successive and implied weighting

[Silvio Shigueo Nihei]

  Dear taxacomers,

My question is concerning to the successive (Farris 1969) and implied
(Goloboff 1993, 1997) weighting approachs used in cladistic analysis.

Under successive weighting scheme, as stated by Farris (1969, Syst.
Zool. 18: 374-385) on page 379, there are 4 main categories of functions
associated to the character weight assignment. With these functions
(e.g. "linear", "convex", "concave and bounded" and "concave and
unbounded") varying according to the concavity value (k).

Although the implied weighting approach works in a different dynamics,
the weight assignment also varies according to the k value used. To
perform implied weighting at PAUP (version 4.0b10), the k defaul value
is 2, and such value can be modified by the user.

My first point is: how can I access the k value of the successive
weighting at the PAUP? and the k value can be modified by the user?

My second point is: under the successive weighting approach, what is the
best index (e.g. ci, rci, ri...) (or the most recommended) to estimate
the character reliability used to assign the weights? The PAUP default
index is rci, although Farris (1969) have originally used the ci. Is
there some work with comparative analysis using different index?

My third point is: is there some work analysing and comparing the
results of cladistic analysis under different k values? The only work I
know is that one of Pape (2001, Syst. Ent. 26: 133-171).


Kind regards,


Silvio Shigueo Nihei

--
Entomology Graduate Studies
Department of Zoology
Universidade Federal do Paran?
CP 19020
Curitiba, Paran?, Brazil
CEP 81531-990

www.silvionihei.hpg.com.br

ssnihei at bio.ufpr.br
silvionihei at yahoo.com