Posted on 12/31/2001 9:30:22 PM PST by jennyp
It should. I said the same thing in post #8 on this thread.
Let's get the order straight:
The "salting" on scratching posts was done in the 1998 'study'.
The 'control samples' were sent later, to cover their a$$ after they got wind down the grapevine that they had been snitched off by some of their cohorts.
Right. Exactly my point.
Thanks for coming out for all here to see.
You just have no intellectual honesty, do you?
Let's get the order straight:The "salting" on scratching posts was done in the 1998 'study'.
The 'control samples' were sent later, to cover their a$$ after they got wind down the grapevine that they had been snitched off by some of their cohorts.
Where do you get that from? None of the articles say that, do they? The 1998 study was discredited because the lab screwed up. Nobody has claimed fur was planted in the 1998 study. If you think otherwise, then please point me to the article that says so. The hairs were sent in in 1999 & 2000.
No, I understand your point that these guys were recklessly acting on their own & not following a good protocol in any sense. I was referring to where NMGA pointed out that if their story was true, they should've been salting the data with bobcat hair instead of lynx hair, to smoke out false positives. Salting the data with lynx hair, as they did, would only smoke out false negatives - which contradicts their story.No More Gore Anymore pointed out that if they were suspicious of false positives coming from the lab, they should've planted a bobcat sample to see if the lab would score that as a lynx. That makes perfect sense!
It should. I said the same thing in post #8 on this thread.
Efforts to manage wildlife under the already controversial Endangered Species Act have been hurt by the seven federal and state scientists who sent bogus lynx-hair samples to a lab.
The incident wounded the scientific credibility of state and federal agencies that turn to science to defend politically unpopular decisions, from turning off irrigation ditches for salmon to silencing chain saws for spotted owls.
It sure does. IMO, this faltering or fraud (call it what you will) is just one more reason why the government should own as little land as possible. Let private sector business utilize it. How about a network of FreeRepublic Patriot's Parks all across the country.
If a person or group of people want to protect a species let them buy the land and preserve it for that purpose.
You are right, the articles that you have inserted do not say that, but it is the only logical conclusion that is supported by all of the facts available.
When a jury deliberates, it is usually what pertinant facts they were NOT told that they have to figure out, is it not?
Jenny, what one does in a single-blind test is submit test samples to the lab from two populations, known and unknown. In this case they would have been captive samples and field samples. Whether the known samples were lynx or bobcat is immaterial to the point.
Immaterial to whether they acted improperly, I agree! Look, in light of your background explanation of what a proper procedure would entail, it's obvious they acted improperly. At this point I'm trying to figure out if they were merely lazy, arrogant, & incompetent (what they're essentially arguing) or if they indeed were trying to fraudulently plant lynx into the actual data.
Whether it was lynx or bobcat should be material to that second question, because it reveals whether their stated rationale is internally consistent.
If you suspected false positive hits for lynx & wanted to test the lab for it during the next survey (assuming you were arrogant/lazy enough not to care about doing it right as part of a standard protocol), which kind of hair would you insert - lynx or bobcat? It seems obvious that it would be bobcat, to confirm your suspicions that the lab was misreading it as lynx. But they didn't do that, they did the opposite: They inserted lynx hairs. I'm just saying their specific actions contradict their cover story.
Which is why I'm starting to agree with you that it must be a coverup on some level. Don't you see that?
LOL, yes, I do see that you are coming to some conclusions but one has to be certain ;-) All experimental designs are subject to several factors. The sampling requirements for a false positive or false negative are different. Large numbers of trials necessary for statistical certainty can get expensive. One seldom gets many samples of an endangered species. This means that one must have enough confidence in the lab to regard their conclusions to be totally reliable. That means that one should qualify the lab in advance of the study.
If this were a lab qual, I would first make a call on the lab to see if they have validated their procedures by indepedent third-party audit. I would cross their process documentation with the handling of a sample that was already in the lab. I would look at calibration data, request their SPC charts, check the training logs on the technicians, and go through their audit file. If they've got all that, it looks clean, it cross-checks when I call the auditor, and I know their reputation or have experience with them, I'm done.
Lacking any of that, I might perform a set of extreme vertex screening experiments to see if I wanted to go into further tests. I would ask them what the minimum mass of the sample must be. Some of the submissons might be marginally below that (to see if they react to their own specs). I might take the two fur samples and break them into four tests: lynx, bobcat, both, and a second bobcat (assuming that lynx samples were harder to get). I would NOT tell them what I was sending. The vials would be merely numbered. (Remember, you are sampling DNA, the lab tech is unlikely to recognize the fur.) This test verifies that the lab is capable of identifying both species and individuals, and that they will react to out-of-range data (the mixed vial) with a response of, "invalid sample, but it could be this." If I get all that, I would repeat the test with but two of the samples from one of the same individuals as before. That way I get data from different sample preparation procedures and perhaps different technicians and I get to see how well the equipment repeats in separate trials.
Screening experiments are an artform because few can afford the number of samples required for statistical certainty. The setup I just described is a pretty rugged indicator of the lab's performance.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.