Tabular data arrangement is a raw dump that expects the viewer to carry the full burden of analysis and interpretation. This makes sense when the software can’t be expected to anticipate how the user may choose to evaluate the information.
Tables are good when:
- All data columns are of equal or unknown importance.
- Horizontal space is not a limitation.
Tables are a fallback when you have no reasonable control over the data points being displayed, ie the user can assemble their own table from a grab bag.
Opinionated data arrangement (that’s what I call your lists) establishes a hierarchy of data and makes more flexible use of space. What we now widely refer to as “cards”, made famous by Google Now, is one example of opinionated arrangement. In this case, users are looking for a known structure and priority of information and the presentation can aid their evaluation.
“Lists” are appropriate when:
- You know what will be displayed.
- There is a useful hierarchy to the data.
Arranged presentations can accommodate a limited set of user configurable data points. The trick is to set up a framework for what types of data will go where. You still need to understand the potential variations in hierarchy.
Sorting is not an issue
The concern of sorting a non-tabular presentation is a pattern that has been solved successfully in plenty of ways. Look around for filter and sort patterns. Just a few solutions available:
- Limited sets of sort options can be presented as buttons for quick switching.
- A combo box / drop down solution works for long lists.
- An exposed list in a left or right column can coexist with a filter controls.
If there is any added weight to the sorting process, the assumption is that an ordered presentation of data makes up for the cost.