When less is sometimes more: Optimal learning conditions are required for schema acquisition from multiple examples

While it is usually claimed that multiple examples for the illustration of problem categories are a necessary prerequisite for schema acquisition, there is a lack of conclusive empirical evidence supporting this claim. Moreover, there are findings indicating that carefully designed one-example conditions may allow for profitable processes of example comparison as well. In line with this reasoning, we present an experiment – that builds up on a series of studies conducted by Quilici and Mayer (1996) and a previous experiment by our own research group (Scheiter, Gerjets, & Schuh, 2004) – that demonstrates that multiple examples may only be helpful for learning when sufficient time for their processing is provided and when learners are guided by an instruction to compare examples. On the other hand, learning from single examples proved to be less dependent on optimal learning conditions. Results are discussed in the light of current instructional design theories