CCP Shenanigans on CNN Poll

On Thursday (2/18/2009) CNN reported the results of a poll they conducted with Opinion Research Corp. It ran between Feb. 12 and 15 (basically Valentine's Day weekend) and contacted 1,023 Americans by telephone.

They then reported, "Nearly three-quarters of all Americans think Tibet should be an independent country, according to a new national poll."

Hooray. Yes indeed, "Free Tibet" has become a favored cause, slogan, and bumper sticker in America.

On Monday night, CNN's Larry King interviewed the Dalai Lama, the spiritual leader of Tibetan Buddhism and previously the head of state of the Tibetan theocracy (before invasion from China, which occupied Tibet increasingly from 1950-1959).

The Dalai Lama himself was quick to correct Larry King, saying that he does not seek independence. Instead, he seeks the Middle Way which would give Tibet genuine autonomy within China. (China would be externally federal but internally confederate, keeping the tasks of defense and foreign relations.)

Nuances aside, CNN placed a box poll on the Larry King web page. It's question: Should Tibet be independent? Because of only two choices, Yes and No, the question was coarse-grained, with no room allowed for nuanced replies.

Then the trouble began.

I visited that poll on Sunday night, Feb. 21. At that time, I seem to remember that results were something like 83 - 85% yes.

These box polls on the web are not scientific -- they reflect opinions of visitors to that particular web site and the subset of those who choose to reply. So, there is a sampling error built into any web poll. Respondents will have the demographics of that site's web surfers, not the public at large.

More trouble becomes possible if the web site does not require a log in, and allows people to vote more than once. Then, any anonymous prankster can move the results one way or the other, simply by flooding the poll with bogus votes.

So, the Free Tibet community was able to tweet and urge supporters to go there and vote for Tibetan independence. AND, the Chinese Communist Party was able to tweet and urge nationalist Chinese to go there and vote against Tibetan independence.

On Monday, in advance of the Dalai Lama / Larry King show, came this tweet (from a leader of Students for a Free Tibet):

Please vote on CNN's Larry King show page: Should Tibet be independent? 53% say yes. http://www.cnn.com/CNN/Programs/larry.king.live/

Try to remember, I saw it when it was 83% yes. (And remember, the more scientific poll of CNN/Opinion Research said 75% yes.) This new tweet reflects 53% yes -- a wide swing over night.

On Facebook came the replies to that tweet:

(...) is the Chinese Gov't hacking - again? Google, CNN...

(...) so weird. this morning it was about 80% yes.

(...) It was 83% yes when I voted yesterday, but now it's 44% - you'd think someone was trying to rig the vote, hmmm...

(...) its 44% yes and 56% no now...never knew that CNN is this popular in china. FREE TIBET!

(...) Ha Ha Ha China got a long hand in CNN too.

(...) Wow this suck, all yesterday and this morning at 88% yes, Well not surprised who did this. You can vote as many times as you want, vote a lot!

(...) Within 24 hrs, there were only 800 votes with 88% Yes and 12% No. Now its more than 4000. I am seeing 2000 votes in two hours!! Regular chinese citizens are sleeping now, the votes we are seeing is the propaganda work by chinese govt.

--All of the above was before Larry King Live put the Dalai Lama on the air. I became worried that King might "report the results" of this poll, with its faulty results.

I saw the broadcast, and thankfully the poll did not get mentioned.

After the broadcast, right now I'm looking at the poll. From 33,476 votes, the poll is REVERSED completely from when I first saw it. It says Yes - 15%, No - 85%. There are 28,600 "No" votes -- and I believe, most of those are over votes from the Communist Party.

As we will remember, when there was a more scientific poll, with controlled sampling and demographics, CNN/Opinion Research found Americans to be 75% Yes on Tibetan independence.

So, the Larry King poll reflects nothing scientific, and even in its own results box, the web page says: This is not a scientific poll.