Qualitative research uses non-numerical data to explain what, why, and how something happened. It provides a contextual understanding of people, behaviors, and situations that quantitative studies often can’t.
There are ongoing conversations in the scientific community about what a “good” qualitative study entails. While the specific criteria are subject to debate, there are some widely accepted guidelines to consider. These include:
- Use of theory to inform research questions, design, and interpretation of findings.
- Explaining the decision-making process behind choosing study design, methodology, and sampling.
- Maintaining study quality through transparency and systematicity of the research process, which includes establishing reflexivity, validity, reliability, and generalizability of the research.
Below, each of these are described further.
Researchers should provide a theoretical or conceptual framework that motivates their study’s research questions, data gathering, and interpretation of results. Research questions should add to existing theory, and sometimes also generate new theory. This article in SAGE provides introductions to some of the theories used in qualitative research.
Study Design, Methodology, and Sampling
Qualitative research should discuss why the study’s design and methodology were chosen. Below are some common approaches, with consideration of inherent tradeoffs.
- Interviews: Interviews allow data to be collected from individuals through unstructured and/or semi-structured questions, with flexibility to gain unanticipated knowledge. The researcher has the freedom to structure the interview to be as formal and in-depth based on the needs of the study. Interviews provide information based on what people say but cannot provide insight based on direct observation of how people behave or interact. They can also be time consuming while conducting interviews and during the transcription and analysis process. Transcription costs also increase as the number of participants increase. On the other hand, the amount of time required can be estimated during study design (e.g., one hour per interview times the number of anticipated interviews).
- Groups: Focus groups, panels, and other group-based interviews provide a larger venue for capturing both verbal information and observations of group dynamics and interactions, including how individuals influence each other. Disadvantages include more difficulty getting truthful responses (since sensitive topics may be more challenging to discuss in a group), social desirability biases, and groupthink. Like interviews, focus groups can also be time consuming and costly during the interviewing, transcribing, and analysis process.
- Ethnography: Ethnographic research involves observation of people and culture, allowing for description of what individuals do in their natural environments instead of controlled settings. This method could require a large investment of time, since it is unclear how long it may take to capture and understand authentic behaviors.
- Content/Document Analysis: Content and document analysis reviews a multitude of secondary sources, ranging from written accounts to recorded media. Content analysis examines all sources in which a searched term appears, while document analysis focuses only on written documents. Both methods are particularly good for studying questions of historical significance, including those where participants are no longer living. A significant limitation is that the data were initially collected by others, potentially with unknown biases, errors, and/or omissions.
A qualitative study should also explicitly discuss its sampling strategy, especially if convenience sampling is taking place since it can result in selection bias and sampling error. The goal of sampling should be to reach theoretical saturation, the point at which more participants or documents would not add more unique information.
A qualitative study’s design should be both transparent and systematic. Transparency means providing a clear description of all techniques and processes used to collect, analyze, and interpret data. Systematicity means the use of data collection and analytic methods follow widely accepted research processes and qualitative research guidelines.
A high-quality qualitative study should also address validity, reliability, researcher reflexivity, and generalizability.
- Validity: Validity pertains to how a study’s results would apply to similar populations and settings outside of the study. Some methods to address validity include utilizing description to provide behavioral meaning and context, triangulating (by utilizing multiple research methods, sources, or researchers), and negative case analysis by reviewing contradicting data to the study’s findings.
- Reliability: A study is reliable when stable and consistent results are produced when the research methods are replicated. Researchers should be transparent about their methods and justify them, clarify how the study could be replicated, and ensure data accuracy.
- Researcher Reflexivity: Researchers should reflect on their own biases, judgments, and belief systems that may have impacted data generation and analysis. This is typically addressed in a statement from the researcher.
- Generalizability: The need for generalizability in qualitative research is debated amongst experts. It focuses on how well study findings would transfer to different settings and populations. While study replication would most likely produce some new findings, key concepts and themes can be identified that are reasonably expected to apply across contexts, as well as indicate what it is about their findings that may not generalize. This concept is contested since qualitative research may not always aim to be generalizable but rather descriptive about a certain group or individuals.
By considering these factors and general guidelines, researchers can be more mindful of their own research and when reviewing existing qualitative studies, while reviewers can better assess new qualitative research.