[ACT-R-users] ranges for B and Sji

[Fabio Del Missier]

-- APOLOGIES FOR MULTIPLE POSTING --

Dear all,

I'm wondering if someone could help us answering a ?simple? question.
We are trying to set base-rate activations and Sjis in an act-r model  
of associative word production.
We have estimates of word occurrences and co-occurrences drawn from  
different sources (linguistic corpora and databases, google searches)  
and we want to use them as proxies for Bs and Sjis. The problem is how  
to scale them.
We went through a series of act-r models and papers and noticed very  
different  ranges for Bs and Sjis in different domains (from cognitive  
arithmetic, to the fan experiments, to stroop --wact).
Could you be so kind to suggest us reasonable Bs/Sjis ranges for a  
model of associative word production?
An independent constraint on the ranges --possibly theoretically  
derived-- would help reducing free parameters.
Many thanks and have a great workshop!

Fabio Del Missier & Cristiano Crescentini.


======================================================
Decision Research Lab
Department of Cognitive Sciences and Education
University of Trento
Via Matteo del Ben 5/B, I-38068, Rovereto (TN), Italy
e-mail: delmissier at form.unitn.it, delmisfa at units.it
skype: fabio.del.missier
fax:   +39 0464 483554
voice: +39 0464 483572
======================================================


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.