At the very heart of a descriptive study is the observation of an environment or analysis of data without affecting the nature of the situation under scrutiny. So we can analyze using surveys, ethnographies, case studies, prediction studies, etc..., all of which are qualitative descriptive studies. The strength of this study comes from the observance of a specific situation with clearly defined variables, and of course the weakness of this type of study comes from analysis that can not be replicated because, for example, it may lack explicit instructions, or from coders who do not agree beforehand on the importance of category development, or from researchers who do not question the theoretical assumptions underlying their research, all of which weakens the significance of the findings.
The purpose of case studies is to identify important aspects or variables in natural conditions through the study of individuals, small groups, or whole environments. It is not cause and effect, but rather allows researchers to form theories and hypothesis from observing a small number of subjects in the natural environment. Subjects can be selected, for example from the classroom of a researcher, of from the interviews of researchers. In Lauer & Asher, Emig chose 8 subjects from various types of schools, and Graves selected 8 students from a larger study. In keeping with the small nature of these studies, Deborah Brandt interviewed over eighty people born between 1895 and 1985 and then chose two subjects born sixty-eight years apart to study - both of whom had striking parallels in their lives.
The data is collected a variety of ways, including letters, speeches, TV shows, essays, etc. Emig, for example, collected conversations, tape recordings of students composing aloud, accounts of processes, discrete observations of composing, writing samples, and school records (L&A 26). Graves used folders (these contained tests of intelligence and reading, assigned and unassigned writing, study records and observances of the student both in and out of school) and observations of 53 writing episodes, in addition to interviews. Overall, the researcher of this type of study uses the data to seek out patterns and identify operationally definable variables, and then asks how they relate to one another (L&A 27).
A major aspect of case study analysis is content analysis (which includes coding, or the setting up and labeling of categories and variables). Content analysis, according to Lauer & Asher, is a "major measurement procedure" that allows researchers to substantiate that materials and observations are ultimately quantifiable (27). Here communication data is analyzed, patterns are noticed, and variables are identified. The kinds of generalizations possible include extensive descriptive accounts. These studies should optimally "relate their findings to the research of others," thus increasing their ability to generalize about their study (L&A 33), and as with Graves, furnish questions for subsequent research, which is an approriate and valuable feature of qualitative study. Deborah Brandt notes that her aim is to "extract from their experience lessons that can be applied to literacy learning in other economic configurations - not in order to predict particular outcomes, but to understand better the struggles that economic transformations bring to the pursuit of literacy" (377). So, in essence, she notes that her study is not a cause and effect, but rather an observation that allows a hypothesis that can create further understanding.