Science education in academic engineering programs aims to equip students with scientific knowledge and academic skills to solve complex (socio-)technological problems. This article addresses the critical question of whether traditional science courses effectively prepare for this ability. It starts from the premise that scientific approaches to technological design and development require the ability to construct new scientific knowledge (e.g. scientific models) relevant and adequate to the concrete problem. My central claim is that a dominant view of science, called a physics paradigm of science, hinders the effectiveness of science courses. Next, I propose that an alternative view of science, called an engineering paradigm of science, is better suited to understanding scientific approaches in technological design and development, and to developing more effective science education. The engineering paradigm of science assumes 'useful' knowledge as the goal of science, and thus focuses on the construction of scientific knowledge for epistemic purposes in specific contexts such as technological design and development. The philosophy of science can contribute to a better understanding of the epistemic strategies in knowledge construction. I propose to call this domain of study the context of construction.