One of the first tasks of an instructor is to systematically expose students to scholarship describing careful and rigorous qualitative research in psychiatry. The instructor can choose from many seminal studies that have contributed to the transformation of psychiatry and psychiatric services. Focusing on these studies indicates to students that, when judiciously employed by thoughtful researchers, qualitative methods can produce important findings that can benefit patients and practitioners on the ground.
Many of these seminal studies relate to psychiatric institutionalization. One is Goffman’s (
3) magnum opus "Asylums," another Rosenhan’s (
4) classic paper "On Being Sane in Insane Places" (see Table 2). These qualitative studies describe and analyze some of the depersonalizing and stultifying effects of psychiatric institutionalization. They strongly influenced the movement for deinstitutionalization, and the transformation toward community care. Another seminal ethnography is Estroff’s (
5) study of recently discharged people with severe mental illness. The title richly chronicles the trials and tribulations of living in the community in the context of fragmented services.
Another domain reliant on well-controlled and rigorous qualitative research is cultural psychiatry. Kleinman (
6) has shown the cultural relativity of many taken-for-granted psychiatric concepts through utilization of in-depth interviews and observational techniques. His work on "explanatory models" and the "illness experience" has focused attention on social and cultural correlates of mental illness.
Another example comes from Antonovsky (
7). His qualitative study of stress and coping among holocaust survivors raised awareness of medicine’s "pathological emphasis," which tends to ignore the multifarious ways people live meaningful lives in the presence of adversity. This work somewhat presaged the recovery paradigm, now common in psychiatry. This paradigm is frequently investigated through qualitative methods, and again is worthy of class discussion (
8).
In discussing these seminal works, students learn that qualitative research has a strong tradition in psychiatry, and that it has made a valid contribution to the discipline’s development. It is not a fly-by-night trend, and should not be glibly dismissed as the soft underbelly of research methods. That said, it is worth emphasizing to students that much qualitative research is of questionable quality. However, this usually reflects lack of judgment, training, experience, or interpretive acuity in the investigators, rather than something inherently defective in the method itself. Examples can be found in various psychiatry journals, and can be discussed in class.
Some students may desire to become better accustomed to the practical conduct of qualitative research, in order to design, execute, or assist in qualitative studies. To address such desires, further sessions can be developed that focus on developing practical skills needed to conduct small-scale qualitative research studies in psychiatry.
In teaching such sessions, it is essential to communicate that the structure of qualitative research, if done rigorously, is largely similar to that of quantitative research. All of the same a priori thought and discussion that goes into the development and design of a quantitative study must also be the case for a qualitative study. Like quantitative research, qualitative research is a detailed endeavor, requiring adherence to certain methodological canons. Namely, an innovative research question should be formulated in the light of literature review. A study should be designed that is capable of elucidating the research question. This should take into account the financial, temporal, and human resources available to the investigator. The study should be executed in a timely manner. Incoming data should be interpreted and written up for publication. Stringent training and quality assurance procedures should be in place throughout. All studies should have professional qualitative experts in supervisory or consultancy roles, just as quantitative studies often require professional biostatistical expertise.
Three major differences exist when comparing qualitative to quantitative research. It is important for students to know and understand these differences. These relate to hypothesis-testing, sample size, and design modification. These, and other basic differences between qualitative and quantitative research, are briefly summarized in Table 3.
Qualitative research is rarely hypothesis-driven, but is generally inductive in orientation. Data are explored in light of a research question rather than tested against a predefined hypothesis. This is because qualitative research generally aims to broadly understand a phenomenon of interest (e.g., the impact of psychiatric institutionalization), rather than test an extant theory. The second difference relates to sample size. Though a few studies have large numbers of participants (mostly cross-site), in general qualitative studies have a small sample, commonly between 20 and 40 participants. Some studies have fewer than 10 participants. The careful and judicious study of lived experience in fewer people is considered optimal in qualitative research as it reaches an intimate depth of knowledge unattainable through other methods. Finally, it should be noted that interim analysis and subsequent modification of design is encouraged in qualitative research. This allows the researcher to hone in on an area of interest that may not have been obvious from the outset.
Fortunately, there are many pedagogic resources that can be used to illustrate these points. Numerous qualitative research textbooks exist which can be recommended to students depending on their level (
9,
10). These give helpful advice vis-à-vis study design, sampling, recruitment, execution, and analysis. Likewise, students can be encouraged to examine recent peer-reviewed qualitative articles in "journal club" manner. These describe how studies have been conducted on the ground, providing real-world examples of factors such as sample size and analytic strategies.
Indeed, analysis and interpretation may be the most challenging parts of qualitative research, and this should be imparted to students lest they approach it lightly. Students should be made aware of the congeries of analytical techniques, such as grounded theory, phenomenology and hermeneutics, and semiotics. However, in introductory courses it may be best to focus on generic analytical strategies such as content analysis. Students should also understand that analysis is, in temporal terms, lengthy. Audio recordings need to be transcribed and should be read (and listened to) many times. Transcripts will need to be coded. Computer assisted qualitative data analysis software may be useful; however, this is an organizational aid and does not conduct the analysis in itself. The analyst still needs to categorize, code, and link data, without losing sight of the overall dataset, and indeed the wider background literature into which it should be ensconced.
There are two common analytic pitfalls that instructors should discuss with students. First, inexperienced researchers may risk coming to premature closure after reading only a subsample of transcripts. Second, they may feel drowned in data and unable to engage in thematic reduction. Teaching strategies to overcome these dangers in the abstract is difficult, and may be best saved for mentoring sessions with students conducting pilot projects.
Likewise overall interpretation of data, in the context of the research question and the wider literature, is a skill that may best be learned in contact with real data, rather than through abstract classroom discussion. One useful strategy here is for instructors to ask students to read research papers written by the instructors themselves. Instructors can then talk students through the "behind the scenes" data analysis and interpretation that led them to formulate the results and conclusions. This allows students to interrogate the researcher’s decision-making processes, thematic reductions, and final resolutions. If the instructors do not have qualitative research papers to discuss, they may invite more experienced colleagues from elsewhere to present and discuss their own work.