New Empirical Directions to Evaluate Online Learning

Abstract

A semantic net on Piaget s cognitive theory was obtained from a sample of 75 psychology pregraduate students. This allowed testing for semantic priming effects in this knowledge domain in other 20 pregraduate students. This last sample was required to take semantic priming studies before and after a computer lesson on Piaget s theory (experimental group), whereas another group who did not take the computer lesson was tested only once (control group). The semantic priming study allowed comparison of the word pair latencies based on Piaget s theory against standard associative, categorical and neutral related word pairs latencies. The schemata related word pairs used in the semantic priming studies were selected from connectionist computer simulations of schemata behavior, which in turn was based on the obtained semantic net. Then another connectionist neural net was implemented to test if semantic priming latencies from the control and experimental group provide systematic information to a neural net to discriminate students who learned the computer lesson from those who did not take the lesson. Results showed that the neural net is capable of discriminating students who integrated new information in long-term memory from those who did not by only considering reaction times from the semantic priming studies. Implications for new ways to evaluate e-learning are discussed.