The twelve interaction design spectrums
Constrained to flexible
Software can constrain and guide a person, or it can aim to offer them as many options as possible and let them decide. Opinionated software is more constrained, for example.
Familiar to novel
Familiar interaction design takes advantage of a person’s existing knowledge. Novel interaction design requires a person to explore and become familiar with it.
Operationally to perceptually simple
An interface is operationally simple if there’s one interactive element per action. This can lead to clutter. It is perceptually simple if many actions are combined into one interactive element (e.g. a menu). This can lead to discoverability issues.
Screen-by-screen to all-at-once
You can keep each screen simple and put all elements on separate screens, or you can put everything on one screen. e.g. a website with a different page for each topic vs. a website with everything on one page, which requires scrolling.
Shallow to steep learning curve
A shallow learning curve makes use of more helping hands to help a person become familiar with the interface. It might lead to frustrating “slowness” for some users. A steep learning curve drops a person in the deep end, but might suit people who are happy to tinker to understand the interface.
Static to animated
An interface can use no animation, or lots. Animation can help people to understand how the interface works.
Broad to deep information architecture
A broad information architecture is generally easier to understand, but takes up more space. A deep information architecture takes up less space, but has many more levels to explore.
Single-purpose to multi-purpose features
Single purpose features are designed to do one thing. They might be easier to understand, but there will be more of them. Multi-purpose features are generally designed to support different use cases, but they are naturally more complicated. e.g. one dialogue box for each text setting in a word processor, versus one dialogue box with all of the text settings.
Single-approach to multi-approach actions
How many ways are there to take an action? e.g. you can make some text bold with a keyboard shortcut, a toolbar button, or a menu option.
Close to distant controls
Controls can sit near the element they affect, or somewhere else. For example, a 3D modelling app might have controls that affect a 3D model next to the model, or in a toolbar far away from the model.
Modeless to modal
Do various actions put the software into different modes? e.g. You can show a delete action for every element in a list, or wait until the person turns on an “edit” mode before you show actions like delete.
Suitable to flexible
An interaction can be designed to suitable, so that it fits its context well. Or it can be designed to be flexible, so it works in many contexts. It’s rare if not impossible for a pattern to do both well.