safety inspector Thomas Ronald Baron Dec 8, 2009 20:59:10 GMT -4
Post by JayUtah on Dec 8, 2009 20:59:10 GMT -4
You need to test your hypothesis and see if it is actually possible. If not, then you need to discard it as falsified.
That's why I recommended the bench tests. We do this all the time for hydraulic systems, but the ones I typically do are very expensive because they involve high-end actuators and controllers. Nevertheless you can get a master cylinder, appropriate slave cylinders, and authentic tubing for under $200. Have a Volvo mechanic hook it up, certify it up to spec, then "fail" it in as many ways as you can. Suck all but 30 mL of fluid out of the reservoir and see if you can still exert braking force. Test different leak rates and see where it becomes noticeable to a driver. Drill holes in the metal, abrade the rubber tubes, and open the bleed valves to see what effect that has on braking. I can only do so much with crude drawings and data sheets, and a little math. I can provide first-order approximations and suggest what to have Gary test with a mechanic.
...apply Occam's Razor, the simplest hypothesis (i.e. the one with the least major assumptions) is likely the correct one
"Simple" in the context of Occam means the one with the fewest loose ends. The simplest explanation for the half-inch loss of fluid is simply ordinary leakage. Forcing a leak in one or more circuits would also produce that result, but to test whether that happened to Baron we'd need information we don't have from the original investigation. We can't test that hypothesis, therefore it's not a "simple" one. We don't need to test the leakage hypothesis because we know a priori that it happens.
...investigative journalism tends towards the spectacular because it makes a far better story.
I've observed that, but it's unfair to Gary to impugn his entire profession. We can discuss Gary's methods without wondering whether they are followed by others.
Without proof, hearsay is worthless as evidence because it means nothing.
Not exactly. Hearsay evidence cannot be tested, therefore it has diminished value. If we ask Mogilevsky, "Describe the car that was tailing Baron," his answer would be, "I don't know that; I only know what Baron told me." That limitation prevents hearsay evidence from being accepted in rigorous contexts.
The only value hearsay has is as a marker to help an investigator know where to look in order to locate further possible evidence.
And that's exactly how Baron was treated before Teague's subcommittee. He was asked about observations he himself could make, and he was invited to submit the names of others who could testify in place of Baron's hearsay.
So to his mind, and memory since he did note that he'd have to re-read his report, the biggest issue was a lack of communications. Not really that damaging is it?
Yes and no. Baron was right on that point, and it doesn't take a rocket scientist to know he's right. Any worker will throw up his hands in frustration when he realizes he wasn't told an important piece of information. Communication problems plague all industries at all times, regardless of the pace at which they operate.
Jay has never stated that Baron wasn’t intelligent, but rather that there is not enough evidence to determine how intelligent he was.
Right. To distill the conversation:
Gary: Baron was intelligent, therefore his lengthy report was likely argued well enough to be valuable.
Jay: Do you have any objective measurements of Baron's intelligence?
Jay: Then we can't accept the viability of the report on that basis.
Baron may have been intelligent, but we don't know how intelligent he is. Baron's report may have been well-argued, but not for that reason.
This is how refutations works. If someone makes an assertive claim, we test how well that claim is supported. If the claim fails for lack of support. we don't assert the counterclaim; we simply fail to assert the proposition attached to the claim. The needle simply goes back to center, for lack of a better analogy.
If someone tells me, "It's raining outside," I may ask, "Can you actually see out the window?" If his answer is no, I know that the argument fails to support the claim. But that doesn't mean it's not raining. It only means that that particular argument fails to establish whether it is or not. It may be raining, but additional observation must occur to determine that.
Instead Jay noted that Baron was inexperienced and well out of his depth in the field that he was attempting to work in.
Yes. Baron was right to observe that his organization did not communicate effectively. It is reasonable to suggest that all communication be in writing, for example. However, such a thing is not practical or prudent. When Baron makes those kinds of idealistic suggestions, he belies his inexperience. A more experienced operative would respond, "Yes, that would solve some problems but it raises others that impact the work in other ways."
...very intelligent people can have problems trying to deal with things outside their experience...
Indeed. I have no idea what the frak a "kilt pin" is.
Baron's lack of experience in large-scale organizational management, engineering development, and engineering methodology show in his writings and testimony. They define his expectations, and his expectations in turn drive his judgment of what he observes.
When we have that evidence, it is not mere assumption to propose that those same conditions persist. If Gary proposes that Baron's expectations may have improved from one report to the next, then that incurs a burden of proof. When you have an established context, proposing to change aspects of that context requires proof that the context actually changed, or was likely to.
If you want to argue that Baron's approach and expectations changed between the short report and the long report, you have to point to events that occurred in his life that drove it -- classes, conversations with experts, etc. But Baron didn't seem motivated to change his approach. He gave his report to NASA, which immediately arranged for Baron to discuss his findings with NAA. But Baron wasn't satisfied with NAA's response and went then to the media. Baron's behavior is not consistent with someone who is evaluating the validity of his expectations.
Evidence that change occurred is first-class evidence. I don't see any of it. Evidence that a change was motivated would be second-class evidence. I don't see any of that either. Failing those, it is entirely plausible to propose that the same context in which Baron wrote his short report is the context in which he wrote his longer report. And that context is what's broken. The expectations against which Baron evaluates his former employer's behavior are not well founded.