Thursday, April 17, 2008

Thoughts on Quality Management & Contol

Stories in the media about toxic chemicals in toothpaste, and lead in the paint on children's toys highlights the ongoing importance of quality control. Producing an online database, and the software to interact with it (as in a project for my Systems Analysis and Design class) probably will not affect the health and safety of the users, but it still needs to be a quality product. One of the most critical aspects of quality control when creating online systems is the protection of personal data. Searching the LexisNexis database using the term "identity theft" generated 999 results. Titles included "To Fight Identity Theft, a Call for Banks to Disclose All Incidents," "The Identity Theft Scare," "Employers joining identity theft battle; But many unconvinced job-related assistance outweighs the added work," and "A Banner Year for Identity Theft; 2006 was a banner year for identity theft - at least it was for me." The ease with which a tech-savvy individual can steal information gives me pause in this assignment. Pioneers in "quality management" such as Deming, Juran, Crosby, Ishikawa, Kaplan and Norton have given us a solid foundation of principles and standards on which to build. But the fluid, changing nature of electronic systems, and the security weaknesses that result from these changes, has dramatically increased the importance of having an ongoing quality control process. From a tech librarian's perspective, I wonder how many vendors that build integrated library systems (ILS) have someone on staff to attempt to hack into their systems as they are being developed, checking security.

On a more local note, the system I am prototyping has to be of a certain quality to be accepted and used. In addition to protecting private data, it must be useable and maintainable by non-techies, since my work on the project will end upon graduation. Reliability and usability are the most critical aspects I face in its development. My boss, the Principle Investigator, has a reputation of being able to crash any system you ask her to test. If she is interested in having me develop the system beyond the prototype stage (which is all that is required for my class assignment), it will require careful definition of the end result, and many iterations to result in an acceptable product. I am reminded of a bumper sticker from one of my favorite catalogs, which reads: "Oh no, not another learning experience!"

No comments: