If you're not a regular reader of the Telegraph online, then you've been missing out on an interesting sequence of events. See Damon Kiesow's "Web Notes" for all the details but in essence the story was that there was an attempt -- fairly quickly discerned -- to stuff the electronic ballot box in a web poll the Telegraph was running.

This incident brings to the foreground some issues that have been simmering since the first web polls appeared only a few years ago.

First, the Telegraph, as do many news operations that run web polls, warns that the results are not scientific because it is not a random sample of the public but rather a snapshot of people who read the Telegraph online and volunteer to be in the poll. The Telegraph also notes that a "savvy user" can get around the safeguards to prevent multiple votes by the same person -- which is apparently what happened here. But it was the easiest type of web poll manipulation to detect because the "votes" all came from the same network domain and came in a large volume in a short time. At the time of the incidents that are related in the "Web Notes" blog, the link to the Telegraph's disclaimer had disappeared from the Telegraph home page - apparently the victim of a redesign of the page. A link to the disclaimer is now on the home page, directly underneath the poll question.

If you click on the link, a popup window shows you the disclaimer. But I wonder how many people ever bother to click the link? And, if you have set your web browser to block ALL popups, do you see this window? Today the column where the poll results appear has empty space at the bottom. Maybe the disclaimer should pop into view below the poll results that come up when you participate in the poll. That way, everyone who takes the trouble to respond to the poll would see this warning.

But more troubling, I think, are incidents where groups on one side of an issue or another try to organize big numbers of like-minded people to go to a web site and vote in a poll. This really came into its own in the 2004 election, where emails flew around the internet urging people to vote in "instant polls" that were conducted by major news organizations after each of the Presidential debates. If each vote is by a different person, even if they have been urged by an email to do so, is that stuffing the ballot box or not? I think you can argue it either way.

More generally, for some time now readers at freerepublic.com, a highly popular website, have often responded to appeals to go to particular web sites to vote in web polls. This has happened often enough that online polls that are subjected to this kind of treatment are sometimes said to have been "freeped."

Clearly a good number of people think that reporting of the results of these web polls can sway public opinion - otherwise, why bother with emailing people to get them to respond to the polls? And indeed, you do hear the results of web polls reported, mostly on TV. Sometimes there's a brief disclaimer noting that the results are "not scientific," but I suspect that garners as little attention as most other instances of "the fine print." For example, would diet programs and diet pill makers buy ads that have to state "results not typical" if they weren't fairly sure that most people will skip right over that crucial point? Same thing with saying a web poll is "not scientific" - I suspect most viewers skip right by that and only absorb, "65% of people think x." So there's the danger.

The recent incidents affecting the Telegraph's poll show what the paper's disclaimer has always stated: A "savvy user" can skew, or attempt to skew, the poll's results fairly easily. In this case, I thought the paper handled the whole issue with grace and good humor.

Finally, while these recent attempts were rather crude and quickly detected, they do illustrate a lesson that every critical reader or listener should take to heart: Web polls are a way for readers to express their opinions. They don't prove much beyond that.