Mixing process-generated data in market sociology

Mixed methods researchers often neglect the methodological discussion on data in general and process-generated data specifically. Assuming that the advantages and disadvantages of surveys, interviews and observations are well-known, the author discusses when and why researchers should prefer process-generated data to research-elicited data in market sociology. The author then compares four different types of process-generated data (aggregated statistical data, business directories, newspaper articles, engineering literature). For each data type, the author (a) describes the data type, (b) illustrates what kind of information researchers can gain from this data type and (c) discusses the draw-backs and limits of the data. The article concludes with a discussion of what one can learn for mixing methods and market sociology from this example and why primary data might be useful as an alternative data source after all.