An archive is any collection of records. Traditional archives include library records, courthouse records, and business records. Most archives are valuable so pains are taken to keep them safe. Natural disasters and wars can destroy archives and cause the loss of priceless information. Archival data are collected for a variety of reasons. Because these
archives are usually printed or handwritten, they are called paper archives. With the advent of computers and the Internet, many archives now exist only in electronic form; they are called electronic archives. In the next chapter, we explain how to use a well-known electronic archive, PsycINFO, which lists more than two million records from the 1800s till today.
Examples of Archival Research
Archives often span many years, which enables researchers to study behaviors that develop over decades. Mullen’s (2001) archival research used several paper archives spanning 150 years to investigate ethnophaulisms, which are words that are used as slurs to describe immigrant groups. The archives he examined included 40 years of data from United States immigration quotas records, 30 years of naturalization records, and 150 years of data on words used to describe 19 European ethnic immigrant groups (Allen, 1983). All archives indicated that “smaller, less familiar, and more foreign immigrant groups tend to be cognitively represented in a simplistic and negative manner” (Allen, 1983, p. 472).
Using electronic archives obtained from litigation, Lê Cook et al. (2003) examined more than five million on-line documents related to internal research conducted by tobacco companies over a 35-year period. From that large body of documents, the researchers found 239 that dealt with research on smokers’ psychological needs and market segmentation. The 239 documents were selected by a computerized, index-based word search for terms related to psychological needs, segmentation, and personality. At least two researchers performed independent searches until the results were consistent. The 239 documents were catalogued and placed on-line as another, smaller archive.
Archival data are nearly always plagued by inconsistencies and incompleteness. Cook et al. noted that the research methods used for internal, industrial research are sometimes not explicitly stated, and peer review is not as stringent as in research published in journals. Also, the different tobacco companies conducted research under conditions that were not always consistent so the results may not be directly comparable. Finally, unreleased documents or documents missed in the computer search may exist. Nevertheless, the researchers concluded that the tobacco companies did use sophisticated market research to segment smokers and then market cigarettes and other tobacco products accordingly. Such research is not illegal or unethical in general. However, the researchers questioned whether such methods are ethical in industries in which the products cause addiction and preventable deaths.
Conducting Archival Research
Participants In archival research, there are no live participants as there are in other types of research. Instead, the archives consist of records already collected from “participants” in the past. In the Mullen (2001) study, some of the records were 150 years old.
Apparatus Paper archives come in many forms. They might be the correspondence between two scientists in a box in a library. Such an example is the correspondence between Ivan Pavlov and Robert M. Yerkes at the Harvard Library. Paper archives may also be collections of public or private records such as those found in libraries, courthouses, or businesses. Finding and gaining access to the archives is the first step in archival research with paper archives. Once access is obtained, the archives must be searched, the necessary materials identified, and data collected.
Electronic archives, too, may be public or restricted. Public archives are available to all. Restricted archives may require subscription or payment. As illustrated in the tobacco company archival research, successful use of electronic archives requires knowledge and expertise of computerized search engines.
Procedure Regardless of the type of archive, the procedures for data collection from the archive are similar. Records on the topic must be identified, categorized, and converted into data that are then analyzed with quantitative or qualitative methods.
Results Conclusions from archival research are subject to the limitations that surround the original records. Nearly always, the original records were collected for reasons other those of the archival researcher. For example, many tobacco companies originally conducted research to help them sell cigarettes to particular market segments (young people, women, and particular ethnic groups) but publicly denied doing so. Lê Cook et al. (2003) used the archived data to demonstrate that the tobacco companies had lied about their practices. Another methodological limitation is that we can only study data already archived and cannot collect new data from the “participants” in the archive.
Ethics Because archival data have already been collected, many of the ethical issues of informed consent do not apply. However, the confidentiality of participants must still be protected. Depending on the archive, personal information about the people who provided the data may or may not be available. Researchers and their IRBs should determine how to handle privacy issues before accessing archives. In electronic archives, this problem can be solved by careful selection of archival fields. To avoid privacy concerns, researchers might choose not to access fields such as names, addresses, or telephone numbers. Some archives may be covered by copyright law; researchers may have to obtain permission of the copyright holder to gain access to the archives and publish the results. Naturally, such permission should be secured before accessing the archives.
Tobacco companies Phillip Morris, RJ Reynolds, Brown & Williamson Tobacco Company, and Lorillard were sued by state attorneys general that resulted in the 1998 Master Settlement. The database for the Lê Cook et al. study can be accessed at: http://tobaccodocuments.org.
In the cigarette industry, market segmentation is accomplished by tailoring brands and advertising to specific groups of smokers. Brand marketing and advertising target different groups of smokers. Thus, one brand may appeal to young rural males and another brand may appeal to young urban women.
The archive is available at: http://tobaccodocuments.org/product_design/?field_id=9&field_value=Behavior +Targeting.
Back to Chapter 2 Outline