Refine
Year of publication
- 2003 (1)
Document Type
- Working Paper (1)
Language
- English (1)
Has Fulltext
- yes (1)
Is part of the Bibliography
- no (1)
Keywords
- Aussagekraft (1) (show_all)
Institute
Empirical organizational research has serious methodological limitations if it is based on the perceptions of single informants. Single informant studies are likely to suffer from a systematic measurement error, a so-called method error or informant bias which negatively impacts the validity of their empirical findings. Previous empirical innovation research is based on single informants. This causes substantial doubts regarding the validity of findings in past empirical innovation research based on perceptual, self-reported measures and their managerial implications. This study undertakes a comprehensive theoretical and empirical investigation into the causes and the effects of a single informant bias in empirical innovation research. Based on multiple informant data on various aspects of innovation management in a company, multitrait-multimethod analyses are applied in order to quantify the extent of the informant bias. The results show that the informant bias is substantial and that construct validity cannot be achieved based on the assessments of a single informant only. Only multiple informant data allows the selection of an appropriate informant depending on the organizational trait to be measured based on formal statistical criteria. The higher the level of interfunctional integration and the higher the intensity of vertical communication inside the organization, the lower is the informant bias. This information can be used by researchers to assess whether an informant bias is likely to occur prior to conducting empirical surveys. The findings of this study substantiate the presumed doubts about the validity of past single informant studies in innovation research and demand the use of multiple informants in future empirical studies.