2013
DOI: 10.22237/jmasm/1383278700
|View full text |Cite
|
Sign up to set email alerts
|

The Impact of Continuity Violation on ANOVA and Alternative Methods

Abstract: The normality assumption behind ANOVA and other parametric methods implies that response variables are measured on continuous scales. A simulation approach is used to explore the impact of continuity violation on the performance of statistical methods commonly used by applied researchers to compare locations across several groups.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
3
0
2

Year Published

2015
2015
2022
2022

Publication Types

Select...
3
1

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(5 citation statements)
references
References 16 publications
0
3
0
2
Order By: Relevance
“…Even though the questionnaire yielded ordinal data, parametric analyses were used because our data was normally distributed. ANOVA analyses are relatively insensitive to violations of equidistance if assumptions of normality are met (Lantz, 2013). The ANOVAs showed a significant effect for two of the four variables, namely: anxiety ( F (2,38) = 3.56, p < .05, multivariate eta squared = 0.16) and success ( F (2,38) = 5.02, p < .05, multivariate eta squared = 0.21).…”
Section: Resultsmentioning
confidence: 99%
“…Even though the questionnaire yielded ordinal data, parametric analyses were used because our data was normally distributed. ANOVA analyses are relatively insensitive to violations of equidistance if assumptions of normality are met (Lantz, 2013). The ANOVAs showed a significant effect for two of the four variables, namely: anxiety ( F (2,38) = 3.56, p < .05, multivariate eta squared = 0.16) and success ( F (2,38) = 5.02, p < .05, multivariate eta squared = 0.21).…”
Section: Resultsmentioning
confidence: 99%
“…Thus, we opted to use one-way ANOVAs to compare the means of the survey item responses. ANOVAs have been shown to be the best approach when data are skewed and sample sizes are equal, as they are for the three dosage groups, and when scale points are perceived as equidistant (Lantz, 2013). A nonparametric Kruskal-Wallis test was also run as a robustness check and in case middle-of scale effects were at play (Lantz, 2013); results were the same as for the one-way ANOVAs.…”
Section: Discussionmentioning
confidence: 99%
“…ANOVAs have been shown to be the best approach when data are skewed and sample sizes are equal, as they are for the three dosage groups, and when scale points are perceived as equidistant (Lantz, 2013). A nonparametric Kruskal-Wallis test was also run as a robustness check and in case middle-of scale effects were at play (Lantz, 2013); results were the same as for the one-way ANOVAs. iv Due to the challenges in obtaining responses from staff at a random sample of colleges attended by PUPP alumni, researchers opted to use a convenience sample to obtain sufficient interview data to complete the study.…”
Section: Discussionmentioning
confidence: 99%
“…Apesar de algumas referências indicarem a robustez da ANOVA para pequenos desvios de normalidade (SHARPE, 1970;HARWELL et al 1992;FARAWAY, 2006), estudos mais recentes revelam o impacto causado na violação dessa pressuposição (KHAN;RAYNER, 2003;SCHMIDER et al, 2010;LANTZ, 2013aLANTZ, , 2013b. Isso, aliado à difícil determinação dos limites para pequenos desvios de normalidade, permite inferir que a ANOVA não deveria ser executada quando os resíduos não seguem a distribuição Normal.…”
Section: Discussionunclassified
“…Quando utilizada de forma consciente, a transformação angular se tornou um instrumento eficaz para o analista, numa possível solução do não atendimento às pressuposições. Apesar de ser apontada como solução para o uso seguro da ANOVA (COCHRAN, 1940;RAO, 1960;WINER et al, 1971), ela não pode ser tratada como única e nem ser empregada sem restrição ou análise crítica, pois pode piorar a análise, conforme mencionado anteriormenteisso demonstra que nem sempre a transformação de dados é suficiente para corrigir o não atendimento de uma pressuposição, necessitando o emprego de outras metodologias estatísticas, a exemplo dos MLG (LIX et al, 1996;LANTZ, 2013a) (SAKAMOTO et al, 1989).…”
Section: Discussionunclassified