 |
PM |  | |
Quote | | |
Reply | | |
Full Topic | |
| Joined: | Mon Nov 21st, 2005 |
| Location: | California USA |
| Posts: | 6399 |
| Status: |
Offline
|
| Mana: |    |
If anyone has studied statistics, then, they may understand terms called random sample and representative sample. This web site (my site) is open on the World Wide Web.
I send out messages to people that link back to this web page and therefore the poll here is not statistically random. I do not; however - exclude any one from voting.
Which tactic will alter the accuracy (validity) of the representative sample? In other words; if someone were to exclude all the unwanted answers, then, only the wanted answers will remain. If, on the other hand, someone selected a few of the participants in the poll but did not exclude anyone based upon any judgment whatsoever, then, will the excluding tactic work to gain the desired poll numbers better than the volunteer tactic?
There is no question in my mind that my poll is affected by my personal selection of people receiving invitations to my site. I don’t, for example, contact convicted murders in prison because they are convicted murderers in prison and invite them (because they are convicted murderer in prison) to view this site and perhaps vote on the poll.
My poll is a statistical measure. The actual measurement is ambiguous and not specific. In other words I can’t accurately predict the winner of the presidential race with this poll. I can however affect someone who is on the fence wondering if their vote will be wasted.
Look at the numbers. That is akin to the “Who wants to be a Millionaire” game show where the contestant ‘polls’ the audience before answering the million dollar question and the audience votes 90 percent for answer A.
What does the potential Millionaire pick for an answer?
Torturing mass murderers?