Strengthening statistical usage in marine ecology

被引:69
作者
Beninger, Peter G. [1 ]
Boldina, Inna [1 ]
Katsanevakis, Stelios [2 ]
机构
[1] Univ Nantes, Fac Sci, Marine Biol Lab, F-44322 Nantes 1, France
[2] Commiss European Communities, Joint Res Ctr, Inst Environm & Sustainabil, I-21020 Ispra, Italy
关键词
Alternate approaches; Data treatment/presentation; NHST; p-Values; Statistics misuse; MODELING FISH GROWTH; INFORMATION-THEORY; P-VALUES; MULTIMODEL INFERENCE; CONFIDENCE-INTERVALS; REVIEWING POLICIES; ALLOMETRIC GROWTH; MEDICAL JOURNALS; PUBLICATION BIAS; STOCK ASSESSMENT;
D O I
10.1016/j.jembe.2012.05.020
中图分类号
Q14 [生态学(生物生态学)];
学科分类号
071012 ; 0713 ;
摘要
Although within their own disciplines, the statistical, social science, medical, and terrestrial ecology literatures are replete with accounts of the widespread misapplication and misuse of statistical testing and interpretation, awareness of these issues is weak among marine scientists who are not statisticians, but whose work is nonetheless situated within the expanse of marine ecology. Moreover, the major recent developments in statistical approaches in these fields are, as yet, poorly-represented in the marine ecological literature. We present a non-technical review of (1) the most fundamental, yet pervasive, problems concerning classical statistics, with suggestions for improved practice, (2) alternate, often more appropriate and intuitive, approaches to statistical design and interpretation, and (3) the crucial roles of reviewers, and especially of editors and editorial boards. It is hoped that increasing the awareness of these issues will strengthen statistical usage in marine ecology. (C) 2012 Elsevier B.V. All rights reserved.
引用
收藏
页码:97 / 108
页数:12
相关论文
共 131 条
[1]  
Akaike H., 1973, 2 INTERNAT SYMPOS IN, P267, DOI [DOI 10.1007/978-1-4612-1694-0_15, 10.1007/978-1-4612-1694-0, 10.1007/978-1-4612-0919-5_38]
[2]  
Altman DG, 1998, STAT MED, V17, P2661, DOI 10.1002/(SICI)1097-0258(19981215)17:23<2661::AID-SIM33>3.0.CO
[3]  
2-B
[4]  
Anderson D.R., 2008, MODEL BASED INFERENC
[5]   Suggestions for presenting the results of data analyses [J].
Anderson, DR ;
Link, WA ;
Johnson, DH ;
Burnham, KP .
JOURNAL OF WILDLIFE MANAGEMENT, 2001, 65 (03) :373-378
[6]  
[Anonymous], ERASMUS J PHILOS EC
[7]  
[Anonymous], 1993, An introduction to the bootstrap
[8]  
[Anonymous], 2002, Model selection and multimodel inference: a practical informationtheoretic approach
[9]  
[Anonymous], SIGNIFICANCE
[10]  
[Anonymous], 2010, TESTING STAT HYPOTHE, DOI DOI 10.1201/EBK1439808184