Quality Control Committee
From Digital Libraries - Fall 2008
Contents |
[edit] Committee Members
Paul, Lauren Rodriguez
Contact/point person: Paul Baker (baker5@simmons.edu)
[edit] Schedule/Timeline
Waiting on a Digitization and Web Design schedules to know when we might have something to QC check. In the meantime, we will be discussing what methods we'd like to use in our checking, what procedures we need to employ to make sure everything gets checked, and who will be checking what.
[edit] Meetings and Minutes
Factors to consider in Quality Control:
- Operability: does the site work equally well on Macs & PCs, and on Explorer & Mozilla, etc.?
- Consistency: Are identical things always called the same thing? Especially important for keywords & search terms.
- Synchronization: Are images always accompanied by the correct metadata, etc.?
- Linkages: Do all links work? Do they all take you where they are supposed to?
- Proofreading for spelling errors, incomplete or incoherent sentences, etc.
[edit] White Paper
Quality Control White Paper
Quality control is the last step in line in the process of creating the digital library, or perhaps penultimate to correcting the errors that emerge in the course of quality control. Being very near the tail end of things, quality control is most particularly subject to compression when the schedule gets tight. As such it is needful for the Quality Control committee to be flexible in its goals and expectations. It would take a prodigious effort to guarantee that the library is utterly free from error, more effort than would be sensible even if it were feasible. It always would have been necessary to set some limits, and with much of the library remaining incomplete until the last minute, the quantity and the very quality of the Quality Control that can be done is less than ideal, and perhaps even less than sufficient.
When there isn’t enough time to do a thorough job, there is a degree of triage necessary. Some fixes take priority over others. A broken link is more significant than a spelling error. A glitch on the homepage is more significant than one on an internal scrapbook page. Some problems are easier to fix than others, and the Quality Control committee will not necessarily know which is which.
In practice, quality control is not nearly as exact a science as it might seem. While it is clear when a link does not function or when "Simmons" is spelled "Smimons," it is less clear cut when a passage uses too many or too few commas. A sentence which seems unclear to a reader doing quality control may have been thought quite clever by the writer. A degree of tact is required to avoid unnecessarily stepping on toes. Quality control is essentially a process of finding fault; and while no one wants the project to go live with significant faults, no one enjoys having their faults found. While people might appreciate major faults being discovered earlier rather than later, they are liable to resent their work being inspected with too fine-toothed a comb. It is best for the Quality Control committee to cooperate with the other groups, but not too closely. The other groups will of course themselves be looking for problems within their own work, but neither side should rely on the other to catch things. Quality control should not assume that any task is going to be done right, and others should not assume that quality control is going to find their mistakes.
With the project running up very tightly against the deadline, the window of opportunity for quality control gets very small. One has to work quite closely on the heels of others. Quality control risks crossing the line into kibitzing; you don’t want to lean over the shoulders of others and point out errors that they are already aware of and in the process of correcting. Neither do you want to nag them about minor defects in a finished portion of the project when they are frantically attempting to complete another equally essential portion. It can be difficult to tell what input is constructive and what is intrusive when most of the communication is done through email.
As time grows short in which to complete the project, it becomes necessary to revise the quality control procedure to fit existing conditions. When only a portion of the time anticipated for quality control is actually available, one needs to decide which portion of the process should be prioritized. As the window of opportunity for quality control gradually closes, more and more pieces of the process are set by the wayside. Those parts of the project which are completed early receive a disproportionate level of scrutiny. (I trust that the Content Committee’s members aren’t too thin-skinned.)
As the project has panned out, the Quality Control committee is doing less work than had been anticipated. While it might be expected that this would be welcomed as a relief, in truth it is rather disappointing. We were prepared to put in some intensive work in the last stages of the process, but the other work was not complete in time to leave room for us. It would have been different if the Quality Control committee’s work had proved unnecessary, but it was frustrating to not be able to make a contribution to the project that was needed, and which we were willing and prepared to do.
