What solutions are available for increasing accuracy of the tests? From the source linked above: Carbon-14 is considered to be a highly reliable dating technique.It's accuracy has been verified by using C-14 to date artifacts whose age is known historically.In this answer, I will try to go through this story in great detail, (hopefully) exposing the reasons why this work is not taken seriously by scientists.A research team from the CRSEF, or Creation Research, Science Education Foundation, led by Hugh Miller, has claimed to have dated dinosaur bones using radiocarbon methods, determining them to be no older than several dozens of thousands of years old.Let's look at their research methodology in detail (indicated by bullet points): Let's take a little pause to consider the general issue of misrepresenting your own research. did this, since there would have been a slim chance (at best) of the museum curator providing them with any dinosaur bone fragments if he or she had known what the true intent of the supposed chemists was.In particular, it is implausible that it would have been considered worthwhile to try to use radiocarbon dating methods on these bones, since the rocks that they were taken from were determined to be 99 million years old, as shown in this paper by Kowallis et al."Comparing such different molecules as minerals and organics from the same bone region, we obtained concordant C-14 results which were well below the upper limits of C-14 dating.These, together with many other remarkable concordances between samples from different fossils, geographic regions and stratigraphic positions make random contamination as origin of the C-14 unlikely".
This, of course, raises some ethical questions, but let's brush these aside for now.We proceed with the examination of the research done by Miller and his fellow researchers from the CRSEF.At a horizon of 40,000 years the amount of carbon 14 in a bone or a piece of charcoal can be truly minute: such a specimen may contain only a few thousand 14C atoms.Consequently equally small quantities of modern carbon can severely skew the measurements.Contamination of this kind amounting to 1 percent of the carbon in a sample 25,000 years old would make it appear to be about 1,500 years younger than its actual age.Robert Kalin senior research specialist at the University of Arizona’s radiocarbon dating laboratory, performed a standard independent analysis of the specimens submitted by Hugh Miller and concluded that the samples identified as “bones” did not contain any collagen. These results corroborated established paleontological theories that assert that these fossiles presumably were 'washed away' over long periods of time by ground water, replacing the original bones with other substances such as the minerals naturally present in the water, implying that this sample could ).