Qualitative research is naturalistic inquiry, because the data collection strategies used are interactive to discover the natural flow of the events and processes. Most qualitative research deals with people's individual and collective social actions, beliefs, thoughts, and perceptions.
In this session, we will focus on designing qualitative research: its sampling strategies, phases of data collection and analysis, validity and reliability issues, and finally the qualitative research standards.
In qualitative research, the research sites and participants are selected following a strategy called purposeful sampling. Purposeful sampling, in contrast to probabilistic sampling, is "selecting information-rich cases for study in depth" (Patton, 1990, p. 169) when one wants to understand something about those cases without needing or desiring to generalize to all such cases. Types of purposeful sampling include site selection, comprehensive sampling, maximum variation sampling, network sampling, and sampling by case type.
|Site Selection||Select site where specific events are expected to occur|
|Comprehensive sampling||Choose entire group by criteria|
|Maximum variation sampling||Select to obtain maximum differences of perceptions about a topic among information-rich informants or group|
|Network sampling||Each successive person or group is nominated by a prior person as appropriate for a profile or attribute|
|Sampling by case type|
|Extreme-case sampling||Choose extreme cases after knowing the typical or average case-e.g., outstanding successes, crisis events|
|Intense-case sampling||Select cases that are intense but not extreme illustrations|
|Typical-case sampling||Know the typical characteristics of a group and sample by cases|
|Unique-case sampling||Choose the unusual or rare case of some dimension or event|
|Reputational-case sampling||Obtain the recommendation of knowledgeable experts for the best examples|
|Critical-case sampling||Identify the case that can illustrate some phenomenon dramatically|
|Concept/theory-based sampling||Select by information-rich persons or situations known to experience the concept or to be attempting to implement the concept/theory|
|Combination of purposeful sampling strategies||Choose various sampling strategies as needed or desired for purposes, especially in large-scale studies and lengthy process studies|
Qualitative phases of data collection and analyses are interactive research processes that occur in overlapping cycles. There are no rigid step by step procedures to follow, both data collection and analysis progress simultaneously toward the completion of the research project.
Validity refers to the best available approximation to the truth of propositions. There are two types of validity according to Campbell and Stanley (1963). Internal validity refers to the approximate validity with which we infer that a relationship between two variables is causal. External validity refers to the approximate validity with which we can infer that the presumed causal relationship can be generalized to and across alternate measures of the cause and effect and across different types of persons, settings, and times. Although validity, i.e. internal and external validity, is developed in the context of quantitative research, in order to ensure compatibility, people still use the same terms for qualitative research but not necessarily with the same interpretations.
In qualitative research, internal validity is affected by qualitative research designs, and external validity is related to the extension of qualitative findings.
Qualitative research design validity concerns these questions: Do researchers actually observe what they think they observe? Do researchers actually hear the meanings that they think they hear? Thus, the internal validity of qualitative research is the degree to which the interpretations and concepts have mutual meanings between the participants and the researcher.
The following strategies are used to enhance the validity of qualitative research design:
|Prolonged and persistent field work||Allows interim data analysis and corroboration to ensure the match between findings and participant reality|
|Participant language; verbatim accounts||Obtain literal statements of participants and quotations from documents|
|Low-inference descriptors||Record precise, almost literal, and detailed descriptions of people and situations|
|Multiple researchers||Agreement on descriptive data collected by a research team|
|Mechanically recorded data||Use of tape recorders, photographs and videotapes|
|Participant researcher||Use of participant recorded perceptions in diaries or anecdotal records for corroboration|
|Member checking||Check informally with participants for accuracy during data collection; frequently done in participant observation studies|
|Participant review||Ask each participant to review researcher's synthesis of all interviews with the person for accuracy of representation; frequently done in interview studies|
|Negative cases or discrepant data||Actively search for, record, analyze, and report negative cases of discrepant data that are an exception to patterns or that modify patterns found in the data|
In order to maximize the external validity in qualitative research, the intent is to provide for the extension of findings, as grounded theory or as an analytical synthesis that enables others to understand similar situations and apply these findings in subsequent research. Knowledge is produced not by replication but by the preponderance of evidence found in separate case studies over time or in more structured quantitative designs. The following ten design components affect the extension of qualitative findings:
|Strategy||Adequate Description in the Study|
|Research Role||The social relationship of the researcher with the participants|
|Informant selection||Criteria, rationale, and decision process used in purposeful sampling|
|Social context||The physical, social, interpersonal, and functional social scenes of data collection|
|Data collection strategies||The multimethods employed, including ethnographic observation, ethnographic interview, and documents|
|Data analysis strategies||Data analysis process described|
|Authentic narrative||Thick description presented as an analytical narrative|
|Typicality||Distinct characteristics of groups and/or sites presented|
|Analytical premises||The initial theoretical framework or concept that informs the study and others added as the study progresses|
|Alternative explanations||Retrospective delineation of all plausible or rival explanations for interpretations|
|Other criteria by purpose (after study completed)|
|Phenomenology||Understand the complexity of the situation; generates additional case studies|
|Grounded theory||Concepts or mini-theory relate to social science; generates verification research with more structured designs|
|Critical traditions||Historical revisionism and situatedness to erode ignorance; empowers participants; generates further research; action stimulus|
Even though reliability is not quite applicable in qualitative research, a different criterion, disciplined subjectivity is proposed. Disciplined subjectivity is the researcher's rigorous self-monitoring, continuous self-questioning and reevaluation of all phases of the research process. There are following strategies to minimize research bias, thus to reduce the subjectivity:
|Peer debriefer||Select a colleague who facilitates the logical analysis of data and interpretation|
|Field log||Maintain a log of dates, time, places, persons, and activities to obtain access to informants and for each data set collected|
|Field journal||Record the decisions made during the emerging design and the rationale; include judgments of data validity|
|Ethical considerations recorded||Record the ethical dilemmas, decisions, and actions in field journal|
|Audibility||record data management techniques, codes, categories, and decision-rules as a "decision trail"|
|Formal corroboration of initial findings||Conduct formal confirmation activities such as a survey, focus groups, or interviews|
Qualitative designs are judged by several criteria. Below are typical questions that researchers might ask of their designs or that reviewers may use to critique a qualitative design: