Studies of Welfare Populations: Data Collection and Research Issues. Coding Issues

06/01/2002

Qualitative research of any kind--open-ended questions embedded in surveys, ethnographic interviews, long-term fieldwork with families or "neighborhood experts"--generates large volumes of text. Text files may derive from recorded interviews, which then must be transcribed verbatim (a costly and time-consuming proposition), or from field notes that represent the observer's account of events, conversations, or settings within which interactions of interest routinely occur. Either way, this material is generally voluminous and must be categorized to document patterns of note.

Anthropologists and qualitative sociologists accustomed to working with these kinds of data have developed various means for boiling them down in ways that make them amenable to analysis. At the simplest level, this can mean developing coding schemes that transform words into numeric representations that can be analyzed statistically, as one would do with any kind of close-ended survey data. Turning to the Urban Change project, for example, we find that initial baseline open-ended interviews show that respondents are hoping that going to work will enable them to provide a variety of opportunities for their children. Mothers also report that they expect their social status to rise as they depart welfare and note that their children have faced taunting because of their participation in AFDC; they trust the taunting will cease once they are independent of state support. These findings come from tape recorded interviews intended to capture their prospective feelings about moving into the labor market some 2 years before the imposition of time limits. These responses can be coded into descriptive categories that reflect the variety of expectations respondents have for the future, or the hopes they have expressed about how working will improve their lives.

Most qualitative interview instruments pose open-ended questions in a predefined order. They also may allow interviewers some latitude to permit informants to move the discussion into topic areas not envisioned originally. Within limits, this is not only acceptable, but it is desirable, for understanding the subjective perspectives of the respondents is the whole aim of this kind of research and the instrument may not effectively capture all the relevant points. However, to the extent that the original format is followed, the coding can proceed by returning to the responses that are contained in approximately the same "location" in each interview transcript. Hence, every participant in our study of the working poor under welfare reform was asked to talk about how their neighborhood has changed in the past 5 years. Their responses can be categorized according to the topics they generally raised: crime declining, gentrification reflected in rising rents, new immigrant groups arriving, and so forth. We develop codings that reflect these routine responses in order to be able to draw conclusions such as "50 percent believe that crime has declined precipitously in their neighborhood" or "20 percent object to police harassment of their teenage children."

However, we also want to preserve the nuances of their comments in the form of text blocks that are "dumped" into subject files that might be labeled "attitudes toward the police" or "comments on neighborhood safety." Researchers then can open these subject files and explore the patterned variety of perspectives on law enforcement or the ways in which increasing community safety have affected the patterns of movement out of the home or the hours that mothers feel comfortable commuting to work. When qualitative researchers report results, we typically draw on these blocks of text to illustrate the patterns we have discovered in the data, both to explore the nuances and to give the reader a greater feeling for the meaning of these changes for the informants. To have this material ready at hand, one need only use one of a variety of text-processing programs, including Atlas.ti, Nud.ist, and Ethnograph, each of which has its virtues.(7) Some proceed by using key words to search and then classify the text. Others permit the researcher to designate conceptual categories and then "block" the text with boundary markers on either side of a section so that the entire passage is preserved. It is even possible to use the indexing capacities of standard word-processing programs, such as Microsoft Word 6.0 and above, which can "mark" the text and dump it into subject files for later retrieval.

Most qualitative projects require the analyst both to digest the interviews (which may be as long as 70 pages or more) into subject headings and to preserve the flow of a single informant's interview through summaries that are preserved by person rather than by topic. I typically maintain both kinds of qualitative databases, with person-based summaries that condense a 70-page text to 5-6 pages, offering a thumbnail sketch of each interview. This approach is of primary value to an academic researcher, but it may not be as important to practitioners who may be less interested in life histories for their own sake and more concerned with responses to welfare reform per se.

View full report

Preview
Download

"01.pdf" (pdf, 472.92Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"02.pdf" (pdf, 395.41Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"03.pdf" (pdf, 379.04Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"04.pdf" (pdf, 381.73Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"05.pdf" (pdf, 393.7Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"06.pdf" (pdf, 415.3Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"07.pdf" (pdf, 375.49Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"08.pdf" (pdf, 475.21Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"09.pdf" (pdf, 425.17Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"10.pdf" (pdf, 424.33Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"11.pdf" (pdf, 392.39Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"12.pdf" (pdf, 386.39Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"13.pdf" (pdf, 449.86Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®

View full report

Preview
Download

"14.pdf" (pdf, 396.87Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®