• secondary analysis is neither a specific regimen of analytic procedures nor a statistical technique, but a set of research endeavors that use existing materials

• primary analysis involves both data collection and analysis, while secondary analysis requires the application of creative analytical techniques to data that have been amassed by others

• Meta-analysis integrates the findings from a universe (or sample) of investigations of some phenomenon

  • the study itself becomes the unit of analysis

  • Meta-analysis involves using published research results to compute an overall level of significance for an array of comparable test statistics

• Documents (written materials that contain information) are used extensively in historical research and often provide data over time where no similar data exist

  • While documents themselves may be classified as primary (i.e., eyewitness descriptions of behavior or events) or secondary (i.e., second-hand accounts), document studies are virtually always secondary analyses

  • But they’re used less frequently than surveys for secondary data analysis

Advantages of secondary survey analysis

• The primary advantage of secondary survey analysis is its potential for resource savings

  • it requires less money, less time, and fewer personnel and is attractive in times of economic fluctuations when funds are limited or uncertain.

  • With data already collected, the costs are only those of obtaining the data, preparing them for analysis, and conducting the analysis

  • a researcher can complete a research project independently, thereby eliminating the need for staff

• secondary analysis circumvents data collection problems

Limitations of secondary analysis

• many limitations come from the limitations of survey collection

• A major problem is data availability

  • Despite the development of data archives, researchers sometimes have trouble locating what they need

    • Some topics lend themselves more readily to secondary analysis than others

• the time involved in acquiring data sets from archive

  • but it’s usually still shorter than creating and conducting a survey

• data quality

  • Data files from surveys employing nationally representative samples, properly designed questionnaires, and rigorous procedures for interviewing and coding do not always exist

  • Even surveys of high quality may have measurement problems (invalidity)

• possible inhibition of creativity

  • If researchers use the same data sets repeatedly and are limited by the variables contained therein, scientific progress will be thwarted to some extent

  • continued use of the same indices and data sets may limit the scope of social science research