Supporting Research

The Semantica® products are derived from semantic network theory that supports knowledge construction and visualization. Engaging in knowledge capture with Semantica enhances basic cognitive skills such as automaticity, availability of appropriate problem-solving strategies, quality of knowledge integration, and refinement of knowledge structures. Self-modeling is made possible by the Semantica software, which enables the rapid construction of semantic networks composed of concepts interconnected by labeled relationships.

On the basis of cognitive research, we know that more able individuals in a domain have richer, more interconnected knowledge structures (semantic networks) than do less able individuals (Derry, 1990). Expert's structures are more internally coherent and also more associated with information external to (or distant from) the topic than are novice's (Tweney & Walker, 1990; Roth, 1990). Experts recognize and store more patterns and they organize information into larger chunks (Chase & Simon, 1973; Perkins, 1981). When students use Semantica software as a knowledge construction tool, they tend to move toward more expert-like knowledge structures (Fisher, 2000).

Central to the expert's structures are generative mental models, schemas, or sets of concepts and relationships that are useful for classes of domain problems. Experts can generally make finer discriminations between closely related entities than novices (Klausmeir, 1990; Roth, 1990). The evidence, then, points to a rich, coherent knowledge structure as being a central aspect of competence.

Novices and experts can engage in self-modeling of their own cognitions and can share their perspectives with one another and with their successors. "One branch of cognitive science has been concerned with declarative knowledge structure and representation. This concern with the form and content of conceptual knowledge (i.e., with "knowing that") is most prevalent in semantically rich domains. However, the conceptual representations associated with such declarative structures also have been found to be applicable to such meaningful information structures as procedures, experienced events, visual scenes or patterns, and even complex situations or task structures." (Frederiksen & Breuleux, 1990).

An especially convincing demonstration of the utility of semantic networks in predicting problem solving skill is provided by Gordon and Gill (1989), who were able to predict student performance in two domains with 85% and 93% accuracy on the basis of a conceptual graph (semantic network) representing each student's knowledge about the domain being tested. The utility of a conceptual network is well recognized.

References