1995
DOI: 10.1006/anbe.1995.0069
|View full text |Cite
|
Sign up to set email alerts
|

Practical considerations in the use of simultaneous inference for multiple tests

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
120
0
1

Year Published

1996
1996
2012
2012

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 213 publications
(122 citation statements)
references
References 0 publications
1
120
0
1
Order By: Relevance
“…Table 1) we presented the raw values of statistical tests; we believe that such approach is useful to comparison with other published investigations on metal interactions in tissues in birds and threshold of toxic metals indicative of acute and chronic exposure or toxicosis, see papers in which the matrices of correlations between concentrations different metals are presented without Bonferroni adjusted P-value; e.g., Wenzel et al (1996), Woshner (2001a, b), Kalisińska et al (2004Kalisińska et al ( , 2007, Mochizuki et al (2008), Larison (2002). The accepted Bonferroni adjusted value agrees with the level of significance recommended for multiple comparisons (Chandler 1995). For remaining tests the significance level was 0.05.…”
Section: Discussionsupporting
confidence: 66%
“…Table 1) we presented the raw values of statistical tests; we believe that such approach is useful to comparison with other published investigations on metal interactions in tissues in birds and threshold of toxic metals indicative of acute and chronic exposure or toxicosis, see papers in which the matrices of correlations between concentrations different metals are presented without Bonferroni adjusted P-value; e.g., Wenzel et al (1996), Woshner (2001a, b), Kalisińska et al (2004Kalisińska et al ( , 2007, Mochizuki et al (2008), Larison (2002). The accepted Bonferroni adjusted value agrees with the level of significance recommended for multiple comparisons (Chandler 1995). For remaining tests the significance level was 0.05.…”
Section: Discussionsupporting
confidence: 66%
“…Family-wise error was corrected for using the Dunn's post hoc test in comparisons between species, but only within each variable. The between-variable family-wise error rate as a result of the many Kruskal-Wallis tests was not corrected for using the standard Bonferroni correction because of its tendency to increase the type II error rate, particularly when sample sizes are small (Chandler, 1995;Garamszegi, 2006). We therefore chose to leave the question of biological significance in the hands of the readers, and limited our discussion to those variables that were highly significantly different between species (P,0.001).…”
Section: Discussionmentioning
confidence: 99%
“…A value of ÿ2 P ln P greater than the corresponding c 2 value allows us to reject the null hypothesis of no effect. We used a Bonferroni-corrected critical significance level to correct for the familywise error rate when a family of related tests was performed (Chandler 1995). Table 1 presents the results of the dominance analyses.…”
Section: Other Analysesmentioning
confidence: 99%