Product dissection has been widely deployed in engineering education as a means to aid in student's understanding of functional product elements, development of new concept ideas, and their preparation for industry. However, there are large variations in the dissection activities employed in education with little research geared at understanding the impact of these variations on student cognitive load requirements and, ultimately, student conceptual understanding. This is problematic because without this knowledge, we do not know what components of product dissection impact (positively or negatively) conceptual understanding of the dissected product and how this is related to the cognitive requirements of the dissection activity. Therefore, the purpose of this study was to investigate how the type of product dissected (complexity and product power source), the virtuality of the product (physical or virtual), and the type of dissection activity performed impacted student conceptual understanding and cognitive requirements through a factorial experiment with 141 engineering students. While the type of cognitive load varied between virtually and physically dissecting products, no differences were found in subsequent levels of conceptual understanding. This indicates that virtual environments may be used as a proxy for physical environments without impacting the conceptual understanding of products by students. These results are used to develop recommendations for the use of product dissection in education and propel future research that investigates relationships between example-based design practices and student understanding outcomes. [ABSTRACT FROM AUTHOR]