Data Analysis and Validation

• Cresswell has a section in his chapter on Data Analysis and Validation Procedures that I think is pretty good.
• You can convert qualitative data to quantitative by merely categorising.
• This means that the annotator (researcher) goes through textual data (or film or something else) and says this bit is category A, this bit category B.
• This data is then quantitative and can be used appropriately.
• This leads to problems that the annotator might not be very good, and thus the quantitative data is flawed.
• Multiple annotators and inner-annotator agreement data can help here.
• Exploring outliers also sounds interesting. If you have some data that is really unusual (qualitative or quantitative), this may be particularly interesting to explore.
• Instrument development seems pertinent to mixed methods. Use a qualitative study as the basis for developing a survey.