Authored by: Anonymous on Wednesday, July 18 2012 @ 02:31 PM EDT |
"Congress, I've just learned, is having hearings on patents this week, and
as it happens Professor Chien is going to testify, so if you have information
you'd like her to take with her, so to speak, this is your moment. The purpose
of the survey, in other words, is to help frame policy recommendations, so it's
important, if you care about patents, and I know a lot of you do. Just be sure
to be accurate and precise. "
Or at least that's what PJ's post implies.[ Reply to This | Parent | # ]
|
|
Authored by: Wol on Thursday, July 19 2012 @ 08:47 AM EDT |
And even if it IS a biased sample, the absolute numbers themselves can be very
illuminating.
Let's say you have a town of 1000 people, and you run some sort of survey. 50
people are motivated to respond, and they all tell the same story. Okay, you
have a self-selected, biased survey. But taking other information into account,
you can say "5% of the people felt strongly enough to respond". That
5% is a pretty massive figure in that context ...
Absolute numbers can shout louder than percents ...
Cheers,
Wol[ Reply to This | Parent | # ]
|
|
|
|