Abstract
In an earlier study into learning by discovery, the variables of school grade and intelligence were investigated. This partial replication using a map-reading task consisted of a 2 X 2 X 2 factorial design involving instructional method (egrule and ruleg), school grade (grades 9 and 5), and intelligence (high and average). A set of eight dependent variables was used. Similar results were obtained from both studies and are discussed with respect to other recent findings.