The study examined whether an instruction-based strategy (studying the instructions before attempting the task) or a task-based strategy (attempting the task and referencing instructions) was more effective for procedural performance and learning. Four groups of participants learned to perform macram tasks and assembly tasks, and received detailed instructions at different times in the process of attempting the tasks. Performance was assessed at training and a week later by recording task completion time, correctness, and subjective cognitive load. The strategy for using instructions affected initial performance on the macram tasks, where instruction-based strategy was superior, but not later retention or transfer. This pattern of results was not found for the assembly tasks indicating that characteristics of the tasks influenced the effectiveness of the strategy for using instructions.

Including domain rules and generalities (principles) in instructions for procedural tasks is believed to help learners understand the task domain (or the system), and in turn make them better able to complete tasks. However, equivocal results of prior research indicate that principles are not always beneficial. The goal of the current research was to delineate the characteristics of the conditions under which principles are useful. In two studies I investigated the impact of the timing of principle use, the method used to study the principles, and the specificity of the procedural instructions accompanying the principles. The first study showed that the timing of principle use (studying the principles before, during, or after completing training tasks) did not affect declarative (knowledge of the system) or procedural learning (troubleshooting task performance). Therefore, the commonly advocated idea that principles should be provided before task engagement was not supported. Neither was the hypothesis that using principles while solving tasks would enhance procedural learning. When learners summarized the principles, they demonstrated better declarative learning compared to when they just read the principles. Better declarative learning was associated with better procedural learning, but the relationship between understanding and using a system is likely not as direct as often assumed. In the second study declarative and procedural learning were enhanced when the principles were accompanied by general rather than detailed procedural instructions. General procedural instructions appeared to encourage task engagement and the effective use of principles although this effect was reduced if leaners were required to summarize the principles rather than simply read them. Together the results of the two studies reveal how the learning situation and instructional materials can be constructed to create conditions where principles enhance learning and subsequent performance.