Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Lost to Industrial Secrecy, Statistical Insignificance and Short Attention Span: Dark, Dead, and Dated Data

item Lord, Jeffrey

Submitted to: Society for Invertebrate Pathology Annual Meeting
Publication Type: Other
Publication Acceptance Date: June 6, 2006
Publication Date: September 1, 2006
Citation: Lord, J.C. 2006. Lost to industrial secrecy, statistical insignificance and short attention span: dark, dead, and dated data [abstract]. Society for Invertebrate Pathology Annual Meeting, Wuhan, P.R. China, August 27-September 1, 2006.

Technical Abstract: There are many reasons for data failing to reach publication. Perhaps the most common is that the data are not statistically significant. Yet there can be great value in negative data. For example, there are many published reports of pathogen efficacy where unrealistic doses have been used. Negative results with realistic doses remain in notebooks while questionable positive results are cited. Including negative data in publications along with positive results should be encouraged. Industrial data is published at the pleasure of men who wear neckties. This means that unfavorable results are rarely reported. Because of the commercial and regulatory rather than academic purposes of industrial research, the data are sometimes not taken in a manner that is suitable for publication. Confidentiality agreements can prevent publication of data from collaborative research for up to ten years. Examples of data failing to reach publication for these reasons will be discussed.

Last Modified: 5/5/2015
Footer Content Back to Top of Page