Favorite Files

No favorite files added yet

Recent Posts

1. Comment - The red/blue split for online voters is not 50/50, survey finds

   (Jun 19, 2008 - 3:47 PM)

   This series of blogs got started basically about the accuracy of scientific polls. If sample subjects are chosen truly RANDOMLY, a poll can predict reality with high probability. The crux is how subjects are chosen. This is not a trivial concern in polling. A well thought out protocol as to how individuals are chosen, how to breach the aim of the poll to the respondent, how to treat the answers received, what to do about "no opinion" answers, how to try to prevent as much as possible respondents giving answers opposite to their real intentions by follow up questions, etc. etc...
   
   I know that polls can be designed to prove almost any point. These are not "scientific" polls, as the outcome is more or less predetermined.
   
   During this time many feelings are running high concerning the directions our country may be taking during the near future. If we all try to learn as much as possible about candidates, use our God-given intelligence to try to figure things out, and, most importantly, don't impune others with evil motives, the future will surely be bearable

2. Comment - The red/blue split for online voters is not 50/50, survey finds

   (Jun 19, 2008 - 2:45 PM)

   FOXFYRE, YOU ARE SUCH A SWEET PERSON. WE SHOULD HAVE MORE OF YOUR KIND. LSTEN TO "FAUX NEWS" -- BARE AND BIGOTED. BY THE WAY HAVE YOU EVER HEARD OF AN OXYMORON. I AM READY TO BE INSULTED AGAIN.

3. Comment - The red/blue split for online voters is not 50/50, survey finds

   (Jun 18, 2008 - 3:03 PM)

   Don't confuse me with your opinions. I know that I am right and you are wrong. (How do you like them apples? It sure beats thinking for yourself) Signed: Eggheads for Obama

4. Comment - The red/blue split for online voters is not 50/50, survey finds

   (Jun 18, 2008 - 12:20 PM)

   Sorry to bust your bubble "FOXFYRE", but I was born six months after Hitler came to power and grew up in the shadow of the vilest group of racists, sexists and anti-homosexualists the world has ever seen. To compare anyone today with anyone from that time of that persuasion is tantamount to character assassination and they should have their contribution deleted for foul language and personal attacks.

5. Comment - The red/blue split for online voters is not 50/50, survey finds

   (Jun 17, 2008 - 9:06 PM)

   The one very important time when scientific polling didn't work was during the 2000 presidential campaign. Scientific polling can only predict a range of values for a variable, say, those planning to vote for George Bush. Think of the entire electorate at a certain time as a population of millions from which you RANDOMLY pick a about 12 hundred individuals. This would need to be done state by state due to the electorial system. If almost equal numbers were planning to vote for Bush as were planning to vote for Gore across the population your statistical study would reveal nothing except that it "was too close to call".