4 Sources of quantitative research

In thinking about any kind of data, we need to give attention both to how it was produced and to the purposes for which it is to be used, in other words what inferences it is being or will be employed to make. These two aspects need to be considered together, since each has implications for the other.

It is important to recognise the diversity of sources and forms of data used in quantitative research. One source is available statistics or secondary data sources (including what are often called ‘official statistics’). These consist of numerical data that have been produced by various organisations and agencies – publicly funded bodies of various kinds, commercial organisations, interest groups – as well as nation states and international governing institutions like the European Union. Equally important are bodies of quantitative data produced by previous social science research projects. These are often available in data archives and it may be possible to re-analyse them in order to address somewhat different questions from those with which the original studies were concerned.

Instead of, or as well as, relying on already available statistics, researchers often produce quantitative data for the purposes of particular projects: this is a primary data source. This usually involves data being structured at the point of collection in a form that allows counting and/or measurement. For example, information may be collected about the frequencies of various kinds of events through use of an observation schedule. Similarly, psychological tests or fixed-choice questionnaires may be employed to measure abilities or attitudes across a sample of people. However, it is worth pointing out that some of the sorts of data used by qualitative researchers, such as written documents and transcripts of audio or video recordings, can be subjected to quantitative analysis after being collected. This involves developing and applying a structure of categories that allows counting, ranking, or measurement. One example of this, usually applied to written documents, is content analysis. Answers to free response items on questionnaires, and sections on observational schedules requiring open description, also have to be structured in this post hoc way if they are to be used for quantitative analysis.