Trabajo y proceso de trabajo

Heuristic Evaluation [Nielsen, 1994b] is a well established technique for evaluating inter- faces. It has the very desirable property that it can be conducted quickly, and relatively inexpensively; suitably expert evaluators can make use of a set of heuristic guidelines, or rules, to evaluate an interface. One essential requirement is a suitable set of guidelines. These can be used both for summative evaluation of an interface and they can also play an invaluable role as an aid for designers as they explore the design space: with a good set of design guidelines, or heuristics, they can use these to inform that exploration and conduct very quick formative evaluations of a range of new possibilities.

Certainly, the very broadest design guidelines, which apply to all interfaces are still of value for tabletops. However, established heuristics – including those for conventional and non-conventional interfaces – fall short of some critical interface challenges for the tabletop interface that are not present in traditional interfaces. For example, multiple, collocated users introduce a dichotomy of interface requirements, as multiple people often results in the interface having to support multiple simultaneous user goals [Tang et al., 2006]. Traditional sets of heuristics are typically designed for a single user with a single goal. While there are many sets of guidelines, some generalised [Nielsen and Molich, 1990,

2_{Aesthetics are not a current focus of development – they are considered insofar as to not detract from}

Shneiderman, 1992, Nielsen, 1994a] and others highly specialised (e.g. the Mitre Set [Smith and Mosier, 1986] and others [Brown, 1988, Mayhew, 1992]), there is no set that properly supports the particular class of tasks involved in designing new interaction elements for tabletop interaction.

Scott et al. [2003] conducted an analysis of existing tabletop research from the Human- Computer Interaction (HCI) and Computer Supported Cooperative Work (CSCW)fields. From this, they propose a set of heuristics to facilitate the design of tabletop interfaces, and outline directions for future research. The heuristics proposed are summarised in Table4.1.

1. support interpersonal interaction

2. support fluid transitions between activities

3. support transitions between personal and group work

4. support transitions between tabletop collaboration and external work 5. support the use of physical objects

6. provide shared access to physical and digital objects 7. consider the appropriate arrangements of users 8. support simultaneous user actions

Table 4.1: Tabletop Interface Heuristics from Scott et al. [2003]

A clear focus of these heuristics is supporting transitions (such as between activities or people), rather than usability. As such, the focus of these heuristics seems to lie with the design of the overall user experience, at a high-level. Moreover, they involve both hardware and software. In this thesis, the goal is the design of software to support interaction.

My work is particularly concerned with software design at a low level, where the design is an exploration of possibilities for elements within the interactive experience. Heuristic evaluation at this level, such as for the design of a specific user interface widget or gesture, is important to the design of a highly usable tabletop user interface framework. Scott’s guidelines, being at a very different level, do not address this. In addition, there are some aspects of tabletop interaction, such as clutter, that these heuristics do not cover.

Instead, much of the design influence was gained from Nielsen’s original set of heuristics. While these heuristics were developed with only traditional, vertical computing interfaces (typically data entry interfaces), many continue to be relevant for the highly graphical and immersive pervasive computing interfaces such as those on a tabletop. Nielsen [1994b] suggested these ten, to provide coverage of published usability problems (quoting from [Baecker et al., 1995]):

N1: Visibility of system status. The system should always keep users informed about what is going on, through appropriate feedback with reasonable time. N2: Match between system and the real world. The system should speak the

users’ language, with words, phrases and concepts familiar to the user, rather than system-oriented terms. Follow real-world conventions, making information appear in a natural and logical order.

N3: User control and freedom. Users often choose system functions by mistake and will need a clearly marked “emergency exit” to leave the unwanted state without having to go through an extended dialogue. Support undo and redo. N4: Consistency and standards. Users should not have to wonder whether dif-

ferent words, situations, or actions mean the same thing. Follow platform conventions.

N5: Error prevention Even better than good error messages is a careful design which prevents a problem from occurring in the first place.

N6: Recognition rather than recall. Make objects, actions, and options visible. The user should not have to remember information from one part of the dialogue

4.1. Interface Design Approach CHAPTER 4. USER VIEW

to another. Instructions for use of the system should be visible or easily retrievable whenever appropriate.

N7: Flexibility and efficiency of use. Accelerators – unseen by the novice user – may often speed up the interaction for the expert user such that the system can cater to both inexperienced and experienced users. Allow users to tailor frequent actions.

N8: Aesthetic and minimalist design. Dialogues should not contain information which is irrelevant or rarely needed. Every extra unit of information in a dialogue competes with the relevant units of information and diminishes their relative visibility.

N9: Recognise, diagnose, and recover from errors. Error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution.

N10: Help and documentation. Even though it is better if the system can be used without documentation, it may be necessary to provide help and documentation. Any such information should be easy to search, focused on the user’s task, list concrete steps to be carried out, and not be too large.

Intrinsically, the tabletop brings support for N2 to a new level beyond what traditional computing interfaces can provide. By providing a graphical, immersive and direct interac- tion with virtual objects, a very tight match between the system and the real world can be provided. The interactive graphical environment also enhances recognition of objects and actions (N6), rather than having to rely on recall.

However, complications of multi-user, pervasive interaction, and problems of orienta- tion and text readability exacerbate the straightforward support for many other Nielsen heuristics. For example, should undo (N3) revert only the command of the user invoking the undo, or should it be a “global” undo that takes the entire interface back to a previous state? Shortcuts (N7) are difficult to implement for expert users when a keyboard is not available. Similarly searchable, online help (N10) might not be appropriate for viewing on the table – a peripheral display may be better suited to this task.

We are at a crucial point in time where we have the potential for tabletops to break away from the interaction that has come to be normative; that seen in Microsoft and Apple desktop products. Tabletop hardware is currently evolving quite fast and we would want Heuristic Evaluation to be independent of the hardware3. Because there are no established interfaces yet, this is a time when we can consider a broad range of possibilities without concerns of backward compatibility and we are less fettered by entrenched mental models: users coming to a very different interface may be more open to, and willing to learn about, new and unfamiliar ways to interact.

Nielsen’s heuristics provide a starting point, but cannot be blindly applied to tabletop interfaces. Really, a new set of heuristics is required for tabletop interfaces. Some of the existing heuristics are inappropriate for direct-touch interaction, or the intent needs to be re-purposed to suit the kinds of applications that the tabletop is designed to support, as there is a difference. But this is beyond the scope of this thesis. Such a set should be developed as a community effort and when the field is more mature, with a chance to reflect upon real-world use of tabletop interfaces.

Instead, our approach was to develop our own guidelines to drive the design, as presented in Section 3.1. In this chapter, we give these guidelines more context and further indicate how our design was influenced by these guidelines. This is summarised in Table 4.2.

Core Function

Design Influence Guideline

Select Photo corners, colour shows selecting user G2e, G1a

Move Tight coupling with selection point G1a

Rotate Large activation area in photo corner, bound to photo

G2b

* Resize Same activation as for Rotate (fewer actions to remember); Easy, flexible enlarge when vision is poor

G1b, G2b, G2e

Copy Builds on move (predictable) G1b, G1c

Grouping Postcard influence – attach to back after flipping over, tools to help manage clutter, support limited reach and permissions

G1a, G2e

Delete (Hide)

Avoiding cluttered displays; gradual, reversible ac- tion with continuous feedback

G2d, G1b, G1d

Capture Photo frame or viewfinder G1a, G1d, G2b

* Rotate and resize is a combined action –rosize

Table 4.2: Design Influences on Core Functionality in Cruiser