Field Methods

Complexity theory, developed primarily at the Santa Fe Institute in the 1980s, attracts increasing attention from social researchers. Complexity departs from the tradition of static linear models to focus on nonlinear dynamic systems, difficult to predict, with emergent properties not reducible to their elements. The author is currently working on a project to explain heroin trends. This article explores complexity theory, as reflected in the work of John Holland, to evaluate it as a candidate formal system after which to model this work. After reviewing some of the basic characteristics of the theory, an example is presented to evaluate the fit of theory to case and to outline the kinds of data necessary. At the metaphorical level, the evaluation is a positive one. How well the details of the formalism will fit the details of the case remains an open question.

Although laws and policies developed in the context of off-line research apply to online investigations, questions about the identifiability of human subjects, the conceptualization of privacy, the need for and means of obtaining informed consent, and the applicability of copyright law to computer-mediated communication (CMC) pose special problems for doing research in cyberspace.

The prominence of interpretive, literary, self-reflexive, and hermeneutic approaches in sociocultural anthropology in recent years has led some commentators to argue that the field has taken a humanistic turn. This article examines quantification and statistics in articles about sociocultural anthropology over the past several decades in six U.S. journals, assuming that the proportion of articles using these methods at any given time is a rough indicator of the balance between humanistic and scientific approaches. Although there has been a reduction in recent years in the proportion of articles using quantification and statistics in these journals, the overall drop has not been nearly as marked as many writers suggest. A notable shift away from quantification has occurred in the American Anthropologist, American Ethnologist, and Ethnology. The proportion of articles using quantification and statistics, however, has increased greatly in the Journal of Anthropological Research and remained about the same in Current Anthropology and Human Organization.

Debriefing is not a substitute for ethnographic interviewing but a complement to it. It may be about content, context, or both. It is helpful in situations or descriptions of events in which recording is not possible or permitted. It is one of the best methods for supplying ethnographic context to interviews that the interviewer may have overlooked—it then becomes an interview about an interview. Debriefing may be self-debriefing, one-on-one, one information giver to a group of debriefers, or a group session. Particularly the latter is an excellent teaching tool for ethnographic methods.

SIMSTAT is an inexpensive but full-featured Windows/DOS statistical package. Several features set it apart from its competitors in the same price range. First, SIMSTAT offers extensive bootstrapping capabilities (bootstrapping is a statistical technique that uses resampling to estimate population parameters). Second, WORDSTAT, one of several modules that can be purchased to extend the capabilities of SIMSTAT, provides a wide range of quantitative content analysis features that will be especially useful to researchers who work with relatively small text chunks. SIMSTAT and its modules are relatively easy to use, especially for users who have experience with SPSS, which SIMSTAT resembles.