This essay describes the main aspects of relativist qualitative research in social psychology. Generalised comparisons will be considered between the methodology of relativist qualitative and experimental research, reflecting their different epistemology’s (what can be known about human behaviour), via issues of validity, reliability and generalisability.
In accordance with a logical positivist epistemology, experimental researchers argue that social behaviour is an objective, independent reality, relying on theories which describe that reality (Megee, 2001). Theories depend on constructs and variables, and the relationship between these constructs/variables. Theories are formulated either deductively (based on logic) or inductively ( based on observation) or through a combination of these two approaches (Megee, 2001).
Methods are quantitative and attempt to understand some reality by measuring hypothesized relationships between constructs via isolation of these entities into independent and dependent variables (Smith, 1996).
Although qualitative research is an umbrella term for both relativist and realist study, this essay will focus on the former. Relativist qualitative researchers disregard positivist logic (Hammersley, 1996). They instead affirm an idealist epistemology, endorsing relativism and social constructivism (Megee, 2001). Research is based on description rather than prediction as meaning of a phenomenon is sought instead of cause. Interpretations are typically derived through discourse analysis (Megee, 2001),
Realist qualitative researcher epistemology emanates from quantitative realist and essentialist philosophy, where real underlying entities can be denoted from data (Megee, 2001). However, repute inferences should be drawn from multiple methods of data collection. Robson (1993, as cited in Megee, 2001) proposed criteria for assessing realist qualitative research based around; credibility, verified through the utilization of multiple evidence sources and the provision of comprehensive detail concerning the data collection conditions. Dependability of data collection ensures that findings would be repeated in similar settings. Finally, transferability considers the relevance of results from one situation to another determined on an individual case basis (Mcgee, 2001).
The main features of Qualitative research
In social science we have the problem that social systems are very complex, and it is not possible to identify all relevant variables and constructs and include them in a comprehensive theory of the phenomena under study. Relativist qualitative research tries to address this problem by rejecting the concept of an objective reality, or the notion of underlying variables/constructs that can be generalised across settings.
Relativist qualitative researchers believe language constructs behaviour rather than being a tool to discover some reality behind discourse (long descriptive text using participant dialog) (Gill, 1996). Believing potentiality to understand a phenomena will be decreased through reducing people down to variables, research aims to understand complex social phenomena through discovery of underlying motivations, feelings, values, attitudes, and perceptions that construct cognition (Potter, 1996).
Components of a phenomena are studied within its contextual setting e. g., case studies (Megee, 2001). Methodological principles of interpretation to understand realty within are approached through discourse analysis and theoretical backgrounds (Nikander, 1995). Interest is focused around how speakers deploy interpretative repertoires (way of speaking about specific social behaviours) and how idealogical dilemmas are resolved through oratory . A third accent is upon the analysis of discourses, believing discourse is not inside an individual but within a community and how the role of language maintains forms of social practice. Hence, researchers make no pre-conceived hypothesis, enabling hypothesis to emerge from discourse data instead (Nikander, 1995).
Researchers endeavor to curtail any data loss in translation by ensuring absolute accounts of participant speech and behavior are documented. Attempts are also made to diminish data reduction at the time of acquisition, e. g., refrain from categorization (Willig, 2000).
While quantitative research attempts to avoid subjective influences, qualitative researchers target the particular setting, individuals and subjective views, which they view as critical in their construction of social behaviour.
Evaluation of qualitative compared with experimental research
Validity is primarily a positivist concept, based solely on clearly defined hypotheses verifiable through rigorous objective observations. Internal validity is determined by the degree to which research actually evaluates the association between variables (Megee, 2001). Positivists also advocate reliability of measurement, that is the extent an instrument delivers consecutive information on separate occasions(Megee, 2001).
Researchers of this discipline are critical of qualitative validity and reliability of measurement. They argue qualitative research can not rigorously examine the detailed structures underlying complex natural interactions with subjectivity leading to procedural problems and researcher bias unavoidably built in. Empirical observations and interpretations are essentially subjective and therefore inaccurate (Toren, 1996).
Reliability for qualitative research is poor as two researchers may arrive at different conclusions based on the same behaviour sample. Unsystematic methodology and absence of accessible observed data provide difficulties in replication of the setting. Inadequate criteria as to what would constitute the researched behavior also contributes difficulties for others investigating the same phenomenon to replicate the research (Megee, 2001).
The qualitative researchers principle of inconcludability and reflexivity, as proposed by Woolgar (1988, cited in Megee, 2001) renders these criticisms redundant. He suggests social behaviour can not be reduced to manipulated equations to produce definite understanding as there are numerous contexts allowing countless implications. Any analysis is unfinished and open to development by other researchers (Megee, 2001)
Qualitative researchers argue they have greater validity as observations of a phenomena are in a natural setting allowing researchers to develop a more accurate understanding. Research produces much richer data as it is more intensive and flexible avoiding rigidly definable variables. Also, experiments fail because they can neither adequately define nor accurately measure enough of the variables to understand natural complex interactions. Experimental researchers would find it almost impossible to measure certain phenomena ie, self concept, which could be adequately attempted by qualitative researchers (Hammersley, 1996).
A deep understanding of a phenomena can be achieved through acknowledging both contradictions and confirmations of an interpretation allowed via participant feedback (Willig, 2000). Research is sensitive not to overpower the opinions of the people involved or obscure the relationship between the researcher and participant as the interpretation is shared through feedback and any prejudice from the personality of the researcher is reduced. However there is no guaranteed accuracy of participant interpretation as problems can occur if a participant disagrees or is persuaded to agree with the researchers interpretation through rephrasing or interpersonal relations (Smith, 1996). Also, feedback can be unintentionally contaminated through both the false consciousness of participant delusions of reality (Megee, 2001) or intentional falsehoods.
A qualitative theory is likely to be formed upon the affinity between multiple studies rather than in single isolation, as is the case for experimental research. Any theory is then left open for other researchers to build on. Readers have access to the original raw data and can themselves evaluate the interpretation, which is not the case in experimental research where readers have to rely on the researchers interpretations . Deviant case analysis also increases qualitative validity as exceptions are shown to be different through counter examples (Potter, 1996).
Reliability equates to reflexivity in relativist qualitative research, bending back the research process on itself. Every component of the research process is examined (Willig, 2000). Researchers continuously review and study themselves as they are sensitive to their own role in the interpretation of data (Megee, 2001) thereby weakening their impositions to gain a more authentic understanding.
By allowing theories to form through what people say and do, qualitative research does not impose theories upon participants. After a period of immersion in a situation the qualitative researcher is better able to draw hypotheses which emerge from interview and observation in a natural setting.
Generalisability
Quantitative research believe findings are externally valid if they can be generalised and implemented to the real world.
Relativist qualitative researchers endorse indexicality rather than generalisability. Interpretations from data can only be purposeful in that time and context and can not be directly employed to another situation. (Megee, 2001). Research has greater ecological validity than experimental research as naturalistic data is gathered, rather than applying results to the real world from manipulated variables (Willig, 2000).
Qualitative researchers typically attempt to comprehend the internal dynamics of a case by working with individuals or small groups making representativeness redundant. According to quantitative researchers, complications of qualitative analysis can be encountered if research examines behaviour which is significant to a larger population than those participating in the study. Advocates of qualitative research argue, ‘ if a given experience is possible, it is also subject to universalisation’ (Haug, 1987, cited in Willig, 2000) .
Data drawn through implementation of accumulative techniques enables contributions from multiple related sources, allowing richer interpretations to be derived by means of data integration (Willig, 2000).
Conclusion
Experimental positivist epistemology of realism attempt to find the causes of a phenomena through manipulating variables to test a hypothesis. Advantages of objectivity and reliability allow more scientific conclusions to be drawn. However results are disadvantaged through artificiality and some phenomena are impossible to measure.
Relativist qualitative epistemology of idealism tries to discover the meaning of a phenomena by looking at linguistic constructs, with the hypothesis emerges from data. Comprehensive research provides a rich, in-depth understanding of complex behaviour from a natural context. Methods are not limited to rigidly defined variables making examination possible of complex phenomena which often elude quantitative methods. Limitations are evident with problems of poor reliability, difficulties of replicability and subjectivity with inbuilt researcher bias. The comprehensive approach to data gathering proves expensive, time consuming and limits scope.
Although founded upon different epistemological bases, we are most likely to improve our understanding of social behaviour by combining the best elements of the research methods of qualitative and quantitative research by appropriating different methods according to purpose and circumstance.(Hammersley, 1996).
