Provide detailed descriptions of data collections and data analysis: e.g. Sensemaking theory: how does sensemaking come into play in Try to ground constructs in the data and analysis, e.g. What else could you have done and why did you not do so? Pay attention to alternative explanations: e.g. Describe and explain case selection: sites, participants, events: What kind of case? What is your sampling strategy? Clearly articulate the connection between theory and data: Deductive or inductive approach? Process continues until new cases support the hypotheses Means of assuring research validity: Constant comparative method Researcher should always attempt to finde more cases to test the initial hypotheses Means of assuring research validity: Deviant cases The researcher should always look for and address deviant cases Means of assuring research validity: Comprehensive data treatment Comprehensive description of a specific social phenomenon Criteria for designing and evaluating qualitative research - Situate the research in appropriate literature: e.g.What do we know about the topic/phenomenon? If hypotheses is not confirmed hypotheses is revised Initial explanatory hypothesis is derived Comprehensive data treatment Means of assuring research validity: Respondent validation Bring your research findings back to the people you have studied to see whether the finding is conform to their own experience Means of assuring research validity: Analytic induction Identify essential characteristics of the phenomenon studied: Originated in quantitative research Means of assuring research validity - Triangulation: using multiple methods to get a more accurate, comprehensive representations of the object of study Low-inference descriptors: Tape recordings, careful transcription, presentation of long extracts of data in the report to get all pre-leading factors Research Validity - Accuracy or correctness of the findings At least two researchers analyse independently the same data

Field note conventions Standardized ways of doing field notes, which increase the use of low-inference description e.g.: Verbatim quotes ' ', Paraphrases ' ', and Contextual data or fieldworker's interpretations ( ) Reliability in terms of texts - As the data already exists, texts are often seen as more reliable than observations Reliability in terms of interviews - Interviews can be vey unreliable

Reliability for qualitative research