# How can I statistically evaluate different measure methodologies?

I have a set of data with the measures of Constitutional length. The data is organized by state, so each row has the information of constitutional size for the referred state. I have the data for 26 states.

Now, there are three units, or methodologies, or types, of accessing constitutional size. I can measure it's length by:

1. Number of articles (varies around 300 articles);
2. Number of words (varies around 32000 words)
3. Number of provisions (around 1300)

So what I have is three different methodologies to assess constitutional size, and what I'm after is if there is any statistical way of doing that. Naturally they are very different variables, in different magnitudes, although the idea is that they all measure constitutional size. So it's expected that they perform similarly.

Summing up, I'd like to know if there's any statistical method to evaluate and compare the three measuring methodologies.

However you can compare these methods and see how well these measures are related: for example you could see that the number of words is roughly 100 times the number of articles and that the number of provisions is roughly 4 times the number of articles. You could study these relations further by computing mean and variance of: (number of words / number of articles) and (number of provisions / number of articles).