Evaluating the Evaluator: Modeling Systematic Data Analysis Strategies for Software Selection

摘要

Developers of interactive software are confronted by an increasing variety of software tools to help engineer the interactive aspects of software applications. Not only do these tools fall into different categories in terms of functionality, but within each category there is a growing number of competing tools with similar, although not identical, features. Choice of user interface development tool (UIDT) is therefore becoming increasingly complex. Evidence suggests that, when selecting UIDTs, industrial software developers rely heavily upon right-brain (intuitive) decision-making [13], based on little more than an ad hoc inspection of marketing material, journal reviews, and recommendations from colleagues [9]. Although potentially adequate in some cases, this can result in poor choices, especially when the aforementioned information is not relevant to the context in which the tool is to be used. Acknowledging the existence of, and place for, intuition or non-rational decision-making in the overall process of UIDT evaluation and selection, this research facilitates integrated style UIDT selection; it provides a mechanism which allows evaluators to approach UIDT selection from an analytical perspective but which does not suppress intuitive decision-making so that, where necessary, it can be used to handle areas of uncertainty, such as trade-offs. By promoting an analytical approach to UIDT selection, it is possible to guard against inappropriate use of intuition and thereby increase the effectiveness of data analysis leading to greater likelihood of tool acceptance.