Background

Agency records, secondary data, and content analysis do not require direct interaction with research subjects. Government agencies gather a vast amount of crime and criminal justice data and this information is referred to as data from agency records. In content analysis, researchers examine a class of social artifacts – typically written documents. Information collected by others is frequently used in criminal justice research, which often involves secondary analysis of existing data. In a general sense, most data obtained from agency records or research projects conducted by others are secondary data.

Agency records support a wide variety of research applications. Published statistics and agency records are most commonly used in descriptive or exploratory studies, but agency records are frequently used in applied studies as well. Researchers use a variety of published statistics and nonpublic agency records. Information collected by or for public agencies usually falls into one of three general categories:

  • published statistics
  • nonpublic agency records routinely collected for internal use
  • new data collected by agency staff for specific research purposes

Researchers must be especially attentive to the units of agency records, especially in selecting samples for analysis. Archives and agency records may be based on units of analysis that are not suitable for particular research questions. It is often necessary to select subsets of agency records (samples) for particular research purposes

Understanding the details of how agency records are produced is the best guard against reliability and validity problems. The key to evaluating the reliability and validity of agency records, as well as the general suitability of those data for a research project, is to understand as completely as possible how the data were originally collected.

  • The text discusses a number of general characteristics of the recordkeeping process:
  • Virtually all criminal justice recordkeeping is a social process.
  • Agency data are not designed for research.
  • At the operational level, most officials in criminal justice organizations are more interested in keeping track of individual cases than in examining patterns.
  • Error increases with volume.

Content analysis involves the systematic study of messages and is particularly well suited to the study of communications and to answering the classic question of communication research: Who says what, to whom, why, how, and with what effect? In the study of communications, as in the study of people, it is usually impossible for researchers to directly observe all that interests them. Usually, it is appropriate to sample the units on which content analysis will be conducted. Content analysis is essentially a coding operation and, of course, coding represents the measurement process in content analysis. Coding the manifest content – the visible, surface content – of a communication more closely approximates the use of closed-ended items in a survey questionnaire. Alternatively, the latent content of the communication may be coded – its underlying meaning. No coding scheme should be used in content analysis until it has been carefully pretested. The text provides two illustrations of content analysis: the measurement and characterization of newspaper crime stories and the content analysis of body language.

Data collected by other researchers are often used to address new research questions and this technique has become an increasingly important tool. The advantages of secondary analysis are obvious and enormous: It is cheaper and faster than collecting original data, and, depending on who did the original study, there may be substantial benefit from the work of topflight professionals and esteemed academics. There are disadvantages, however, and the key problem involves the recurrent question of validity.

1