Can different radioactive dating method produce vastly differing results on the same item? After being questioned by a creationist, some of their given evidence seemed to be more poignant to me than it had previously done. notwithstanding carbon dating methods which are highly prone to contamination, the presupposition that other forms of radioactive dating should produce a roughly similar estimate should stand. Are there any examples of such a test being done on some elements that have been carbon dated in the past?
Obviously, the only examples I got were from creationists who run tests on the grand canyon lava flow layers and all the tests provided vastly varying age estimates.
Results:
Method: Potassium-Argon = Isochron age of 516 million years
Method: Rubidium-Strontium = Isochron age of 1,111 million years
Method: Samarium-Neodymium = Isochron age of 1,588 million years
Is this to be expected or can it simply be chucked to the test being irregular or contaminated if they were all testing the same samples? If so, what would be the explanation for this? They argued that without adjusting the decay rate of one method to match another, most methods produce various Isochron ages.
I was also stumped by the example of MT ST Helens. As the mountain erupted in 1986, crystalised rocks were only 10 years old in 1996. Apparently, in 1996 some samples were tested using radioactive methods and the results were as follows:
Using the K-Ar method
Whole rock = 0.35 million years
feldspar = 0.34 million years
pyroxene = 1.7 million years
This was due to the fact that the rocks inherited excess daughter Aragon and according to them this proves the flaws in radioactive dating methods. Could this also be a result of improper testing methods or is there a more accurate explanation of the inconsistencies using more scientific methods or theories?
I hope to acquire more knowledge on the subject should I find myself in the middle of these kinds of debates again.
Thank you.
 A: Something important to highlight is: different methods only work in predefined sets of materials. Take uranium-lead, for instance: it is usually done on $\mathrm{ZrSiO_4}$. This incorporates uranium onto its structure as a substitute to zircon. Therefore you could not use this dating method in such a material that does not incorporate uranium at known rates - otherwise, uranium would only be a contaminant varying in its concentration according to the surrounding geographic formation.
I don't know if this is actually helpful, but I cannot comment on posts yet, as I don't have enough "reputation"; so I'm posting this as an answer. But there you go: I hope it helps.
A: I'm not a specialist in this - you definitely need to find some geologists to ask for details - but I can maybe offer some general comments.
First, on the question of what is going on with the samples. I had a quick look around and found some Creationist stuff that showed data plots that strongly suggest the result is genuine. I couldn't trace it back to a published paper, and it's not impossible it could be faked, but I don't think it is. This sort of confusing result is common enough in real-world science that I don't think they'd need to.
The fit to the straight isocron line looked tight, so it's probably not sample contamination or measurement error, unless it is very systematic. (I've heard that you can get different isotopes diffusing at different rates through groundwater-filled cracks that can result in a systematic real-looking error. I don't know enough to say if it's plausible in this case.)
I think it's a lot more likely that you're getting rocks of genuinely different ages mixed up in the same lava flow. Think about sticky viscous lava flowing through a tube in rock. It's going to heat and melt the walls, right? And drag bits off and into the flow. It might have drilled through lots of layers of different ages. The assumption is that when it remelts it gets thoroughly mixed and homogenised, but how likely do you think that is in reality?
If you got some white cake mix and dark cake mix, chucked them together in the same bowl, and then squeezed it out into a cake tin, you wouldn't get a perfect uniform mixture of some intermediate colour. No, you would have separate strands and blobs of the different colours folded into each other, like marble cake. It would take a lot of mixing to thoroughly homogenise them. If the rocks looked the same, you might not notice.
Real-world processes are often messy, compared to textbook assumptions. You always have to think about all the many things that can go wrong.
Secondly, this is also a question about science and scientists. We are all human. We make mistakes. We misunderstand. Theories that have stood for centuries can get overturned. Measurement methods that we have relied on for decades can prove to be entirely bogus and biased. These things happen. We are all fallible, and the Scientific Method is designed specifically to deal with that fact.
We humans are biased reasoners, unreliable logicians, full of cognitive blindspots, and a tendency to get fixated on favourite theories and then defend them to the death. How can you generate reliable conclusions using an unreliable and error-prone reasoning process? The answer is that you get lots of reasoners with different biases, cognitive blind spots, and favourite theories, and get them to pick holes in one another's claims. Theories gain scientific credibility only by surviving such a process of highly-motivated, competent sceptical challenge.
However, the real world is often messier than the theory. I suspect in this case that geologists are well-aware of the messiness of real geology, and the limitations of such methods, and the problem is simply that in talking to the public they are just sticking to the simple story of "These rocks are a billion year old!" and don't expand on the caveats, assumptions, and uncertainties involved. But sometimes it can happen that there isn't enough sceptical challenge and a mistake has got through. Personal pride, careers and reputations, the need to pay the bills or get grants, or politics can get in the way of scientific openness to examination and correction. They can get too confident in the accuracy of their methods, and then when attacked 'politically' by outsiders like Creationists get over-defensive. There is an unscientific but all-too-human urge not to engage so as 'not to give ammunition' to the 'enemy'.
The Scientific Method thrives on being attacked like this, but politics, the public understanding of science, and the need to pay the bills often take a different view.
So it is just as important when dealing with Creationist arguments to respond in the right way with regard to the Scientific Method, and not just have it be about winning the argument and proving them wrong on technical grounds. Because the biggest problem with their position is not that they have misunderstood some abstruse point about radioisotope dating of rocks, but that they don't apply the Scientific Method to their own beliefs. (Not that it's compulsory, of course. But it's humanity's most effective route towards truth.) Teaching them to be able to say things like "We don't know" and "We were wrong" and "We are fallible" is hard, especially when identities and worldviews are at risk. To some degree, you can only teach that by example. Do unto others as you would have them do unto you. Be humble, and tolerant of other opinions. Listen. Ask questions. Support free speech and freedom of belief. And don't try to shut them up or shut them out, because that itself is totally antithetical to the Scientific Method.
Like I said, I suspect it's just poor homogenisation of rocks of different ages and geologists know all about it, but be open to the possibility that it's genuinely a mistake. And the right response to being mistaken is to acknowledge and correct it, and move on.
