2007
DOI: 10.22237/jmasm/1177992660
|View full text |Cite
|
Sign up to set email alerts
|

Tests for Treatment Group Equality When Data are Nonnormal and Heteroscedastic

Abstract: Several tests for group mean equality have been suggested for analyzing nonnormal and heteroscedastic data. A Monte Carlo study compared the Welch tests on ranked data and heterogeneous, nonparametric statistics with previously recommended procedures. Type I error rates for the Welch tests on ranks and the heterogeneous, nonparametric statistics were well controlled with a slight power advantage for the Welch tests on ranks.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

0
24
0

Year Published

2013
2013
2024
2024

Publication Types

Select...
10

Relationship

1
9

Authors

Journals

citations
Cited by 30 publications
(24 citation statements)
references
References 35 publications
0
24
0
Order By: Relevance
“…The age data were nonnormal and heteroscedastic; therefore, the Welch test was applied to the ranked age values (as was recommended by Cribbie, Wilcox, Bewell, & Keselman, 2007). An omnibus test demonstrated that age differed significantly between the groups, F(2, 382.55) = 350.91, p < .001.…”
mentioning
confidence: 99%
“…The age data were nonnormal and heteroscedastic; therefore, the Welch test was applied to the ranked age values (as was recommended by Cribbie, Wilcox, Bewell, & Keselman, 2007). An omnibus test demonstrated that age differed significantly between the groups, F(2, 382.55) = 350.91, p < .001.…”
mentioning
confidence: 99%
“…Because some species were more common than others, sample sizes were unequal and criteria for normality and heteroscedascity were not met. We therefore utilised the Welch test on ranked data to test for overall significance and the Ryan-Elinot-Gabriel-Welsch (REGWQ) test for pairwise comparisons; both of these tests have been shown to control Type I error rates and maximise power when distributions are skewed and variances are unequal (Cribbie et al, 2007).…”
Section: Statistical Analysesmentioning
confidence: 99%
“…In the literature, there are many studies in which the comparison of ANOVA and its alternative tests was made with respect to type I error rate and power when the assumptions of ANOVA are not satisfied. The effects of sample size (balanced/unbalanced case), non-normality, unequal variances, and combined effects of non-normality and unequal variances were investigated in details (Wilcox, 1988;Cribbie et al, 2012;Lantz, 2013;Gamage and Weerahandi, 1998;Parra-Frutos, 2013;Cribbie et al, 2007). In the light of these studies, we conduct a Monte Carlo simulation study in an attempt to give recommendations for applied researchers on the selection of appropriate one-way tests.…”
Section: Introductionmentioning
confidence: 99%