CCE Theses and Dissertations

The Failure Analysis Matrix: A Usable Model for Ranking Solutions to Failures in Information Systems

Date of Award

2000

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Graduate School of Computer and Information Sciences

Advisor

Laurie Dringus

Committee Member

Sumitra Mukherjee

Committee Member

John A. Scigliano

Abstract

Failure Modes and Effects Analysis (FMEA) is a Six Sigma tool for identifying, analyzing and prioritizing failures and solutions. FMEA is widely used in many different industries worldwide. The FMEA model is neither easy to learn nor easy to use. Effective FMEA automation has been an elusive goal for some time. The scope of meaningful FMEA automation has been limited to specific proprietary or academic application domains where substantial time and effort have been invested. Commercially available FMEA software packages do little more than reduce clerical effort. There has been no published research on the usability of FMEA. Most of the recent FMEA research has applied various artificial intelligence technologies. The vast majority of FMEA research has been directed toward manufactured products rather than information systems. The author has examined the use of FMEA for software development.

The author's goal for this dissertation was the creation of a usable failure analysis matrix (FAM) model for prioritizing solutions to failures in information systems. The two-dimensional F AM worksheets are smaller than the linear FMEA worksheets, requiring less scrolling. The F AM is an alternative method to help identify the most important potential failures or failures and help prioritize alternative solutions using approximate expected costs. This new tool, implemented in Excel, provides integrated data entry and reporting. The F AM may be used without having detailed information.

The author validated four usability attributes for the F AM: usefulness, ease of use, ease of learning, and satisfaction. The validation process employed expert reviews, usability testing, and a usability questionnaire. The final mean value of 5.30 for usability questionnaire items compares favorably with a neutral value of 4. The F AM underwent expert reviews with reviewers from three different domains of expertise: usability, FMEA, and Six Sigma. Usability test subjects were 20 GE Industrial Systems information systems professionals. The test subjects applied the F AM to real information systems failures.

The author has improved professional practice by applying usability engineering to a problem analysis tool for software development. Expert reviews and usability testing both proved to be applicable to the design of the F AM. The expert reviews resulted in the inclusion of several important attributes of the FAM. The usability testing resulted in shorter task completion times and less severe usability problems.

This document is currently not available here.

Share Feedback

  Link to NovaCat

Share

COinS