A. Huberman

Drawing valid meaning from qualitative data: some techniques of data reduction and display.

Several major points are made about techniques of data reduction. 1) Data reduction or data transformation occurs at all points in a study from design to data collection and write-up. It is analysis of a form which organizes and clarifies data in such a way that final analysis can occur coherently. 2) Data reduction and data analysis have as an indispensable accompaniment some form of data display and the display modes chosen will inevitably condition the processes and conclusions for analysis. The most frequently adopted and typical display mode for qualitative data the narrative text is also the most cumbersome. Matrices and figures of several sorts can be generated that meet the need to display data coherently and compactly. 3) Single-site and multisite analytical processes are complex but not arcane or obscure. 4) It is possible to understand processes if they have been documented accurately using some reasonably standardized scheme. Such documentation permits an external audit it allows reproducibility of findings and replicability of studies; it can support dialogue among researchers struggling with qualitative analysis that can lead to something resembling shared methodological canons. However methodological canons also result in endless refinement of effort intense socialization of novitiates into a received orthodoxy and preoccupation with methods rather than the substance of inquiry. Nevertheless self-documentation is labor-intensive and not a total substitute for verbal elaboration. Filling out forms usually takes at least 15% of the total analysis time. Thus although the present documentation method is actively recommended it is hoped that it will help supply some of the common language and syntax that qualitative researchers will need in elaborating a well documented and credible methodology.