Petermann, Sören: Secondary analysis (April 2009)
Advantages and disadvantages of a secondary analysis
The places in which my colleagues do their field work are surprising to me. Their research blogs reveal stories of mantras in Skopje (Goran), games of patience in Stuttgart (Boris) and the hustle and bustle of life in Hackney, London (Susanne). I am particularly envious of Susanne’s field work, and her tales of observing life from a street café.
How can quantitative social research, being much more distanced from its objects of study, measure up to that? Social scientists working quantitatively perform secondary data analysis with data that research agencies have collected. Such distance, does have its own advantages, though. It saves a lot of time and energy, it is less expensive (saving, at times, taxpayers money) and it generates results that are inter-subjectively comparable and what is commonly known as representative. Any decent textbook for empirical research will highlight the fundamental importance of these advantages, particularly the last one, for the generation of scientific knowledge.
So, while my colleagues undertake their fieldwork in Skopje, Stuttgart or London, I save academic time and money and simply search for a dataset that is relevant to my research question. As I am interested in the social, ethnic and spatial inequalities that shape social relations and networks, finding a relevant dataset should not pose a problem. There exist, after all, thousands of datasets in central archives such as the “GESIS-Zentralarchiv für empirische Sozialforschung“.
Currently, I am working on a research project on residential mobility and social capital. This involves the theoretical modeling of the investments and de-investments of social capital, at both the previous place of residence and at the place of destination in the process of (internal) migration. This endeavor makes a dataset, containing both social networks, i.e. social capital as well as biographies of migration, mandatory.
After having thoroughly searched for available data on the research areas that I am interested in (ethnic diversity, social cohesion and integration, mobility and spatial inequality), I am forced to face the surprising reality of the nonexistence of a relevant dataset about inequalities and networks. This fact is initially very surprising to me, as both of these phenomena belong to the very core of sociological research. After a very tedious searching process, I finally choose to use a dataset which unfortunately does not cover all the information required for testing the hypothesis. This is, of course, the downside of secondary analysis, as the better textbooks on the subject will attest to.
Despite the limited data (which opens the doors for the objections of the critics), the analysis leads to unexpected and surprising results. I find myself in a quandary! My analysis of the data supports the predicted and theoretically modeled process of (de-)investments in social relations. At the same time, however, I cannot invalidate justified criticisms. This is a severe dilemma of secondary analysis (and one that is not discussed in even the best of textbooks). I find myself once again pondering the advantages of field work. That being said, I must insist on the relevance of even such limited data for the advancement of social scientific knowledge. Instead of imagining the results of my secondary analysis being thrown into the recycling bin, I am searching for additional data that could, at least partially, invalidate persistent objections. And I compensate for the “place advantage” of ethnological field work by occasionally working on our terrace, while enjoying a cup of coffee in the sun…
Sören Petermann, April 2009