Quote:
Originally Posted by Cannon Shell
The sample size is not big enough to make good determinations about the individual horses themselves let alone any sweeping generalizations about bleeding or lasix. You could scope 55 different horses and get the exact opposite results.
This is stuff that any statistics class goes over right after reading the syllabus.
|
We were told 90+% of horses bleed (I don't remember the exact number). I've taken enough statistics classes to know the results in the article would be enough to think that number is totally bogus. 15 out of 41 bled, not near the high percentage that was quoted. That would be statistically significant given the base of 90+% that has been cited. In other words, if horses really bled at such a high rate, the odds of finding a sample of 41 where only 36% match the criteria are basically nil.
And 10 out of 14 WITH Lasix bled? Yeah, that doesn't sound too effective to me. And we are talking the best (and best cared for) horses.