Skip to content

1 Human centered design

Jump to bottom
SOLV-Code edited this page Oct 13, 2023 · 13 revisions

Key Concepts

  • Human-centered design (HCD) focuses on developing tools/systems that help actual people get work done in their real work environment, rather than building some hypothetically optimal technological solution for individual tasks.

  • A highly sophisticated data visualization tool designed to support complex fisheries decision will not become part of the operational routine if it is not fed with the required data in a timely manner. A fundamental bottleneck is the information flow from raw data into the tool, and the whole decision support process will fail if it is based on unrealistic expectations of individual contributions to the information system.

  • Think of data contributions not as a stand-alone task that happens in isolation, but part of a web of computer-supported activities, and for each specific product, such as a summary report or model input file, think about the full production lattice, which includes all the people, data collection, review steps, and computer-based tools required to get to the final product (Kling and Scacchi 1982). When all the pieces of the production lattice work smoothly together, they are generally invisible to the end-users and the successful product is credited to the program that handles the final step (e.g., "this new database program created a really useful dashboard"). When the final product is incomplete, or delayed, or out-of-date, the blame typically gets assigned to people ("X didn't enter the data right, Y was on leave so the updates weren't signed off in time, Z doesn't have the right database version to generate the dashboard").

  • In human-centered design we look at the whole production lattice, and the focus shifts to finding the most efficient combination of process and tools to accomplish the actual work requirement. For example, instead of trying to improve just the data entry task, which is only a small component of what needs to be done, try to build a process that makes it feasible for the people who coordinate and implement stock assessment field studies to efficiently share both the data and important context for each data point with minimal additional burden.

  • When a data management system is designed as a discrete entity, with a focus on technical sophistication, then it comes with the implicit assumption that all the other components of the production lattice will just fall into place, or can be ordered to "just make it so". This amounts to assuming a best case scenario, where every person in the sequence from doing the surveys to the final estimates will have time available, will allocate available time to this task rather than another, and will have the required software and hardware to do the task when they make time for it. In contrast, when approaching an information system from a human-centered design perspective, we look at the whole production lattice, and ask "what is likely, given the current work environment for each individual contributor?"

  • Human-centered design differs from usability studies. In usability studies the analysts/developers examine how a user interacts with a specific tool, in order to find improvements to the tool (e.g., "how can we improve the layout of the data-entry template for the database and how do we best train staff to use the database?"). (Gasson 2003)

  • Many computer interventions / information systems fail, but common causes for failure are social/organizational, not technical (Bell 2000, Warren 2000).

  • In fisheries organizations, barriers to timely data contribution include some technical issues (e.g. software compatibility, learning curve for database interface) but often go beyond any specific challenge with the particular data set. Most commonly it is matter of the total workload faced by key individuals in the stock assessment programs. Technical improvements or new software can save some time in the long run, but require time commitments in the short term that are often not feasible. In the short-term, we need to work with existing tools that contributors are comfortable with.

  • The international standard for human-centered design (ISO-9241-210:2010) identifies six guiding principles: (1) the design is based upon an explicit understanding of users, tasks and environments; (2) users are involved throughout design and development; (3) the design is driven and refined by user-centered evaluation; (4) the process is iterative; (5) the design addresses the whole user experience; (6) the design team includes multidisciplinary skills and perspectives (Maguire 2001; Thomas, Remy and Bates 2017)

  • Beyond the specific application to fisheries data compilation we are focusing on here, the fundamental concepts of human-centered design are being re-evaluated and expanded to include considerations like social fairness, carbon footprint, and full product life-cycle from manufacture to disposal (Thomas, Remy and Bates 2017).

Selected papers (follow the link for extracts)

Additional References

Checkland, P. 1981. Systems Thinking Systems Practice, John Wiley & Sons, Chichester UK.

Maguire, M. 2001. Methods to support human-centred design. Int. J. Human-Computer Studies (2001) 55, 587-634. Available online.

Clone this wiki locally