Strategies for Research in Design

Carnegie Mellon University, Spring 2007

Mark D Gross, Shaw (Computer Science), Finger (Civil Engg), Herbsleb (Software Engg)

Prerequisite:       None



Graduate course to prepare PhD students to understand how to conduct design research.


Why it is worth understanding design processes (the argument for “research in design”)?

Design spaces: design process as exploring a multi-dimensional space, bounded by constraints, to optimize an objective function; “classical” engineering model of design, Pareto-optimality, design when you don’t know a-priori the objective function; sparse spaces, adding dimensions to the design space during designing. Design-as-search.

Design decisions are coupled. Identifying relationships between design decisions (e.g., adjacency matrix to represent dependencies); clustering to identify closely-coupled decisions.

Dealing with other designers: You are not alone. Hierarchies of control in design: design as a social negotiated act over time. Designers control different levels of a design; relationships among the levels; “dominance” and capacity testing.

Design for change: What you design today may be around in ten, twenty, thirty years. How to design for evolutionary lifecycle change?

Pattern Language – from Alexander to the Gang of Four — how to turn observed relationships between design and effect into a structured representation that can be applied in future designs. Representing knowledge about a specific design domain.

Participatory design – involving end-users and other stakeholders in design processes.

End-user designing: recognizing that end-users want and need to extend designs – seeing end users as designers too (relates to “hierarchies of control”); how to provide end-users with simple and powerful means to work within systems to extend them.

Problem-seeking; identifying the problem is part of design: “system-analysis”; eliciting requirements; setting specifications.

Design rationale; accounting for design decisions; structuring representations of design alternatives, from IBIS to hypertext; “design knowledge management”: design as argument.

Representations for designs: Representations for quickly exploring and comparing alternative designs. Sketching and prototyping during early design;


Course Requirements:

class participation in discussion.


Required Texts:



Learning Objectives:

familiarity with approaches in design research