Few researchers have attempted to model group support system meeting behavi
or mathematically. Using Abductive Information Modeling (AIM), we show that
group size and idea generation type are primary predictors of group proces
s satisfaction. While similar to artificial neural networks, abduction freq
uently provides simpler models and yields weights for links among the model
variables. Results show that the interrelationships among the model variab
les are non-linear. (C) 2000 Elsevier Science B.V. All rights reserved.