In particular Young-earth creationists have 3 main criticisms they frequently cite about radiometric dating, these are as follows.
They claim that scientists ignorantly assume that:
- The initial conditions of the rock sample are accurately known.
- The amount of parent or daughter elements in a sample has not been altered by processes other than radioactive decay.
- The decay rate (or half-life) of the parent isotope has remained constant since the rock was formed.
These assumptions are known by geologists, and are taken into account, and I'll go through the refutations to each of these criticisms.
Isochron dating methods do not assume that the initial conditions are known. This assumption is only made in very simple radiometric dating, the kind that you'd use only if you did know the initial ratio of the parent/daughter minerals, such as with radiocarbon dating. To explain Isochron Dating I'm going to give you a paragraph from TalkOrigins that explains it very well. Though understanding this concept requires a basic understanding of statistics and also of radioactive decay.
With isochron dating, we also measure a different isotope of the same element as the daughter (call it D2), and we take measurements of several different minerals that formed at the same time from the same pool of materials. Instead of assuming a known amount of daughter isotope, we only assume that D/D2 is initially the same in all of the samples. Plotting P/D2 on the x axis and D/D2 on the y axis for several different samples gives a line that is initially horizontal. Over time, as P decays to D, the line remains straight, but its slope increases. The age of the sample can be calculated from the slope, and the initial concentration of the daughter element D is given by where the line meets the y axis. If D/D2 is not initially the same in all samples, the data points tend to scatter on the isochron diagram, rather than falling on a straight line.Other types of radiometric dating do make assumptions about the initial levels of parent/daughter minerals but these assumptions are justified. For example C-14 dating assumes we know the level of C-14 in the atmosphere when the organism died, and this does vary quite substantially. Lucky for us, for dating samples within the range of C-14 dating, we can use other methods to calibrate the accuracy of it, for example ice-core samples and tree-ring dating (dendrochronology).
The second criticism that scientists assume that a rock is a closed system is almost laughable. Absolutely closed systems do not exist even in ideal conditions, but many types of igneous rock come about as close to a closed system as you can possibly imagine. As was mentioned in the paragraph from TalkOrigins, multiple samples are taken in any dating procedure, and the fact that they consistently produce results within one percent of each other is testament to the closed nature of most igneous rocks. Even if a particular lava flow was contaminated, the chance that all samples were contaminated equally, so as to give results within 1 and 3% of each other is extremely low.
Not only are most rocks barely capable of being contaminated, even if they were, isochron dating methods are capable of detecting contamination and even correcting for it. Geochronologists are well aware of potential rock contamination and take precautions to avoid samples that may have been contaminated, for example they will not use a sample that has been weathered.
The third criticism is just stupid. All scientific research that has been done on this idea has conclusively come to show that radioactive decay is a very consistent process, and does not fluctuate under different conditions. Some creationists claim that cosmic rays or neutrinos would affect the half-lives of minerals, but there has been no evidence to show this is the case. This is simply a bald assertion from the Creationists, showing once again that they are not interested in doing real science, but only to propagate falsehoods and lies.