QUOTE(Judaculla @ Feb 15 2004, 08:30 AM)
It's a reliability check.
The frequency of sightings by state in the Green data line up very well with the frequency of data in the BFRO database, even after you account for the difference in total number of sightings. If the databases are largely separate (and it looks like they are with a handful of exceptions), that tells us whatever is being measured by one is also being measured by the other.
Reliability isn't the same as validity, so please no one get in a huff about that. Both terms have narrower technical meanings in natural and social sciences. Validity requires reliability, but reliability doesn't require validity.
Here's are some undergraduate class notes from a social scientist Ph.D. at Arizona State University explaining reliability, validity, and correlation. If you want a dry read, knock yourself out.
Reliability, validity, and correlation Ok....this strikes me suddenly as a rather important, and possilbly overlooked, aspect of study...(overlooked at least, in that I have not seen anything along this line really discussed before).
Statistical analysys using seperate data streams to test for data reliablility..something to chew on for sure!
Thank you Judaculla...I think you are on to something here.