Backward versus Forward Chaining in the Acquisition of a Keyboard Skill

Abstract

Although backward chaining has been found superior to whole-task training, the effect might depend on the ordering of difficult and easy segments of the task, and forward chaining requires investigation. The experiment uses a musical keyboard task to test these possibilities, counterbalancing difficulty order with direction of chaining in a comparison with whole training. Performance was scored for melodic errors and for rhythmic accuracy and consistency. Both of the part-task methods proved superior to the whole method during training trials, in criterion trials on the whole task, and during retention after one week. However, forward chaining conferred a greater advantage than backward chaining on most measures. The results weaken several explanations for the superiority of backward chaining but confirm the advantages of segmentation methods in part-task training.