NotesFAQContact Us
Search Tips
ERIC Number: ED449210
Record Type: Non-Journal
Publication Date: 2001-Feb-1
Pages: 18
Abstractor: N/A
Reference Count: N/A
An Introduction to and Generalization of the "Fail-Safe N."
Long, James
The fail-safe N is typically a "what if" analysis applied to studies rather than to a single study. This statistical procedure provides information regarding the stability of a meta-analysis by demonstrating how many nil-null articles would be needed to change the statistically significant results to a statistically nonsignificant finding. The relevant issues such as statistical significance testing, the "file drawer problem," and how they relate to the fail-safe N are discussed. The paper also explains the fail-safe N procedure in some detail using concrete heuristic examples, and a generalization of the method for use in a single study. (Contains 2 tables and 17 references.) (Author/SLD)
Publication Type: Reports - Descriptive; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A