Skip to content

How important is to have a balanced design for adonis?

2 messages · Gian Maria Niccolò Benucci, Frederico Faleiro

Original
Gian Maria Niccolò Benucci
# 
Hello dear ecologists,

I have a question for you in here. How important is to have a balanced
design for adonis? In other words, how sensitive is adonis() to differences
in sample numbers within the groups I am comparing?

To link this question to the reality, I am trying to compare the microbial
communities of 3 groups of samples (my treatments) and they have 22, 24,
and 24 samples each. I have always thought that an even design would have
been the rule but I found several comments online that were suggesting the
opposite.

What do you think?

Thank you very much in advance,
Frederico Faleiro
# 
Hi Gian,

the classical PERMANOVA have poor results with unbalanced design (Anderson
and Walsh, 2013) and a solution was proposed by Anderson et al. (2017) to
handle it. However I do not know if it is implemented any software.

Anderson, M.J. and Walsh, D.C.I. (2013) What null hypothesis are you
testing? PERMANOVA, ANOSIM and the Mantel test in the face of heterogeneous
dispersions. Ecol. Monogr., 83, 557?574.
Anderson, M.J., Walsh, D.C.I., Clarke, K.R., et al. (2017) Some solutions
to the multivariate Behrens?Fisher problem for dissimilarity-based
analyses. Aust. N. Z. J. Stat., 59, 57?79.

Cheers,

Fred


On Mon, Oct 29, 2018 at 11:53 AM Gian Maria Niccol? Benucci <
gian.benucci at gmail.com> wrote: