The fourth workshop on end-user software engineering

Abstract

1. WORKSHOP DESCRIPTION A recent survey has shown that the number of end-user programmers in the United States is 11 million, as compared to only 2.75 million professional programmers [5]. Unlike professional programmers who typically work with generalpurpose programming languages like Java, C, and C, most end-user programmers work with languages and environments such as spreadsheets, web authoring tools, MatLab, and macro languages. The software developed by end users include spreadsheets, dynamic web applications, resource coalitions, scientific simulations, educational simulations, and many others. End users have very different training and background when compared with professional programmers, and few have backgrounds in software engineering practices. End users are unlikely to invest time and effort in learning software engineering techniques and approaches. This is because end users have very different motivations and work constraints than professional programmers. End users are not, in most cases, striving to deliver software. Instead, their focus is on the results of the software. This difference may seem subtle, but the result is that, since the software itself is not viewed as an important asset, there is little perceived need to learn software engineering techniques. Numerous studies have reported that a bulk of the software developed by end users has faults in them (e.g., [3, 4]), and quite a few of these errors result in substantial financial losses to companies and other institutions [1]. The primary challenge for software engineering researchers working in this domain is to find ways to help support end users in their work while respecting these differences. We need to develop approaches that assist end users within the contexts they are familiar with, rather than attempt to train end users to become software engineers [2]. For example, most end-user programming environments are geared towards supporting