As our business matures, and we get more clinical and manufacturing clients, they keep asking about quality. They have QA/QC teams, who design the process with the intent of taking all subjectivity out. The FDA asks IQ/OQ/PQ documentation, for Installation, Operation and Performance Qualification. They generate voluminous compliance reports, generally of the same bead data every day, so that they can track changing baselines. Which leads me to the conclusion that what industry views as quality is really all about quantity.
I'm all in favor of hard numbers, reproducible experiments, quantified controls, rationalized procedures, and the like, but those are NOT qualitative attributes. Don't try to measure quality, because the more you do, the more it slips away. Quality is a more nebulous aspiration, immeasurable yet visceral. Quality is like art: you can feel it when it moves you or it doesn't, but no reproducible measurements are going to tell you what makes it good. Robert Pirsig (Zen and the Art of Motorcycle Maintenance) said, "Quality cannot be defined because it empirically precedes any intellectual construction of it." So there! Potter Stewart put it more succinctly, when he wrote: "It may be hard to define, but I know it when I see it."
This is just the opposite of the Quality Control industry definition, where Quality is defined as parts per million of materials meeting specifications (Six Sigma). They should call that precision, or standardization, or homogeneity, but not quality. You have to wonder about an entire industry that sets its acceptance range based on alliteration, without any apparent consideration of what they are measuring. It shouldn't be too surprising that these same engineers proudly boast of their methodology as The Toyota Way.
My definition of quality is refreshed every Tuesday, which is Jazz night with the Adam Harris Trio. Adam Harris is a Tree Star employee, who handles most of your free trial licenses to FlowJo, but that's only his day job. His real job is the tenor sax, which he plays with a virtuosity well beyond his years. Since returning from stints in Seattle and Boston, has become a new star of our local music scene. He has a progressive jazz band, which I go hear regularly. Every week it reenforces in me the difference between making a widget, and making a contribution.
Download Link "You Don't Know What Love Is" - Adam Harris Trio
It's quite a stretch to even compare what I do with modern jazz. Everything I create is defined in numbers. Perhaps the common thread is found in the concept of patterns. If pop songs are a sequence of notes, jazz is made of patterns, subject to interpretation and context. The jazz musician has the liberty to interpret the pattern, to offset it, to modulate it, to syncopate it. Similarly, software is the rearrangement of many small patterns. Subroutines and modules are logical patterns that are interwoven to solve the problem at hand. A high degree of individualism and discipline is required to learn one's vocabulary of patterns.
Download Link "Blues In F" - Adam Harris Trio
A required read for any literate programmer, designer, or even biologist (and, admittedly, my favorite bathroom reference material) is Christopher Alexander, et. al., A Pattern Language.The book is a simple taxonomy of patterns found in architecture and urban planning, but the authors delve into the effects different qualities of buildings and towns play upon the residents. They simultaneously detail and humanize the art of building. The software community (among others) has recognized this as a generalized analysis of natural systems, with a concern for quality. Whether the patterns take the form of riffs, or apartment houses, or dialog boxes, the quality of a work comes from the psychological impact that patterns convey, and the subtle effects of its communication to the viewer, listener or user.
Readers of the "Tufts list" of cytometry understand that music reviews are an important part of the cytometry dialectic, because music is a great example of what quality really is. It's contextual and ephemeral. It's a fleeting moment, when the musician doesn't play what you expect, but instead harmonizes with what you expect, or, completely defies what you expect, and, in the process, shows you something of yourself and your assumptions.
My aspiration is to provide exploratory software that doesn't just display the same old process you are accustomed to, but through interactivity gives you some new insight into your data. It may not be measurable, quantifiable, or adequately expressed by the statistics, but at some level, there is a contextual illumination of the data that spans the multiple levels of abstraction between photon count and cell function.
So run your beads and standardize your data acquisition, by all means, but don't ignore the patterns and natural variation. The anti-data of the experiment. It may be noisy and discordant, but you have to learn how to listen.
If you want my claim of Quality Assurance, the Adam Harris Trio will be performing at CYTO 2010 in Seattle.
Download Link "Fantasy In D" - Adam Harris Trio
-Adam Treister
|