Method factors represent variance common to indicators from the same data source. Detecting method factors can help uncover systematic bias in data sources. This article employs confirmatory factor analysis (CFA) to detect method factors in 23 democracy indicators from four popular data sources: The Economist Intelligence Unit (EIU), Freedom House, Polity IV, and the Varieties of Democracy (V-Dem) project. Using three different multi-dimensional concepts of democracy as starting points, we find strong evidence for method factors in all sources. Method-specific factors are strongest when yearly changes in the scores are assessed. The sources find it easier to agree on long-term average scores. We discuss the implications for applied researchers.
Archived Files and Locations
|application/pdf 656.6 kB ||
|application/pdf 501.2 kB ||
|application/pdf 365.9 kB ||
access all versions, variants, and formats of this works (eg, pre-prints)