Comments on: Veterinarians stepping up for feral cats http://www.petconnection.com/blog/2007/11/11/veterinarians-stepping-up-for-feral-cats/ The Web blog of the Pet Connection, a pet-care feature syndicated internationally by Universal Press. Tue, 06 Jan 2009 08:37:30 +0000 http://wordpress.org/?v=2.6.3 By: Gina Spadafori http://www.petconnection.com/blog/2007/11/11/veterinarians-stepping-up-for-feral-cats/#comment-154127 Gina Spadafori Sun, 11 Nov 2007 16:15:48 +0000 http://www.petconnection.com/blog/2007/11/11/veterinarians-stepping-up-for-feral-cats/#comment-154127 Lies, damn lies and statistics ... :) Lies, damn lies and statistics … :)

]]> By: The OTHER Pat http://www.petconnection.com/blog/2007/11/11/veterinarians-stepping-up-for-feral-cats/#comment-154028 The OTHER Pat Sun, 11 Nov 2007 14:01:17 +0000 http://www.petconnection.com/blog/2007/11/11/veterinarians-stepping-up-for-feral-cats/#comment-154028 I hate it when people use "percent" in describing the results of small surveys. At least they use the actual numbers to begin with ("73 respondents", "Eight respondents") but they then jump to using percentages. which makes the numbers "feel" more significant than they are. Rework: 50 percent indicated that they never had the opportunity 25 percent indicated that they did not support TNR 25 percent did not give a reason for not helping into: 4 respondents indicated that they never had the opportunity 2 respondents indicated that they did not support TNR 2 respondents did not give a reason for not helping and it "feels" a lot different, doesn't it? I don't recall the guidelines for how large a sample has to be before you can legitimately report "statistical findings" as "percentages", but I'm pretty sure it's in the hundreds at least - not 81. Imagine I were to ask two people whether they liked peanut butter, and one of them said that no, he hated it. And then imagine I went on to report that "In a recent survey, 50% of respondents said the smell of peanut butter makes them nauseous." It's just not quite forthcoming to use "percentage" results when reporting on a very small number of responses. I hate it when people use “percent” in describing the results of small surveys. At least they use the actual numbers to begin with (“73 respondents”, “Eight respondents”) but they then jump to using percentages. which makes the numbers “feel” more significant than they are. Rework:

50 percent indicated that they never had the opportunity
25 percent indicated that they did not support TNR
25 percent did not give a reason for not helping

into:

4 respondents indicated that they never had the opportunity
2 respondents indicated that they did not support TNR
2 respondents did not give a reason for not helping

and it “feels” a lot different, doesn’t it?

I don’t recall the guidelines for how large a sample has to be before you can legitimately report “statistical findings” as “percentages”, but I’m pretty sure it’s in the hundreds at least - not 81.

Imagine I were to ask two people whether they liked peanut butter, and one of them said that no, he hated it. And then imagine I went on to report that “In a recent survey, 50% of respondents said the smell of peanut butter makes them nauseous.”

It’s just not quite forthcoming to use “percentage” results when reporting on a very small number of responses.

]]>