At about 2:00 PM on April 6, 2007, I submitted a comment to a Tancredo blog post that was driving people to a Laura Ingraham poll (April 4, 2007). Here is my comment:

“These polls have no scientific value. They are self-selected samples and therefore reflect only the views of the limited number of people who take the time to respond to them.

Since they do not employ a probability sample, the results cannot be inferred to anyone who did not take the poll, let alone to the voting public.

Generally, these polls attract a biased sample that leans outward toward the extremes.”

And that is not to mention whatever bias already exists in Laura Ingraham’s audience.  All this survey can do is measure how many Laura listeners and Tancredo-driven supporters clicked through on the poll.  That is it.

The stated code of conduct for comment approval, according to the site is “In an effort to maintain civility and good order, all comments will be approved before permanently posted to the blog.” 

My comments, though critical, are entirely grounded in the Central Limits Theorem, which governs the use of samples for statistical inference.  You can exercise your own judgment as to whether they meet the stated code or not.  I think they do.

Let’s see if it gets approved for posting.  It is 1:09 PM on Saturday, April 07, 2007.