Authors
Jeffrey Carver, Marc Fisher, & Gregg Rothermel
Abstract
Spreadsheets are one of the most commonly used types of programs in the world, and it is important that they be sufficiently dependable. To help end users who create spreadsheets do so more reliably, we have created a testing and debugging methodology and environment for use in spreadsheets, known as the WYSIWYT methodology.
Our prior experiments with WYSIWYT show that users can utilize it to ensure that their spreadsheets are more dependable, but these experiments to date have considered only an unfamiliar prototype spreadsheet environment, and have not involved spreadsheet creation tasks. In this work we conducted a controlled experiment that addresses these limitations.
The results of this study indicate that the use of WYSIWYT did not affect the correctness of spreadsheets created by users, but it did significantly reduce the amount of effort required to create them. Further, the subjects' evaluation of the help provided by WYSIWYT was very positive. Our results provide several insights into the use of the WYSIWYT methodology by end users.
Sample
The subjects who used WYSIWYT took significantly less time to complete the task than those not using WYSIWYT.
This result suggests that the facilities provided by WYSIWYT allowed the students to reach a conclusion that they had performed enough validation activity on their spreadsheet more quickly than they otherwise would have.
Publication
2006, ACM/IEEE International Symposium on Empirical Software Engineering, September, pages 278-287
Full article
An empirical evaluation of a testing and debugging methodology for Excel
