Description Usage Arguments Details Value References
Function that computes an evidential value in favor of dependent observations versus independence
1 | evi.val(meanVec, sdVec, n.fl)
|
meanVec |
numeric vector containing means for each factor level |
sdVec |
numeric vector containing standard deviations for each factor level |
n.fl |
(numeric) integer indicating cell size/nr. of replicates per factor level |
Function was included without modification from UvA report by Peeters, Klaasen, and van de Wiel (2015).
Assumes only summary/descriptive statistics are available (from publication under investigation)
Assumes ANOVA-type setting with one-way factorial design
Assumes there are 3 factor levels
Assumes the ANOVA is balanced (nr. of replicates same for each factor level)
Assumes following restriction on group means: mu1
2*mu2 + mu3 = 0
Assumes the input vectors are ordered according to linearity of effect, e.g., meanVec[1] < meanVec[2] < meanVec[3]
Assumes dependence between the measurement errors undermines (the assumption of) veracity
Computes the evidential value in favor of the hypothesis of a dependence-structure in the underlying data (i.e., correlated measurement errors, indicating incorrect data collection: H_f) versus the hypothesis of independence (the standard ANOVA assumption: H_i)
The evidential value as calculated is defined in Klaassen (2015).
single-element vector with V statistic, or a list with upper and lower limits for V
Peeters, C. F. W., Klaasen, C. A. J., and van de Wiel, M. A. (2015). Evaluating the Scientific Veracity of Publications by dr. Jens Förster. Retrieved from Retraction Watch website.
Klaassen, C.A.J. (2015). Evidential Value in ANOVA-Regression Results in Scientific Integrity Studies. Arxiv: 1405.4540v2.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.