Selecting the data to be used in a qualitative study is one of the first steps in the research process. Once a researcher determines what to investigate, he will then determine what he will need in terms of data, to make his conclusion. For instance, if a researcher intends to investigate the nature of interactions among high school students, his data may be collected via the completion of a researcher created rubric or through the completion of student surveys.
The researcher creates the rubric for researcher observation based upon his research questions. For example, if the researcher is investigating the interactions of high school students, he will include on the rubric or on the survey items that align with his research questions or intended targets of investigation. Suppose in his investigation, he will attempt to determine who most often initiates social interactions among high school students. He will include on the rubric categories for the recording of what is observed relative to the initiation of social interactions. If he is structuring a survey, he'll include questions to provide insights on the same issue.
Qualitative data analysis can be objective or subjective. More often than not, it is a fairly subjective process. The first step in analyzing qualitative data is to comprehensively review the data set before attempting analysis. This is important because it will help the researcher gain an overall picture of the data set before beginning analysis. After the review, the researcher can then code the data set and identify themes for easy and consistent analysis. His codes will allow him and his fellow researchers to group like patterns, ideas and themes into categories for interpretation.
Once the review and coding is complete, the researcher can then use the coded data to make inferences. This is performed by reviewing each code to look for themes or patterns. To continue with the high school student example, when the researcher reviews the survey, he may note that most males indicate X and most females indicate Y on a particular item. Additionally, he determines that the older the female is, the more likely it is that she will indicate X. Coded data makes interpretation much easier, as the researcher is able to disaggregate the data set and narrow further the analysis of each item.