Three experiments investigated the effect of various kinds of reflection (within-task and post-task reflection) on learning a process control task. Also, two ways of providing learners with alternative ideas about task behavior, exposure to other's ideas and providing hints for task solution, were examined. The task involved a simulated "sugar production factory" in which the learners sought to control sugar production by manipulating workforce size. It was predicted that combining within-task reflection with exposure to alternative task ideas would lead to superior task performance through integration of experiential and reflective knowledge. Contrary to the prediction, within-task reflection consistently interfered with learning and knowledge integration by causing learners to acquire overly general and invalid rules (reflective knowledge) about the relations among task variables. These results were interpreted as evidence of a fundamental tendency (reflective abstraction error) of people to seek simple relations among variables in complex systems when engaged in within-task reflection. The most efficient learning occurred when reflection was discouraged during task performance, learners were given access to alternative ideas about task behavior, and they were given an opportunity to discuss their task experiences with other learners post-task. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]