Whereas traces may be considered indications for behavior that are "simply there," archives and other types of records are arranged intentionally. This may be related to research questions, for example, in observation studies (Shaughnessy & Zechmeister, 1990). Here we are primarily interested in archives that exist independent of specific scientific investigations. Data files in administration as well as sales records may be subsumed under this category. Moreover, as archival data, material written by persons who are the target of research may also be investigated. Not only are descriptions of accidents and complaints about nuisances interesting from a psychological aspect, but so are private materials like letters and other forms of written communications (Laucken, Mees, & Chassein, 1988). Basically, when the generation of the material is not affected by the research process, the method is considered highly nonreactive and likely represents the techniques of Type 5.
We must bear in mind, however, that archival data may be generated under circumstances that can be considered highly reactive. Actors may very well know that important data are being documented, and they may be motivated to influence this data. With respect to the considerations on nonreactive methods, we can offer two solutions concerning this problem. First, the process and the result of a participant's intentionally influencing archival data may be the target of research. This is possible when different perspectives are available such as in records of conflict in which, for example, mutual accusations directed at the involved actors or third parties are of interest. Second, if biases occurring during the creation of archival data are identified by the investigation, we may cope with the effect particularly when interpreting archival data.
Compared with physical traces, archival data may be more suitable and effective for research purposes, but they are more susceptible to undesirable effects at the same time. Traces may be less simple to read and to interpret, but they are usually less susceptible to those effects we intend to minimize.
Was this article helpful?