2.6 Interaction Design Characteristics
2.6.1 Interface Design
The TUI interface design and interface functions must meet a balance. A balance exists between being too easy to use and being too difficult to understand. Balanced systems should be designed until more research is conducted to clearly define the circumstances where complexity of the interface is linked to success in completing goals.
A debate rages as to whether intuitive easy to use interface designs, or difficult to understand interface designs, are better for learning and problem solving. The design elements of metaphors and the aesthetics of the tangible objects affect the level of complexity of the design of the interface. The representations and metaphors in a
tangible user interface design provide significant contribution to the overall ease of use of the interface. The metaphor of objects (i.e. the resemblance of interface objects with their real world counterparts), paired with the semantic directness (the resemblance of interface objects to the users intent E.g. floppy disc icon for save) influences how well users interact with the system.
Easy to use interfaces possibly make us complacent. O’Malley & Stanton Fraser, (2004) report that research shows ‘transparent or really easy-to-use interfaces sometimes lead to less effective problem solving.’ This is substantiated by Marshall, (2007) who asserts ‘The easiest to use or most concrete interface does not necessarily lead to the greatest performance in problem solving and learning’. The justifying argument is that the mind is not engaged because the operator does not have to think about the interface, therefore they will not think creatively when the interface is applied. The author believes that concrete interface elements look and appear complete because they have sufficient detail to be perceived by human nature to be a finished product, whereas a hard to use interface, or rough representation, is clearly not a finished product thus encouraging the mind to fill the gap, consequently demonstrating problem solving and learning.
Complex interfaces are inefficient and frustrating (Sharp et al. 2011) . In an abstract, or difficult to use interface, users become confused using the interface. They do not have a working knowledge of the interface, so they are restricted in how they can apply the program, thus their results will be diminished. A user still may know what they want to do, but not be able to do it because of a difficult to use interface (O’Malley & Stanton Fraser, 2004). This may well be true for novices. People with limited experience in the program will focus on learning the interface, rather than completing their work that must be undertaken using the program. However, an experienced, or expert, user is highly efficient in using the interface and can produce higher quality output simply because they know how to most effectively use elements of the program. This is in contrast to a novice who will produce output using restricted functionality working only within their limited knowledge of the interface. A common example is Blender, an open source 3d modelling program.
A complex application will likely have a complex interface (Benyon 2010). An easy to use interface is not the most efficient use for an expert, or experienced user, however it may allow novices to work in a knowledge domain where they do not have experience or expertise. A good example is the field of Computer-aided Design (CAD). Originally, CAD
was solely the domain of experts in construction and design (Architects, Engineers and Draftsmen), but is now accessible to the general public via easy to use limited function interfaces. Is this wise?
‘If you close it all in and specify everything there is no room for the imagination to play. You must leave big enough holes for the imagination to play.’ (Buxton 2007, p. 407) Buxton’s statement may be more of a comment aimed at the design of applications, rather than poignantly aimed at learning applications where it would disagree with research by (Marshall, 2007, p. 167):
Interfaces that constrain the ways that learners can use them or which introduce interaction costs can lead to increased planning and reflection, which can in turn lead to improved learning. It is possible that if tangible interfaces support easy manipulation of concrete objects, that they could in turn lead to decreased reflection, planning and learning.
Marshall’s counter argument is an inference using a known result to imply the opposite could be true. As yet, there is a no empirical evidence to justify this inference, which is why Marshall, (2007) qualifies it by starting with It is possible that and including could lead to. Still, Marshall’s suggestion in his highly cited paper has the effect that it brings into doubt positive benefits of intuitive, easy to use design, within a tangible interface. This side of the debate as presented by Marshall (2007) make it difficult to determine which elements contribute to learning. It needs to be acknowledged that Marshall (2007) is frequently cited as validation in the opposite of the conclusions of the paper.
The debate continues and requires empirical evaluation to determine the best level of design for progress and efficiency (Heijboer & Van Den Hoven, 2008). ‘The point is to channel the learner’s attention and effort towards the goal or target of the learning activity, not to allow the interface to get in the way.’ (O’Malley & Stanton Fraser, 2004, p. 4). It is this author’s opinion that an intuitive natural user interface is a better system for both learning and general productivity. The interface should aid, not detract from, the user experience.