“Quality is value to some person”
I’ve long hated the sentence in the title, for a number of reasons. I threw away a long blog post about that yesterday, but let me just issue this challenge. How do the meanings of these three sentences differ?
Quality is value to some person.
Quality is quality to some person.
Value is value to some person.
July 12th, 2010 at 11:09 am
1. We define quality in terms of value to an individual.
2. We define quality subjectively. Or recursively.
3. When we talk about value, we should always do so subjectively.
I think the three are quite different. The importance of the first, if it has any, certainly includes the subjective character of the others, but to me it also makes it a bit harder to justify gold plating. That might just be because I’ve thought about it.
July 12th, 2010 at 11:15 am
I first came across the phrase in question as a way to shake people out of their complacency that they understood quality as some narrowly defined thing; the kind of thing that QA people /assured/.
Lo! These many years later it has just become some other thing to become complacent about.
July 12th, 2010 at 2:52 pm
I could provide alternatives to Ron’s parsings, but I can’t improve upon them.
2 and 3 are recursive and constraining; even if they’re not persuasive, at least each is modest.
1 relocates the problem of understanding quality to the problem of understanding subjective value. This does little to clarify things for me; quite the contrary, actually.
That’s a criticism of 1’s form; I also dislike its content. There are useful aspects of “quality” beyond what it suggests. Many things in complex systems are best understood using aggregate measures. When the pollen count goes up–a measurement of “air quality”, notably–it’s impossible to detect, on an individual basis, what difference that makes. Twice as many kids might come to the ER with asthma attacks, per unit time, but even if you work from that highly selected sample, you can’t examine any given kid and say, “Oh, she’s here because her pollen threshold was 89, and today we hit 90.”
Thus, I’d submit that 1 is simultaneously too vague (what is value?) and too narrow (some person? which one?!), which is a rather neat trick.
July 14th, 2010 at 7:34 am
“I know it when I see it?”
I like Mark’s idea of considering aggregates. It seems that we, as a species, have an intuitive understanding of “goodness” without being able to specify it in the particular.
Perhaps it works better in the negative:
-) a software defect is an attribute (of the software) which displeases a person
It displeases someone - a specific person - who may or may not be “correct” in their opinion. If you only have one customer, then their opinion carries a good deal of weight. If you aggregate all of the opinions across the class of customers, then you could end up with an aggregate definition of quality.
-) Quality represents a lower number of attributes that displease some person