Date of Award

5-2019

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Mechanical Engineering

Committee Member

Joshua D Summers, Committee Chair

Committee Member

Gregory Mocko

Committee Member

Nathan McNeese

Abstract

Requirements are at the heart of the engineering design process as they define the problem and validate the solution. Requirement definition as an activity has not been well investigated, especially with respect to how to teach this to novice engineers. This study explores the influence that an intervention has on novice engineers in generating requirements for a design problem. An experiment was performed in a fourth-year mechanical engineering design course by giving the participants a design problem for which they had to generate a list of requirements. A lecture on requirements was delivered and then the students were given a second problem. The two problems were tested for similarity in design and in outcome. The data was analyzed using modifications to the ideation metrics of quantity, quality, variety, and novelty. Quantity addresses the number of requirements generated and the results indicate a statistically significant positive influence by the intervention (p = 4.76E-14). Completeness evaluates the grammar of the requirement by identifying whether it has a subject, verb, or modal. It is found that the intervention had little influence on the completion. Variety is assessed using eighteen categories to classify each requirement. More categories were addressed after the lecture than before. Novelty is evaluated on the level of uniqueness of the requirement against the complete set generated, based on both syntax and semantic filtering. No novel requirements were found in the “before” treatment but were found in the “after” treatment.

A brief initial experiment comparing novice to practitioners (as defined by engineering employment for at least 3 continuous years) is done. Findings from this second experiment suggest that practitioners and novices without training performed similarly (p = 0.71), and that novices with training performed better than the practitioners (p = 0.0014) with respect to quantity (Count 3). This could serve as a reason for academia to equate the performance of experiments with students without significant training to practitioners.

Share

COinS