We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Why Journals Should Print Notes and Comments More Conspicuously: the Case of Kendall's W.
- Authors
Arabie, Phipps
- Abstract
The problem of using Kendall's coefficient of Concordance W for sociometric data, excluding self-rankings, has surfaced repeatedly in the literature. Professors Gordon H. Lewis and Richard G. Johnson showed how to compute W in this case and reported Monte Carlo studies with eleven selected significance levels each for N lying between 4 to 10. Probably as a consequence of the regrettably inconspicuous format used by the American Sociological Review, Lewis and Johnson had apparently overlooked a comment by researcher Richard H. Willis on Jones, noting that researcher Kräupfl F. Taylor had already derived the appropriate formulae with a correction for ties, and researcher Ben Willerman had also provided the formulae and a table of critical values for N lying between 4 to 15.
- Subjects
STATISTICAL correlation; MONTE Carlo method; NUMERICAL analysis; JOHNSON, Richard G.; WILLIS, Richard H.; MATHEMATICAL statistics
- Publication
Quality & Quantity, 1982, Vol 16, Issue 2, p163
- ISSN
0033-5177
- Publication type
Article
- DOI
10.1007/BF00166882