Students' scientific investigations have been identified in national standards and related reform documents as a critical component of students' learning experiences in school, yet it is not easy to implement them in science classrooms. Could science demonstrations help science teachers put this recommendation into practice? While demonstrations are a common practice in the science classroom and research has documented some positive effects in terms of student motivation and engagement from their use, the literature also shows that, as traditionally presented, science demonstrations do not always achieve their intended outcomes. This, in turn, suggested the value of investigating what design elements of demonstrations could be used to promote specific instructional goals. Employing action research as a methodology, the proposed study was developed to explore how science demonstrations can be designed so as to most effectively promote student engagement in scientific investigations. More specifically, I was interested in examining the effects of using a discrepant event as part of the demonstration, as a way to create cognitive conflict and, thus, increase interest and engagement. I also investigated the relative merit of the well-researched POE (Predict, Observe, Explain) design versus employing demonstrations that appear to the student to be unplanned (what I will refer to as NOE, or a Naturally Occurring Experience). This study was informed by Constructivism, Situated Cognition and Conceptual Change as theoretical frameworks. The project included the design, implementation and study of an intervention consisting of three instructional units designed to support students' learning of the concepts of density, molecular arrangement of gas particles, and cohesion, respectively. In each of these units, lasting a total of two 80-minute class periods, students were asked to design and conduct an investigation to gain a better understanding of the concept under study. In one case, though, the investigation was preceded by a discrepant event demonstration using POE, in another case the investigation was preceded by an NOE discrepant event demonstration, and in the third case the student investigation was preceded by an interactive lecture (Lecture/Inquiry, or LA) instead of a demonstration. The intervention took place in Fall 2009 in three sections of the same middle school science course I taught. Data from these experiences were collected and analyzed to evaluate the impact of each unit on (a) students' interest in learning more about the scientific phenomenon under study; and (b) how students designed, conducted and interpreted their own investigation to explain the event. These findings were further compared across experiences to identify similarities and differences connected with the three design approaches utilized--i.e., inquiry following a discrepant event demonstration using POE, an NOE discrepant event demonstration, or an interactive lecture. Data sources included: audiotapes of each lesson, students' written work, teacher's written reflections, observer's field notes, audiotapes of a final class reflection and semi-structured student interviews. Qualitative analysis was employed to analyze the data with the goal of revealing emerging themes addressing each research question. Findings from this study show that discrepant event demonstrations can indeed generate student interest and inform worthwhile student-led science investigations without requiring great time commitment. Furthermore, each lesson design used (POE, NOE, L/I) offered distinct benefits in the classroom, influencing student engagement and learning outcomes in valuable and distinct ways. This, in turn, suggests that science teachers should choose specific design elements when planning to use demonstrations to achieve specific objectives. [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://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]