h a l f b a k e r yOn the one hand, true. On the other hand, bollocks.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
This morning on the news I heard the phrase, "Doctors are concerned by the rate at which this is spreading." It made me wonder whether the reporter had actually performed a poll or just asked his mate who once went out with a nurse.
If details of professionals with different fields of expertise,
sportsmen and members of various associations were gathered then these people could be asked their opinions on various matters of the day. Media companies could pay for statistics and get them first and others could wait a couple of days and look at the free on the web.
The contributors could be encouraged to answer questions by charitable donations from the polling company made according to the number of responses received.
[was 'Job Specific Internet Poll'. Changed to 'Specialist Internet Poll' as I though this decribed the idea better]
Half-Baked
http://www.halfbake...0Current_20Thinking
[lubbit, Jan 23 2002, last modified Oct 17 2004]
[link]
|
| |
Obviously this would work better in the "30% of doctors feel that..." case rather than the "Drug dealers in the Brixton area are concerned that..." or the "The avarage age of someone who commits benefit fraud is..." cases.
A good idea though. It would have the flaw that the poll of doctors or whoever would be self-selecting, but this would be no worse a method of selection than a straw poll. |
|
| |
In order to validate the polls, would there have to be at least 100 doctors, 100 sportsmen, 100 members of "various associations",etc., to assuage the public that you were not intentionally skewing the results? This seems impractical, at best. |
|
| |
Last week, for example, I juried a case where the defense effectively argued that 100 randomly selected persons sampled was nowhere close to a representative sampling when it came to gauging impairment levels for DUI purposes amongst American adults with over .08% Blood Alcohol Content, regardless of the standards set by statute and debated interminably in courts of law for the last thiry years. (Sidebar: The Defendant essentially won his assertation and beat the charge.) |
|
| |
I sincerely doubt that you could quickly put together any group for the purposes of the media that would be timely enough, large enough and "pure" enough to give your statistical sampling validity under close scrutiny. If the polling company is making donations of any kind, then that only makes the results derived even more suspect, regardless of charitable intentions. |
|
| |
hippo: Thanks for the anno. It helps expain the idea more fully. |
|
| |
jurist: The idea is to recruit these people beforehand and have between one and three questions waiting for them when they log on. If this is a painless task hopefully there would be a significant population who logged on weekly and a samller population who logged on daily. It is this population that would be most useful for rapid response polls. |
|
| |
jurist, Rods: Suspect stats and misuse of address information is rife in almost any field. You'd have to deal with this here as anywhere else. I could into detail but then I'd be breaking the 'Sealy Law' of 90% of the text here being mine. |
|
| |
[jurist: Not hungry, merely surprised at the lack of fishbone from Rods after a mainly negative annotation. Practicality, as always, is appreciated in everything but humour and daydreams.] |
|
| |
St. Threef, if you're hungry for fish, I'm sure we can accomodate you. In the meantime, we were just trying to be practical. |
|
| |
I've discussed the same basic idea with others, but an unscientific poll shows Media prefers Alarmist approach. Perhaps you may wish to undertake a more scientific approach regarding that aspect - Media. See what results you get - you would certainly need something of that type in order to help argue your own 'case' should you propose the original idea in a venture of $ort$. |
|
| |
I know what you mean. I thought about paying paying people directly but that would encourage people to create opinions where they have none and for some create multiple accounts with automated 'bots providing answers. |
|
| |
No money = people with an axe to grind.
Money = fraudsters. |
|
| |
I figure a small charitable contribution might provide some middle ground. |
|
| |
So, perhaps the upshot of all this discussion is that you should not poll "talking head" specialists who may (or may not) have a professional stake in the results of your poll, but instead should only poll people whose ideas have never before been recorded. I think Appalachia Bell is going to like this idea for improving their business prospects. |
|
| |
How about this:
Allow people to sign up (volunteer) in advance. When people sign up, they indicate their (one) area of specialization. When a poll is required, a random subsection of those with the appropriate specialization are chosen and sent an e-mail containing the poll question(s). As not everyone who is queried will respond we can use a larger-than-necessary sample. |
|
| |
Since we're choosing a random sampling of available experts, those with an 'axe to grind' will be accounted for in the overall results (i.e. They won't get to vote on every issue). If people want to claim specialization in more than one subject they can register with a different e-mail account. Payment shouldn't be necessary - everyone has an opinion and most want it to be heard. |
|
| |