last two problems are basically the same that you would expect in a classic one-to-one open-ended interview.

One of the solutions we tried namely: switching to a guided form of interview or to a dialogue could have some negative impact on the validity of the study as the researcher then claim more of the interpretative power, but it could on the other hand, if this is what the respondent feel more comfortable with, lead to a greater ecological validity. It is a question of weighing the negative against the positive, a guided interview is still possible to analyse phenomenologically with MCA. However, it is difficult to know if the question will lead to responses that include more of the preconceptions of the researcher, and less of the respondent’s intentions, and we cannot analyse what has not been said.

As shown in this article we felt that we had no good tools to handle these sorts of problems when we run into them in the field. All possible solutions come with a whole new set of practical problems and biases, and often the ones we were aiming to avoid by using the open-ended interview to begin with (e.g. limiting influences of the researcher's preconceptions). What we need now is a discussion about how to handle these problems. Perhaps we should rethink the idea of the open-ended interview as a contextually independent method, making ourselves aware of its potential pitfalls and when necessary find other ways to gather the same quality and type of information/narratives, a way which is better suited to the Chinese way of interaction between researcher and respondent, focused on the respondent’s own lifeworld as it appears, as free from preconceptions from the researcher as possible. This has to be done in cooperation with Chinese respondents. Giving the cultural group who are going to be interviewed a more direct involvement in the research methodology has had good results before (Ojeda et al., 2011). Practical tips and warnings should be readily available in the literature; methodological issues on the practical level should be admitted and discussed in articles, not taken as a sign of weakness, but of admitting reality into the research method. We cannot persist in excluding the non-Western perspective from our methodology discussion, and we have to be able to think outside the box in these issues. What we are interested in when making intercultural or indigenous research is to get comparable, qualitative data, or to understand narratives in their own right. This does not necessarily mean using the same type of method, because using the same type of method would be, as said in the beginning, to take the social out of social sciences. Using the open-ended interview in China will not give you the same quality of the narratives/data as in Europe, and data sprung from intercultural research using this technique are therefore not methodologically comparable, even if the method is the same.