Secondary Data

In Sociology, primary data is derived from the work the sociologist does directly himself or herself. The sociologist chooses, plans, executes and evaluates the research.
Secondary data is work that already exists from any other source. This work must be evaluated before use.
Note that these definitions of primary and secondary are not the same as those used in the subject of History. History has its own definitions of primary and secondary. In history, a primary source is that which was created at the time for the purpose involved. In history, a secondary source is that written second hand from a primary source and becomes another source. A history textbook is a secondary source regarding its history data and is a primary source about the teaching of history at the time the book was published. Those who study history and sociology need to be clear about these differences.
Examples of secondary sources are:

  • Official statistics (published by government and government connected bodies)
  • Unofficial statistics (from private organisations like businesses or charities)
  • The mass media and its journalistic stories (its presentation of its work from its sources: its sources often stay hidden)
  • Life documents (such as diaries, journals, letters, emails, websites)
  • Other research work by others in many academic fields (including source data and analysis/ presentation of the data)

Evaluating Secondary material:
the guidelines...

From the point of view of a sociologist using secondary sources, exclusively or as support for primary research, critical evaluative questions must be asked. These include:
  • How useful are the specific secondary sources?
  • How rigorous was the work?
  • Why was that research done?
  • How was it collected?
  • Who presented the data and the context: might it be in their interest to do so?
  • Has a previous sociologist used it as a secondary source and how can that be evaluated?
  • What are the dangers of using these secondary sources (beyond a literature review)
In one sense all sociologists must use secondary sources before doing primary work. They need to know the sociological tradition so far: who has done what work and what is the research that needs to be done to advance the tradition.

Evaluating Secondary Material:

Secondary sources are also justified on practical grounds of speed and cost. If the work is already done, and can be evaluated as being reliable or valid, there is perhaps no need to repeat the research (unless there seems to be a need to test the previous work's reliability).
Secondary sources are justified on historical grounds - these are research studies that illustrate the past (again, beware of getting caught up in historians' definitions of primary and secondary). This is sheer practicality: no one can go back in time.
Secondary sources may provide an additional theoretical stance. This facilitates methodological pluralism, broadening out the data, for example providing another method this time of reliability, or this time validity. This means quantitative reliability added to a qualitative approach, or qualitative valid insight on to a quantitative approach. They also facilitate triangulation, providing a critical examination of the primary research.

Evaluating Secondary Material:

Secondary sources involve lack of contact with developing the research, and this means the theoretical difficulty in assessing either validity or reliability.
Ethically, secondary work, when it was done, needed to be funded, and this creates issues of bias of intention especially if the funding body is not known. A further ethical and sometimes theoretical issue (say in interpretivist methods) is that secondary work creates difficulty for the new user in assessing the original researcher's motivation and bias
Secondary sourced work often appears for a purpose, and again involves bias but this time of use - government statistics being a prime example. Even if the content is properly researched, and arguably valid and capable of being reliable, the use of the statistics is ethically questionable. When is it a day to bury bad news; when is a claimant count the same as people who are unemployed?
Ways to analyse Secondary Material:

  • Formal Content Analysis
  • Thematic Analysis
  • Textual Analysis
Overall points to consider:

  • A positivist/ falsification view of the work (is it objective, value-free, controlled data collection, representative, reliable?)
  • An interpretivist view of the work (is it open to respondent responses, valid, that something purporting to be value free is not?)
  • As ever - consider the Practical, Theoretical, and Ethical


Adrian Worsfold

Pluralist - Liberal and Thoughtful


Last updated on April 09, 2006