Cue Competition Effects: Empirical Tests of Adaptive Network Learning Models