This engineering education research project seeks to develop a set of modular, open-source applications, or "apps" to teach engineering design. The software will be designed to allow simple introduction of design problems into early-level engineering courses. The project's evaluation plan is designed to measure the effectiveness of the software for improving student learning outcomes on understanding course concepts.
The broader significance and importance of this project is to investigate a method which can introduce design into early courses in an engineering curriculum in an affordable manner. The project seeks to promote broad dissemination of the software through an "app store" modeled off of a popular smart-phone company. This approach is designed to mesh with the experience of the target audience of students.
Exploration is an essential element of engineering design problems. However, exploring a problem’s design space is often burdened by the complexity of engineering analysis. The ultimate goal of this research was to overcome this barrier and improve the design experience of students in early engineering education. This goal was met by exploring effective initialization and feedback strategies in computer-based educational software that exposed undergraduate engineering students to open-ended design problems. Towards this goal, design Modules were constructed, tested, and improved to gauge the impact of various factors on the student’s design experience. During the period of this award, over 300 undergraduate engineering students in the early portion of their curriculum participated in three experimental studies. Engineering design Modules were created for core courses in the mechanical and aerospace engineering curricula, and will provide the basis for an engineering design thread at NC State. The Modules created as part of this project are designed to be freeware and used at the freshman/sophomore level to explore design variable interactions at the discipline-level. A major effort during this award period was to understand the most effective ways to redesign the Modules and gauge their effectiveness. Ten criterion developed during the re-design of a water rocket propulsion Module were used to assess the effectiveness of the changes. Assessment was completed using both qualitative and quantitative datasets. Updates to the Module led to improvements in eight of the ten criteria. Several factors were then investigated for their influence on these improvements. The first significant factor was the inclusion of a viable starting point. Initial experiments did not give a starting design, leading to a large number of infeasible evaluations early in the exploration. In the updated version, students choose a starting design from a Latin hypercube sample. This allowed students to generate useable results in a timely manner and increased ease-of-use. Including localized sensitivity information was a second significant factor. In the updated Module, students were shown the mapping between the change in a design variable and the resultant change in a performance parameter. This sensitivity information aided the students by providing insight into the magnitude of the effect associated with a change in solution. This allowed the students to make the most effective changes first, and lowered the amount of time the students needed to effectively move throughout the solution space. Sensitivity information also allowed the students to better understand the relationships between the design variables and performance objectives, resulting in a larger proportion of evaluated designs that improved upon the most recent design. This ability to keep improving the design meant that the sensitivity information increased the amount of engagement the students had with the module in terms of time spent and evaluations logged. The final factor considered was time limit. In the first two experiments, students were given 10 minutes. In a third experiment the time limit was raised to 20 minutes. Students remained highly engaged as both their design time and number of evaluations increased. However, these students took longer between design evaluations, suggesting they spent more time thinking about their next move. This slower pace made the module appear easier to use as a large percentage of the evaluations improved upon the starting design. Additionally, the higher percentage of designs that improved on the most recent design indicates an increased understanding. The outcomes from our studies suggest that the Module was improved from a usability standpoint, as the feature additions made the design exploration process better. However, there is still work to be done from an educational perspective. A series of quizzes were administered before and after Module usage to measure knowledge gain. The students demonstrated that in the short term they were able to learn effectively using a handout. In all three experiments the students scored the highest on the quiz taken immediately after reading the handout. In the final experiment – where all improvements were included - usage of the module increased quiz scores compared to students that did not use the module but took the same quiz. Lastly, it was found that students who read a handout explaining fundamental theory prior to using the Module scored better on their quiz than students that were just told to use the Module without any prior instruction. This supports the argument that these Modules are meant to supplement traditional lectures, not replace them entirely. Future work should explore the educational value of this tool when used repeatedly in a course instead of a single-use exercise. Next, additional Modules that have been created need to be interfaced to allow more complex problems to be explored. This would introduce students to multidisciplinary tradeoffs as they try to barter with their teammates for more design flexibility in order to reach performance goals for their specific module.
