Steacie Science and Engineering Library at York University

Some researchers gain access to large amounts of data without interacting with a single research participant. Instead, they use existing records to answer various research questions. This type of research approach is known as archival research. Archival research relies on looking at past records or data sets to look for interesting patterns or relationships. These records may include newspaper articles, personal ads, censuses, sports statistics, speeches by public figures, and even tweets!

In an example of archival research, the measurements of the models photographed for Playboy magazine’s centrefolds were studied from 1953 to 2001 (Voracek & Fisher, 2002). The researchers were able to use the published data on height, weight and measurements for the bust, waist and hips that accompanied the photographs. These data then allowed them to calculate body mass index, waist-to-hip ratio, and a rating on an androgyny index. The results showed that body mass index decreased from 20 to 18, and waist-to-hip ratios increased from just below 0.7 to just above 0.7. The effect of these changes means that more recent models have become more androgynous in appearance.

Because this study used a nonexperimental research method, we cannot make cause and effect claims. We can hypothesize why the centrefolds have become increasingly androgynous but we can’t say so conclusively. Maybe the change reflects a change in the readership of Playboy. Today’s readership is 83% male, earns a median income of $53,000, and is mostly between 18 and 44 years old (75%). How does this compare to the readership in the 1950s? Alternatively, the changes may reflect changes in western cultural tastes and standards regarding female appearance. Only more research using more experimental methods will be able to illuminate these questions.

Advantages of archival research

  • Archival research minimizes the response biases of subjects because the researcher is not present while the data is recorded.
  • Archival data is very plentiful and has already been collected. This makes it easier and often less costly than alternative research methods.
  • Archival research is effective in helping to confirm that the results and theories derived from experiments reflect the “real world” and do not just exist in artificial or simplistic laboratory settings.
  • This approach can help researchers create new ideas for hypotheses and experiments. For example, based on the archival study of the measurements of Playboy models, an experiment could be designed to assess current preferences for female body shapes.

Limitations of archival research

  • A limitation of archival data is referred to as a selective deposit. Archival records may be biased. For example, letters to advice columns are not randomly selected, so they may not accurately reflect all letters that were submitted by readers. Government records that are thought to accurately reflect activity in Parliament can be altered to be more politically correct.
  • Not all archives endure, and those that do not may not have been randomly lost. When the survival of records is selective, there may be bias in the remaining archival data.
  • People make mistakes in entering data in archives.
  • Sometimes definitions change so that even though long‐term records use the same label, what is being included may change over time. For example, the definition of family may change from families with a mother, father, and child to same‐sex couples or single‐parent families.
  • There may be biases when data is recorded. For example, letters to the editor that involve corporate ethics or terrorist activities may not have been published ten years ago but are now considered important enough to publish. The increase in the number of letters on these topics in newspaper archives may reflect changes in editors’ biases and not changes in the actual number of letters sent to the newspaper. Another example of bias is that suicides may be recorded as accidental deaths, to help maintain the privacy of the victims’ families.
  • Perhaps most importantly, we cannot conclude causal relations from archival research. After all, the researcher did not control and manipulate the variables that may have played a role. Because the data were not assembled to answer a particular research question, it might not be exactly the data the researcher requires.
activity

Open Textbook Reading Activity

Read the section on “Correlational Research” in Chapter 7 of your textbook (Nonexperimental Research) and complete the discussion exercise listed at the bottom of the page.