A pilot study of customer requirement identification methods among engineering students
MetadataShow full item record
After companies gathered raw data from customers, the resulting customer need statements must be interpreted in terms of requirement statements so that designers can use these as inputs to generate technical specifications. There are two well-established methodologies that guide designers on the derivation of requirement statements (Pahl/Beitz's "Systematic Design" and Ulrich/Eppinger's "Product Design and Development") and one emerging concept that applies to identifying customer requirements (Affordance Modeling). To date there has not been a framework that compares the efficacy of these customer requirement derivation (CRD) methods. This work develops and applies a framework that evaluates CRD methods, specifically in terms of the tools' influence on the quality of information at the interface of the customer space and the engineering space. The framework comprises of eight evaluation metrics that were derived from Pahl/Beitz, Ulrich/Eppinger, two Affordance Modeling approaches (Maier/Fadel, Cormier) and related literature. The metrics are Number of Requirements (Nreq), Completion Time (CT), Relevance, Clarity, Specificity, Wording, Representation and Solution-independence. A scoring scheme was created for these metrics and applied to student submissions in a class study. The performance on these metrics determines the goodness of the CRD methods. Results indicate that there are differences in the number of correctly derived requirements for all metrics except Clarity and Specificity. No CRD method is dominant. Consideration of metric sub-categories, correlations among metrics and influences such as complexity of raw data offer more insights. In addition, the work develops an approach for evaluating the acceptance of CRD methods by designers. The Technology Acceptance Model (TAM) was adapted from management / marketing science literature to get insights into what the main drivers of students' attitudes about the methods are. A total of 23 questions were composed in a post-study survey to indirectly measure the constructs of TAM elements. Structural Equation Modeling (SEM) was X conducted in IBM SPSS ® AMOS to gain multiple squared correlations for TAM elements and regressions weights for the paths connecting these elements. A basic TAM model and a full model connecting different levels of elements were simulated. Results are that Incentive to Use and Technical Support are the main influence factors in either model. (Perceived) Usefulness of CRD methods greatly influences the Attitude toward use while the direct influence of Ease of Use is insignificant and has a different polarity among methods. In the full model multiple squared correlations are increased and new paths are identified. The survey revealed that all CRD methods are considered useful. Students perceive one affordance-based approach (Cormier) as more difficult and are uncertain about their own performance using this method. The framework allows making inferences from results of one means of analysis to another. These inferences can be used to validate results or to examine specific factors such as the role of designer experience. As it has been shown that the right use of design tools is positively correlated to the success of product development projects, the framework can help companies make decisions about which CRD method to use when and how to improve the acceptance of a specific CRD method. The framework also offers insights for engineering education. Finally, suggestions for improving the framework itself will be given. This work can therefore been seen as basis for further studies on customer requirement identification methods.