Abstract
Surveys are an important methodological tool in public management research. Multiple informant surveys are held to have considerable methodological advantages over elite surveys (the practice of surveying a top manager, e.g., a chief executive). Although in principle multiple informant surveys can provide a more accurate organizational picture, problems of data aggregation arise in practice. To promote better use of multiple informant surveys, this article reviews approaches to aggregating organizational data. It provides the first empirical test of echelon methods of data aggregation for public management research. We find significant differences between echelon aggregations, elite surveys and unstandardized forms of aggregations (e.g., a simple mean). These results support our argument that careful theoretical and empirical analysis of multiple informant surveys data is required to provide valid and reliable measures of organizational properties.
Original language | English |
---|---|
Pages (from-to) | 229-253 |
Number of pages | 25 |
Journal | Journal of Public Administration Research and Theory |
Volume | 19 |
Issue number | 2 |
Early online date | 12 Sept 2008 |
DOIs | |
Publication status | Published - Apr 2009 |
Externally published | Yes |
Funding
This article was commenced under ESRC/EPSRC (331-25-0006) and continued under ESRC (062-23-0039). Data are from a project funded by the Office of the Deputy Prime Minister.