Human-Centered Design (Part III)
A Summary of Don Norman’s Design of Everyday Things
In this summary, I discuss Chapters 5, 6, and 7 of Don Norman’s The Design of Everyday Things [1.]. The fifth chapter (“Human Error? No, Bad Design”) provides a detailed look at the various kinds of errors humans make, the sixth chapter (“Design Thinking”) explores the human-centered design process and the role of design in the broader product development lifecycle, and the seventh chapter (“Design in the World of Business”) explores the corporate (and consumer) factors influencing design and the position of products within an increasingly human-designed world. While Chapter 5 primarily continues the work of previous chapters (arguing in favor of a psychological approach to design through its presentation of the ways human psychology can butt up against bad design to produce potentially disastrous results) Chapter 6 serves as the book’s culmination, presenting the espoused human-centered design and related concepts, and Chapter 7 contextualizes the incremental nature of human-centered design in light of predictions for a future that is different from both past and present and yet grounded still in human psychology (augmented though it may become).
The discussion of error presented in Chapter 5 is centered around the post hoc investigation of error. Norman states that, all too often, when an error occurs, the error is traced back to its most proximate human cause, and once that cause is found, rather than seek remedy for the source of human error (to prevent errors of the same kind from recurring), the error is viewed as an entirely avoidable thing, and no changes are made to the process in question. Norman disagrees with this method of root cause analysis, however, and presents in contrast the Swiss cheese model of accident causation, which claims that accidents are seldom the product of a single root cause and instead derive from several causes that line up just right (as the holes in a sequence of Swiss cheese slices would need to align for a dowel to skewer them neatly). As such, Norman argues for a holistic model of error causation that appreciates the various types of errors that occur and accepts the responsibility of design for errors (as errors often result from deliberate violations of standards that are accepted (and sometimes rewarded) under everyday circumstances).
Regarding error types, while his ontology expands to deeper levels of classification, Norman states that, at the highest level, every error can be classified as either a slip or a mistake; slips occur when a person intends to do one thing and does another instead (and they subdivide into action-based slips and memory-lapse slips), and mistakes occur when a person establishes the wrong goal or the wrong plan to reach the right (or wrong) goal (and they subdivide into rule-based, knowledge-based, and memory-lapse mistakes). Furthermore, while mistakes are more likely to affect the novice who has not yet committed processes to memory, slips are more likely to affect the experienced practitioner for whom familiar actions require little thought to accomplish correctly, leaving them open to flawed execution. Norman claims that, by understanding common types of error, designers can strategically incorporate redundancy and complexity into design toward the reduction of error.
The presentation of error concludes the foundational argument for a psychological approach to design, and thereafter, Norman presents human-centered design as an approach to design thinking. Tying back to his discussion of error, Norman states that design begins with the identification of the correct problem to solve, and he depicts the design process as a double diamond that diverges before it converges to the correct problem and then diverges before it converges to the correct solution. Represented in a different way, the human-centered design process has four main steps: (1) observation, where design requirements are determined; (2) idea generation, where potential solutions are brainstormed; (3) prototyping, where simple versions of the potential solutions are built; and (4) testing, where the product is tested on its intended population. Furthermore, because Norman emphasizes the importance of iteration in design, he describes these steps with the expectation that they will be revisited throughout the design process.
Despite the perhaps common belief that technological change is typically the result of radical shifts in paradigm, Norman presents it as something that takes place over decades or even centuries, with many early attempts to innovate failing to find a foothold in the market. This aspect of design is particularly interesting to consider in light of the recent advances in text generation. More specifically, while text generation technologies such as chatbots are decades old (simple rule-based chatbots have been around since the 1960s when the ELIZA chatbot was developed by researchers at MIT [2.]), the technologies that are currently being brought to market are being positioned as general purpose tools that can be used for linguistic tasks other than chatting (e.g., classification, summarization). Thus, perhaps only now, after early attempts at innovation failed to make a large impact in the technological landscape, chatbots are ready to take their place as a viable commercial technology (or perhaps not — their fate at present remains in the air).
Although Norman presents numerous ideal models and frameworks throughout The Design of Everyday Things, he is at his core a realist who acknowledges the complexity of design in practice. Perhaps best exemplifying this realism is his law of product development, which states that the product development process is behind schedule and over budget from the day it begins. Along similar lines, Norman discusses the difficulty of establishing firm standards in industries that are undergoing rapid change (i.e., by the time a standard is developed, it might be outdated due to advances in the industry). Although standards (when established in time to be useful) can serve constructive purposes such as setting user expectations and ensuring product quality, competitive forces can result in a false form of standardization with opposing effects: featuritis. Norman defines featuritis as the temptation of companies to overload a product with features by adopting the features of their competitors — regardless of the needs expressed by their users — to avoid falling behind.
In conclusion, the final three chapters of The Design of Everyday Things describe the types of errors that occur when humans use tools to complete tasks, the human-centered design process and key techniques such as the double-diamond model and iterative design, and the business considerations that influence the design process. Through these chapters (and the book more broadly) Norman presents design as an aspect of product development with real importance and serious potential to affect not only the initial adoption and long-term usability of products, but also the safety and moral obligations of products. In this way, design is grounded beneath aesthetic appeal, as something foundational and nonnegotiable in the development of the products, environments, and abstract structures that people must learn to navigate well to be at home in their world.
References
- Norman, D. (2013). The Design of Everyday Things (revised and expanded edition). New York: Basic Books.
- Weizenbaum, Joseph. (1966). “ELIZA — a Computer Program for the Study of Natural Language Communication between Man and Machine.” Computational Linguistics.
