Design and Evaluation
Andy Smith1 and Lynne Dunckley2
(1) Department of Computing and Information Systems, University of Luton, Park Square, Luton, LUI1 3JU, UK. Email: andy.smith,@luton.ac.uk (2) Faculty of Maths & Computing, The Open University, Walton Hall, Milton Keynes, MAA 3JU. Email:L.Dunckley@open.ac.uk
This paper analyses methods for obtaining a combined usability metric for applications. The rationale for the development of this research emanates from the specification of LUCID, a novel interface design method which has been proposed ( Smith and Dunckley, 1998). It is however suggested that the techniques proposed have generic applicability.
In developing tools and techniques for the determination of a combined metric there are two main issues. Firstly a range of potential metrics need to be ascertained. Secondly the major potential metrics need to be prioritised in some manner, so that by allocating weightings to each an overarching value can be ascertained. These issues will each now be further explored.
There are two options for the process of proposing a range of metrics for possible inclusion within a combined metric. Firstly it may be sufficient to rely on developer introspection. The evidence gained from this work and other related studies ( Smith and Dunckley, 1999) would suggest however that collaboration between a number of developers is to be recommended above the introspection of a single expert alone. The alternative approach is to directly involve users and developers in a participative environment.
There are also two options for the process of prioritising the metrics that are proposed. Firstly Lin, Choong, and Salvendy's ( 1997) single usability index