miranadobe
Posted : 8/15/2007 11:59:57 AM
When you make a claim about data, such as you have, you have to have some way of proving that you are right and the data is wrong.
Bob, there's a difference between "wrong", and "incomplete/inconsistent"... or "limited". By data provided in another post/thread, this person reports the following "data" about his experiences with small dogs:
1- pleasant non-terrier
2 - pleasant terrier
4- unpleasant non-terrier
1 - unpleasant terrier
2 - unpleasant terrier-types - one is a terrier based dog, the other a terrier mix
1 - unpleasant unknown breed
All things added, that data equates to 3 out of 4 small terriers/terrier-types he knows are unpleasant = 75%. 5 out of 6 other/unknown small dogs are unpleasant = 83%. By his reported data, the problem is statistically with Chihuahuas, because two of the 4 unpleasant small dogs he's encountered are Chihuahuas. Get another 5 people from his neighborhood giving realtively the same assessments, and, by the application of arguments, Chihuahuas should be banned from his neighborhood. (and, serious limits on viability of the data b/c, this doesn't account for measuring "pleasant versus unpleasant" equally, nor the dogs in the neighborhood that nobody knows about, hiding away unexposed to neighbors in a negative or positive way.)
Sorry, it was just too much fun to use this data to exemplify my point.[
]