Zeno's Paradox
While critical thinking may not make up for a lack of knowledge, it is essential for gaining knowledge.
|
|
Thursday, September 26, 2002:
Critical Thinking, Science:High-flying physicist sacked for falsifying data (NewScientist.com)A high-flying young physicist (Hendrik Schön), who rapidly published a series of remarkable discoveries, has been sacked by Bell Labs, New Jersey, for "scientific misconduct". Scientist fired for fabricating data (The Skeptic's Dictionary)The scandal also illustrates how much faith scientists have in each other and how much faith journals such as Nature have in the integrity of scientists. Dr. Donald Kennedy, editor-in-chief of Science, noted that detecting fraud has never been an objective of the peer-review process. The scandal also illustrates the fact that science is ultimately a self-correcting enterprise, and that errors--whether due to fraud, mistake, or incompetence-will eventually be found out. Science works, though perhaps not at the speed that some would like.
Related Information:
How a Scientist Thinks by Prof. Ian Parberry.
The Undergrowth of Science: Delusion, Self-Deception and Human Frailty by Walter Gratzer (ISBN: 0198507070)
-
Ron
9:16 AM
Wednesday, September 25, 2002:
Human Factors, Critical Thinking:Fix an Incomprehensible User Interface?
I found Scott Adams' Dilbert comic of May 11, 2002 to be particularly interesting:Dilbert: [pointing to a picture of money going into a computer] We can fix our incomprehensible user interface for a million dollars, [pointing to a picture of a person with his eyes tightly closed] or we can close our eyes and wish real hard that our users won't care."
Wally: [sitting next to his manager, whose eyes are tightly closed] "He's saving a million dollars. What did you do today?" Funny as it may be, the manager is probably correct in not choosing the first option...
Consider the following analysis of the comic:
1) There is a serious problem: "incomprehensible user interface"
2) There are two options on how to address it:
2a) It can be fixed at a certain cost, "a million dollars"
2b) or we can just ignore it "and hope that our users won't care"
3) Management does not choose option 2a, thus "saving a million dollars"
From Dilbert's perspective, there is only one reasonable solution, to solve the problem. The reality of such situations is quite different and far more interesting:
1) There is a serious problem: "incomprehensible user interface"
What does it mean that they have an "incomprehensible user interface"? It sounds like a scare tactic. Who identified the problem? How? What evidence exists to support that there is indeed a problem and that the problem is actually serious?
If the people who identified the problem are the very same people who will be rewarded for fixing the problem (monetarily or otherwise), they have definite bias and conflict of interest in identifying "serious problems" of any kind. How does their credibility and expertise in identifying such problems weigh against their bias?
Does the problem actually exist? While usability testing is absurdly unreliable at identifying individual problems, a good usability test conducted by qualified evaluators should be able to identify if problems exist at all. And of course, sometimes it is painfully obvious that there are problems...
Is the problem serious? Unfortunately, identifying the seriousness of a problem through usability testing is even less reliable than identifying a problem. If it isn't obviously serious, then beware of self-interest motivating the identification of "serious" problems.
2a) It can be fixed at a certain cost, "a million dollars"
Is the problem's cause obvious and easily fixed? Fixing usability problems is much more difficult than identifying them, especially without introducing new problems. Of course, if you erroneously identified problems that don't actually exist, or mis-identified minor problems as serious, then attempts to "fix" them will likely just introduce real problems. Likewise, if the problem’s cause was not correctly identified, then the product may be made worse at the cost of a million dollars (assuming that the cost-estimate was reliable and the fixes can be done within budget).
What does the million-dollar fix entail? To be successful it should include redesign, testing, development, more testing, documentation, still more testing, production, more tests still, delivery, follow-up on delivery, etc. while at the same time taking resources away from whatever was planned prior. Were the complete costs included?
2b) Ignore it "and hope that our users won't care"
Of course, ignoring or denying potential problems is always risky, usually irresponsible.
2c) There are always other options. The better ones will include assessing the problem, finding alternative solutions, and then choosing an appropriate one. The best ones will weigh the costs against the benefits to both the customers and the business. (Obviously, I don’t want to go into detail. Perhaps a future weblog entry…)
3) Management does not choose option 2a, thus "saving a million dollars"
Solve a problem for a million dollars? Sounds like a con...
Summary thoughts:
- There may be a problem and it may be a serious one. Don't dismiss or deny a potential problem.
- It is unlikely that the problem and its severity have been correctly identified unless they are obvious.
- It is also unlikely that the costs of fixing the problem have been reasonably assessed, unless the team already has considerable experience fixing such problems and has thoroughly assessed previous costs.
- It is extremely unlikely that the problem can be fixed without introducing more problems.
- It is highly likely that other other measures can be used to minimize the effects of the problems. (Yes, a future weblog entry…)
- What caused the problem in the first place?
-
Ron
6:52 PM
|
|
Copyright © 2002-2005 Ron Zeno
|
|
|
|
Musings not completely unrelated to human factors, management, critical thinking, medicine, software engineering, science, or the like.
|
|
|
|
|
|
|
|
|
|
|