1973
DOI: 10.1080/01621459.1973.10481335
|View full text |Cite
|
Sign up to set email alerts
|

An Evaluation of Ten Pairwise Multiple Comparison Procedures by Monte Carlo Methods

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
66
0

Year Published

1976
1976
2015
2015

Publication Types

Select...
9
1

Relationship

0
10

Authors

Journals

citations
Cited by 533 publications
(66 citation statements)
references
References 11 publications
0
66
0
Order By: Relevance
“…Pre-drinking sensitivity to effects of ethanol on play fighting, social investigation and social preference were analyzed in low and high drinkers using separate 2 (sex) X 2 (drinking level) X 3 (ethanol dose: 0, 0.5, and 0.75 g/kg) ANOVAs, with ethanol dose treated as a repeated measure. In order to avoid inflating the possibility of type II errors on tests with multiple factors (Carmer & Swanson, 1973), Fisher's planned pairwise comparisons were used to explore significant effects and interactions.…”
Section: Methodsmentioning
confidence: 99%
“…Pre-drinking sensitivity to effects of ethanol on play fighting, social investigation and social preference were analyzed in low and high drinkers using separate 2 (sex) X 2 (drinking level) X 3 (ethanol dose: 0, 0.5, and 0.75 g/kg) ANOVAs, with ethanol dose treated as a repeated measure. In order to avoid inflating the possibility of type II errors on tests with multiple factors (Carmer & Swanson, 1973), Fisher's planned pairwise comparisons were used to explore significant effects and interactions.…”
Section: Methodsmentioning
confidence: 99%
“…Total VFA concentration (^moles/ml) and molar percentages of each of the three acids were calculated. In all experiments, multiple comparisons between treatments were statistically tested by an F-test at an a = .05 level (Carmer and Swanson, 1973).…”
Section: Methodsmentioning
confidence: 99%
“…This approach was adopted with the present limited sample size of 34 classes because it seemed to combine the good power characteristics of individual t tests with the protection against large, experimentwise Type I error afforded by the requirement that the overall F also meet the 0.05 significance criterion (Carmer & Swanson, 1973;Cohen & Cohen, 1975, p. 162). In future studies involving larger samples, however, better protection against Type I errors could be obtained by using Tukey's or Newman-Keul's tests instead of t tests.…”
Section: Differences Between Four Forms Of Iceqmentioning
confidence: 99%