3/19/2023 0 Comments Hvad er plingInstead, societal and institutional filters influence which data are produced in what way and if and how data are stored. While they resemble survey data in many ways, they also raise specific methodological problems, as the researcher does not control data production. Social bookkeeping data are one of the oldest data sources for historical and social research. At the same time, they allow triangulating one sample with the other, thus controlling biases and overcoming limitations of a single sam-ple to a certain extent. These sampling strategies provide different solutions to the above stated problems. We finally demonstrate two contrasting ideal-typical strategies for handling this problem. Re-searchers also have to decide how many newspapers, years, issues per newspaper and year and articles per issue are to be sampled. Generally, sampling strategies are multi-staged. We then suggest different strategies for applying these tech-niques for sampling newspaper data. Instead, we suggest using sampling procedures usually applied when sampling qualitative data. Specifically, random sampling usually is no suitable sam-pling technique. Bias and the definition of target population reduce choice of sampling procedures. Thus, humans have to actively preserve data for later use. Data pre-servation is biased, too, as data can be both destroyed and decay. Process-generated data are biased during data production because they are produced for other purposes and under other contexts than scientific research. Depending on how the target population is defined, different sampling prob-lems arise. This data provide us with a rich source of information because newspapers are an actor, arena and archive of public discourses at the same time. For this purpose, we centre on the media discourse on unemployment as an example, here par-ticularly on the news coverage by newspapers. In this paper, we discuss sampling problems and strategies for process-generated data.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |