What to Model

From Shelley Evenson's conceptual models class: The types of things you want to think about modeling and why:

• Processes. Procedures and particular courses of action or the performance of some composite cognitive activity.

  Why model processes? Breakdowns become obvious. You can tell what is and what isn't part of the process. Discussions can be had about where/when an intervention needs to be made by a designer. Comparing documented processes among parallel approaches can reveal why one is more of less successful than the other.

• Attitudes. A complex mental state involving beliefs, feelings, and values.

  Why model attitudes? People change their attitudes based on lots of things ranging on how they feel physically to how competent they feel about a task. Ideally, experiences should map to whatever attitude the person has at a particular moment (so it feels designed for them personally). Or so that you can choose to not design for an attitude.

• Approaches. Plans of attack. Ways to start doing a task or set of tasks.

  Why model approaches? Because they reveal what the general perception of something is and make explicit the methods used or plan made. Understanding different approaches (especially from superusers) may reveal unique strategies that can benefit all users.

• Elements. All the little bits of something.

  Why bother modeling these? To see the relationships, to get quantitative data (like how many of something there is), to see opportunities amid clusters.

• Archetypal Users. AKA Personas.

  Why model these? Only after synthesizing user data can we begin to create patterns of interactions that smoothly match the behaviors, mental models, and goals of users. Can illuminate lifestage, experience level, motivations. You need them to drive scenarios.

Originally posted on Monday, January 31, 2005