people more familiar with are quantitative research. However, in this article, the main focus is on qualitative research, which is widely used in academic and professional areas (Holliday, 2002). So, what is quantitative research? Mason (2002:1) believes that:
“Through qualitative research we can explore a wide array of dimensions of the social world, including the texture and weave of everyday life, the understanding, experiences and imagining of our research participants, the ways that social processes, institutions, discourses or relationship work, and the significance of the meaning that they generate.”
Therefore, from this perspective, it could be derived that for qualitative research, pure statistic and numerical data are not sufficient. It needs more description evidence to discover the subjective qualities that govern human behaviours (Holliday, 2002). The so called ‘thick description’, which could be collected through interview, observation or other methods, is the basis for qualitative researchers to understand and explore the social world(Draper, 2004). Due to the complexity of modern society, there are many problems emerged when conducting qualitative researches, for instance, the ethical dilemmas. In order to get a more critical view of qualitative research, the distinction between qualitative research and quantitative research will be highlighted first. Then two typical qualitative research methods, in-depth interview and participate observation, will be critically examined. Followed by that, three examples will be used to further illustrate the challenges of qualitative research. In the end, a conclusion regarding the merits and demerits of qualitative research will be drawn.
Primarily, the philosophical basis for qualitative research and quantitative research are different. Both Holliday (2002) and Draper (2004) state that quantitative research believes that by using right quantitative measurements people could reveal objective facts, while qualitative research rooted in interpretive tradition, which focus on exploring the underlying meanings of social phenomena. As a result, qualitative and quantitative research differ in their analytical process, research design, methods of collecting data and the approach to analyse and interpret data(Draper, 2004). Rather than by using a deductive process, qualitative research prefers inductive procedure (Gephart, 2004). Therefore, qualitative research provides extra opportunities for researchers to discover insight views and get new understandings of social phenomena (Willig, 2008). Also, Gephart (2004) states that the ‘thick description’ provided by qualitative research enriches the basis for understanding the social world. However, qualitative research also has its own limitations. As Willig (2008) points out that qualitative research do not ensure certainty, the objectivity of researcher is not realistic and the results are not predictive, since qualitative studies always carry out with small sample size but in depth. Moreover, similar qualitative studies due to the different use of methods are impossible to compare or integrate (Willig, 2008). As a consequence, Draper (2004) suggests, although qualitative and quantitative researches are different from their ontological foundation, they could complement each other. In addition, recent studies (Bryman, 2006; Kinn and Curzio, 2005; Sherman and Strang, 2004) show that there is a growth trend of combining these two methods together, in spite of the continuation of debate about integrating qualitative and quantitative methods. After examining the merits and demerits of qualitative research, it is the time to discuss the two dominant methods used in qualitative research, in-depth interview and participant observation.
Initially, it is necessary to know why people choose in-depth interview instead of others, like structured interview. Firstly, by saying ‘structured’, it means standardization, follow a schedule strictly that allow little deviation from a pre-established protocol (May, 2001). In contrast, in-depth interview, which is also known as semi-structured interview, is more flexible. During in-depth interview, the sequence and structure of the interview will be influenced by interviewee’s answers, interviewer need to listen carefully to follow interviewee’s lead (Esterberg, 2010). So from the above statements, it signifies that in-depth interview gives interviewee more room to present their own opinions, and do a great contribution to the ‘thick description’. As a result, “in-depth interview are particularly useful for exploring a topic in detail or in constructing theory” (Esterberg, 2010:73). Additionally, Willig (2008) concludes that the popularity of in-depth interview is due to its easier arrangement than other methods, especially participant observation.
After considering the advantages of in-depth interview, one might ask what the drawbacks of in-depth interview are. Rather than talking about drawbacks, here is preferred to list the challenges raised in carrying out in-depth interview. The chief challenge faced by interviewer is how to encourage interviewees express themselves more freely and openly, and to make them fully understand what is expecting in the interview (Willig, 2008). This is depends on interviewers’ skills and experiences. In addition, there will be circumstance that response effect occurs. Response effect refers to a situation when interviewee tries to please or resist interviewer (Dunne, Pryor and Yates, 2005). This is a problem associated with researcher’s identity and position in the research. As Dunne et al. (2005) advises that researchers need to be self-conscious about their position in the research from the beginning of research design to the end of interpretation, including conducting an interview. Furthermore, if a rapport is built, interviewee would be more willing to talk (Esterberg, 2010). This is also an opinion held by Willig (2008), and she points out that rapport is delicate; it could be disrupted easily, for instance, when interviewer stops to change audiotape. Finally, in-depth interview share two challenges with participant observation, access problem and ethical dilemmas, which will be discussed later.
Primarily, a key point of participant observation is that it requires researchers involve in various activities, such as participation, documentation, informal interview and reflection (Willig, 2008). As a result, participant observation enables researchers learn more about social society than simply asking questions. As Mason (2002) argues that ethnography provide depth, sophisticated and multi-layered data for social explanation and arguments. Moreover, May (2001) contends that, ethnography results in a sentiment understanding of social phenomena, and reduce the bias of researcher’s preconceptions.
In spite of those strong points provided by participant observations, there are several difficulties remain. The first problem is the decision of appropriate dimension of participation. One needs to involve enough to figure out what is going on, but not so deep as emotional problem will affect the study (Willig, 2008). Actually, Dunne et al. (2005) proves that the researcher has an impact on the whole context once they present in the research, no matter as an active participant or outside observer; and this effect is impossible to eliminate, also the quality of the effect is associated with researcher’s identity. This kind of effect is refers to reflexivity, which will affect the process as well as the outcome of the research (Thorpe and Holt, 2008). Next, the main problem of participant observation is gaining access, a question related to informal consent. It will be more problematic when enter a more private setting is required. It may involve negotiation and building relationship (Mason, 2002; Laine, 2000). So far, the access issue has been discussed; this is also a problem facing by interview method. This is because interviewers need get the informal consent of interviewees, and it is a moral concern problem (Mason, 2002; Dunne et al., 2005; Corbin and Morse, 2003). Due to the importance and complexity of ethical issue, it will be discussed in a separate section.
Principally, it is undoubtedly that ethical dilemma is not only an issue in interview or participant observation, but also a universe problem through the whole process of research (Mason, 2002). So, what is the rule of being ethical? To answer this, it is perhaps better to look at an ethical dilemma happened in real research study. Morrison conducts a case study to investigate the life quality of obese adolescent boys (Morrison, Gregory and Thibodeau, 2012). He successfully built rapport with selected boys, and after finish the research he finds that two of them encounters emotional distress, one named MB gets a feeling like ‘breaking up’, and the other named EO has a feeling of angry and lashes out through emails (Morrison et al., 2012). From this example, it might be derived that despite the process is conducted according to ethical codes, there will be potential emotional problems. In addition, Corbin and Morse (2003) find that some participants are likely suffer from emotional distress after discussing a sensitive topic in unstructured interview, while others feel emotional release. Nowadays, the popularity of ethnography together with the occurrence of sensitive topics results in dilemmas of ethical implication (Laine, 2000). Furthermore, ethical codes are not adequate to solve the increasing ‘grey area’ problems (Laine, 2000). Also, Redmond (2003) indicates that the ethical codes should be varied with countries or cultures; there are no universal ethical rules. As a result, it might be concluded that ethical issue at some extend is a paradox, it requires researchers to follow its codes; whereas the codes are varied and always not sufficient as an appropriate guidance. Ethical dilemmas would be further revealed after examining the following three examples in sequence.
To begin, Russell (2005) carries out a research concerning student resistance to schooling, together with reflexivity and rapport issue implied, by using participant observation method. Russell successfully gains the permission to enter her chosen settings. However, she gets a problem to achieve the informal consent with teachers. Since Russell put her research information sheet in staff’s pigeonholes instead of holding an initial meeting with teachers, those teachers who had not read the information sheet introduce her in the class incorrectly. They label her as visitor, but actually Russell want to hide in the background. This then in turn affects her trust relationship with teachers. As Russell (2005:189) states:
“Some teachers saw me with disruptive students and consequently assumed I was there to mediate their behaviour. Thus, when I failed to intervene with troublemaking behaviour, I was sometimes frowned upon causing friction and confusion.”
This is consistent with what Willig (2008) indicates, the rapport relationship is delicate. Also, it shows the importance of getting informal consent, which is advocated by Mason (2002) and Roulston (2011). In contrast, Russell does well in getting informal consent of students. She fully discloses the research purpose and explains what is expected from them. Once the students ensured Russell would not inform teachers if they break school rules, Russell and students has successfully built a trust relationship. Although, a trust relationship is important in this research, Russell faces an ethical predicament, as she could not report to teachers when students break school rules, for example, when she sees students take illegal substances on school premises, she feels struggle. Furthermore, this predicament also influences the balance of her relationship with teachers and students. This is also a question associated with Russell’s identity. Russell as a white, female with northern Lancashire accent, aged 23-24, is a naive ethnographer. Her particular property has a two side effects. From one aspect, her naive makes students feel more close to her; from another aspect, she always encounters difficulties in managing relationship and dealing with unexpected incidents. In addition, Laine (2000) demonstrates that a new researcher might be failed to get an appropriate appearance, and thus one could feel anxiety. As Russell (2005) points out that she handles better in the third school than the previous two, because she gains more experience. Furthermore, Russell’s point of view (2005:194) :
“My gender, nationality, use of language, size and age influenced how students and teachers related and interacted with me. Rapport was gained faster with some and relationships changed and developed throughout the fieldwork, determining the type and quality of data collected.”
is supported by Dunne et al. (2005), as a researcher’s identity do affects the data collected. Finally, when the research is end, both students and teachers suffer from emotional upheaval, such as shocked, crying, and upset. Even Russell suffers when leaving the third school. This is the same emotional response with the case done by Morrison et al. (2012), and this is what expressed by Corbin and Morse (2003) as emotional distress.
Next, the second example is from Kamenou (2008), who investigates the experience of minority women balancing their work and person life, by using the method of in-depth interview. The first problem met by Kamenou is the difficulty to recurring samples, as one of the organizations withdraws from the study by claiming they are restructuring. This organization’s action is tricky here, because this study is aimed at investigating the equality issue, and to see if the advocated equality opportunity matches the reality. Rather than believes the excuse they use, maybe the reason that it is a sensitive topic for organization would be more reliable. This is a typical problem concerning access barriers from subjects (Laine, 2000). Another key issue in sample selection is that the author find ‘independent group’ through personal networking and no formal access is gained in participants’ organizations. Obviously, this independent group sample is highly possible associated with author’s identity. Thus, this sample might be less appropriate representative and results in data bias (Mason, 2002). Also, as no formal access gained, Laine (2000) argues that without a formal relationship and trust, people may provide less than candid answers as they do not treat it seriously. Moving forward, from reading the article, it probably could assume that the researcher is a white female. Therefore, the researcher and minority women are in different ethnicity, culture, religions and education attainment. These differences may lead to communication difficulties, for instance, the researcher and the researched might experience language variation (Oliver, 2010). The special identity of Kamenou compared with whom she investigates probably may impact the response of participants and thus affect the quality of data. The final point to note is the words used in this article. Kamenou (2008: s103) states:
“Ethic minority groups are not necessarily non-white, this study focus on non-white women, namely African, Caribbean, Indian, Pakistani and Bangladeshi.”
These words may make those participants or those non-white people feel uncomfortable. To use the appropriate words in the article is an ethical issue researcher should take into consideration (Oliver, 2010). Moreover, the researcher publishes names, ages and working positions of participants, also the two organizations’ names. This is considered as ethically inappropriate as names, ages are private information. Researcher should be conscious about the confidential of this information and it is also an issue related to informal consent (Oliver, 2010).
Then, the final example is from Brown, Humphreys and Gurney (2005), they conduct a case study at Laskarina Holidays, try to understand organizational identity through an analysis of shared identity narratives. The methods they use are in-depth interview, together with informal interview, observation and visual method plus documents. The unique visual method used by this research compared with the above two, has certain benefits. The combination of interview, visual method and documents, as well as other methods are called triangulation, the aim is to reduce bias and increase reliability (Thorpe and Holt, 2008). In addition, Mason (2002) speculates that visual method and documents enable one get an ontological and epistemological position, that means the evidence from visual method are part of social world and they prove the existence of themselves. So the collection of photos, documentation and internet pages, is a way to enrich ‘thick description’ and guarantee the quality of data. This is a standpoint supported by McNaughton (2009), who conducts a research with video recording. Another key issue concerning this research is the sample size, since thirty nine interviews were conducted only. In qualitative research, sample size are always much smaller than quantitative research. Instead of investigating thousands, qualitative research often merely focus on small scales (Willig, 2008). So whether thirty nine is a suitable number for the research is depends on the level of data adequacy provided by them. As Mason (2002) concludes that the sample should on the right focus, provide access to enough data for addressing the research questions. Besides that, there are practical problems one should consider, such as time, fund availability and the feasibility of samples. Moving one step further, the key issue concerning this research is about moral problem. A similar dilemma with the previous example is the publication of organization name and participants’ names in the article. Moreover, the sentence cited in this paper are mostly personal opinions, even some contain personal judgements on other people or certain things. For example, in the Epilogue section, they directly put two officers’ names in the paper, and state the situation after this research. There is a relatively high possibility that this part of information is published without informal consent of the company and the two officers. It is again a question related to privacy. According to Oliver (2010), informed consent process contains the discussion of confidentially, and oral promise by researchers is not enough. Additionally, Oliver (2010) suggests that explicit statement is necessary for participants to read, and participants need to be clear about the plans of retaining the data. Therefore, what this research present in the article is of high moral risk.
To sum up, qualitative and quantitative researches differ from their philosophy along the way down to data interpretation. Although qualitative research gains advantage in using textual evidence to deeper explores the underlying meanings of social society, it has drawbacks in prediction, comparison and integration. Therefore, currently more and more researches tend to integrate these two methods. The methods used in qualitative research face various challenges, from selecting sample to interpreting the data. Those challenges affect the research process as well as the outcomes. The most notable two issues are reflexivity and ethical dilemmas. There is no perfect solution for both of them. What researchers could do is to predict in advance and endeavour to minimise the impact of them. Researchers need to warn in their mind all the time about these issues, especially moral problem. Incontrovertibly, the debate concerning challenges of qualitative research will continue, and researchers will make efforts to improve.