Stark et al. (1995) showed that multiple teaming contexts can facilita
te learning to control a dynamic system whereas the acquisition of str
uctural knowledge was impaired. In an experiment the generalizability
of this dissociation was tested. The context was manipulated by presen
ting relevant as well as irrelevant system variables to the experiment
al group (n = 29) but only the relevant variables to the control group
(n = 31). By means of the irrelevant variables input-effects which ar
e important for problem-solving can be analysed by multiple examples.
Consistent with expectations this learning context had positive effect
s on control performance but negative effects on the the acquisition o
f structural knowledge. It was suggested that learning contexts requir
ing additional cognitive ressources for problem-solving can have diffe
rential effects on the acquisition of structural knowledge and the acq
uisition of control knowledge.