Random Ideas \ Thoughts \ err \ posts!

Live life to the fullest!

Bad Science?! August 13, 2008

Filed under: Everything — Jason Olson @ 10:45 am
Tags: , , , ,

I recently heard on NPR radio that there is a growing problem with the way studies are conducted on TV viewership of certain programs. The problem is the research firm is still only surveying about 1,000 people nationwide for what programs they are watching. That is significantly too small a sample size, especially as we evaluate that there is nearly 500 different TV channels available today. Because of this, some regional favorite ‘prime time’ network shows have been canceled, because while they were extremely popular in a particular market, they were not within the very small sample size of 1,000 users. This returned me to my thoughts on Carbon Dating…

While I do not have a degree in science, of any kind, there is something fundamentally flawed in the way we use carbon dating to ‘date’ things millions of years old. The basis for carbon dating is that we know the rate of decay. However, it is impossible to assume the rate of decay over a period of time hundreds of thousands of times longer then we’ve been able to observe that decay. We have only been able to reliably observe carbon decay for a few decades, and can infer over a couple of hundred years, but beyond that, it is simply a guess that it is a predictable rate. However, this would be similar to measuring the height of a person over a period of days, or perhaps hours and using that to infer the rate of growth of a person. No mater how constant you observe that rate over a set of people in that same age range. We know that kids go through growth spurts. So if we evaluated a child between 23 and 24 months, when we look at someone who is 80 years old, but we’re using the data collected from months 23 to 24 months, we would ‘date’ that person to be 14 years old or so. Moreover, go the other way around, evaluate the growth of a 40 to 41 year old, the growth (or reverse growth) is taking place at an almost immeasurable rate — if we were to simply use that sample, how old would we ‘date’ that person to be?

Absurd? Yes! So how could we use data from such a small period of time, to infer out to millions of years? If we cannot reliable observe, test and repeat this, over a meaningful sample size, how can we accurately use this information? This is simply bad science!