between raw data and abstract theory: All abstract ideas should be tied to specific parts of the data. In these approaches to qualitative data analysis, it is not adequate for researchers to report on “general impressions” from the data; this would be considered to have low validity. Instead, memos should identify the lines in the field notes or other memos to which they relate so that there is a clear link between bits of data and more abstract conclusions. The memos should also identify what codes or concepts they refer to and be dated, so that it is clear when the thoughts occurred in the research process. Some of the memos, or at least parts of them, may later be incorporated into the final research report.
Beyond saying that memos are abstract and general, it is difficult to specify their content because it is so highly variable. A memo might clarify an idea or an existing coding category, suggest some new coding categories or subcategories, or link data from different parts of the field notes or possibly even from other research projects. A memo might propose a hypothesis or a new pattern code, or it might identify something puzzling that does not fit in with the conceptual framework emerging in other memos. As the memoing process continues, what begins to emerge, either gradually or in some cases quickly, is more formalized and coherent set of propositions or explanations of what is found in the data. We discussed the development of propositions and theories at length in Chapter 2. You can see here how the propositions in qualitative data analysis emerge out of the data in a very inductive process. In fact, some qualitative researchers eventually force themselves to state some tentative propositions, which can then be reviewed in light of all the memos, coded, and data to see if the propositions have any validity. In some cases, propositions may emerge early in the process of data analysis. Some propositions will eventually be discarded, as they are shown to be inconsistent with much of the data. But eventually a set of propositions emerges that seems to be consistent with all or most of the observations. Propositions can take a number of different forms. Such as :