The effects of implicit, explicit, and synergistic training on learning an artificial grammar
Thomas J. Domangue · 2002
Participants were trained to generate exemplars of an artificial grammar by bubbling-in letters from exemplars (implicit training), observing a diagram of the grammar then reproducing it (explicit training), or tracing the path of exemplars through a diagram of the grammar (synergistic training). Performance was measured using a cued-generate task. It provided a template for an exemplar with two letters filled in. Participants attempted to generate exemplars that fit the template. The computer corrected the exemplar when it matched at least 70% of the letters in a valid string. Results showed that both explicit and synergistic training led to generation of better quality exemplars (closer to 100% match). However, implicit and synergistic training led to generating more exemplars good enough (at least 70% match) to fit into a wide variety of contextual cues. The author concluded that for both quality and generativity of exemplars synergistic training seemed the most beneficial.