Framing and Tracing Human-Centered Design Teams’ Method Selection: An Examination of Decision-Making Strategies

Abstract

Designers' choices of methods are well known to shape project outcomes. However, questions remain about why design teams select particular methods and how teams' decision-making strategies are influenced by project- and process-based factors. In this work, we analyze novice design teams' decision-making strategies underlying 297 selections of human-centered design methods over the course of three semester-long project-based engineering design courses. We propose a framework grounded in 100+ factors sourced from new product development literature that classifies design teams' method selection strategy as either Agent- (A), Outcome- (O), or Process- (P) driven, with eight further subclassifications. Coding method selections with this framework, we uncover three insights about design team method selection. First, we identify fewer outcomes-based selection strategies across all phases and innovation types. Second, we observe a shift in decision-making strategy from user-focused outcomes in earlier phases to product-based outcomes in later phases. Third, we observe that decision-making strategy produces a greater heterogeneity of method selections as compared to the class average as a whole, or project type alone. These findings provide a deeper understanding of designers' method selection behavior and have implications for effective management of design teams, development of automated design support tools to aid design teams, and curation of design method repositories.

How do Conversational Agents Transform Qualitative Interviews? Exploration and Support of Researchers’ Needs in Interviews at Scale

“…Prior work presents analysis of typical scenarios and decisions that designers make about data analysis (e.g. [1,14,20]. When we understand the needs of researchers in the interview analysis, we will be able to build systems that support such work by summarizing the content into the topics [3,23], extracting sentiments [6] or sequences of Part-of-Speech tags, and other more advanced methods of NLP.…”