[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Prolacertiformes and Protorosauria

----- Original Message -----
From: "Mike Taylor" <mike@indexdata.com>
Sent: Monday, May 18, 2009 9:23 AM
Subject: Re: Prolacertiformes and Protorosauria

No, wait, what?  You can't go throwing out characters just because
they're correlated -- only if you can see a functional connection that
indicates that you're effectively coding the same character twice.

And that's what is usually meant by "correlated characters". It's also what I mean, sorry for the confusion.

But it's perfectly possible that two unconnected characters might
happen to be correlated by coincidence, in which case they are both


Indeed. Characters that, in the matrix, show too little correlation to any others most likely lack phylogenetic signal. There's an extremely complicated test for this which has only been implemented once because it's very hard to program... (O'Keefe & Wagner 2001, Syst. Biol.)

"This control [...] would permit the construction of a worm of
unparalleled devastation" -- Security firm @Stake, describing
a security hole in Microsoft's dam' _paperclip_.  I ask you.

<dramatic removal of glasses>